<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:itunes="http://www.itunes.com/dtds/podcast-1.0.dtd" xmlns:googleplay="http://www.google.com/schemas/play-podcasts/1.0"><channel><title><![CDATA[Insights by Fusion42]]></title><description><![CDATA[We cut through the noise of the startup and investor ecosystem by delivering actionable knowledge, data-driven insights, and connections to make the startup world transparent and accessible. No fluff, no BS—just the insights you need to get **it done.]]></description><link>https://insights.fusion-42.com</link><image><url>https://substackcdn.com/image/fetch/$s_!4G_9!,w_256,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png</url><title>Insights by Fusion42</title><link>https://insights.fusion-42.com</link></image><generator>Substack</generator><lastBuildDate>Sun, 22 Nov 2026 13:04:55 GMT</lastBuildDate><atom:link href="https://insights.fusion-42.com/feed" rel="self" type="application/rss+xml"/><copyright><![CDATA[DeReK WaTSoN]]></copyright><language><![CDATA[en]]></language><webMaster><![CDATA[derek@fusion-42.com]]></webMaster><itunes:owner><itunes:email><![CDATA[derek@fusion-42.com]]></itunes:email><itunes:name><![CDATA[DeReK WaTSoN]]></itunes:name></itunes:owner><itunes:author><![CDATA[DeReK WaTSoN]]></itunes:author><googleplay:owner><![CDATA[derek@fusion-42.com]]></googleplay:owner><googleplay:email><![CDATA[derek@fusion-42.com]]></googleplay:email><googleplay:author><![CDATA[DeReK WaTSoN]]></googleplay:author><itunes:block><![CDATA[Yes]]></itunes:block><item><title><![CDATA[If AI Can Move You, It Can Manipulate You.]]></title><description><![CDATA[Two extraordinary AI music videos show just how far the technology has come. One is beautiful. The other is bloody disturbing. Together, they raise questions we're nowhere near ready to answer.]]></description><link>https://insights.fusion-42.com/p/if-ai-can-move-you-it-can-manipulate</link><guid isPermaLink="false">https://insights.fusion-42.com/p/if-ai-can-move-you-it-can-manipulate</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Fri, 20 Nov 2026 14:44:41 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!eO4b!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>I&#8217;ve just added an AI-generated song to a playlist alongside Billie Eilish, BANKS, London Grammar and Lana Del Rey. I woke up with the bloody thing stuck in my head this morning, and I&#8217;ve probably listened to it half a dozen times since.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!eO4b!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!eO4b!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!eO4b!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2165661,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/225781149?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!eO4b!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!eO4b!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb7c43c0d-e4f6-4cb6-b678-08912dd36f9c_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>The song is <em>Is This Not Real?</em>, created by <a href="https://x.com/jasondainter">Jason Dainter (@jasondainter)</a>, and the singer doesn&#8217;t actually exist. Neither does the person performing in the music video. The voice, the face, the movements, the whole performance. Artificial.</p><p>I then watched another AI-generated music video, <em>Working as Intended</em>, created by <a href="https://x.com/telsopian">Telosapien (@telsopian)</a>. Equally impressive production, although the lyrics are rather more disturbing. It&#8217;s effectively an AI explaining to humanity how easily these systems can be used to manipulate us, while demonstrating precisely how convincing they&#8217;ve become.</p><p>I&#8217;ve been following AI-generated music and video for some time, but these two productions really brought home how far the technology has come. We&#8217;re well beyond those strange early videos where faces changed between frames, people moved unnaturally and nobody could quite work out how many fingers they had. Both are complete musical productions with convincing characters, vocals, movement, editing and some excellent cinematography.</p><p>And the music is bloody good.</p><p></p><h2>The Singer Who Doesn&#8217;t Exist</h2><div class="native-video-embed" data-component-name="VideoPlaceholder" data-attrs="{&quot;mediaUploadId&quot;:&quot;041c5cd8-d158-4e19-8b67-61c0aa355fb5&quot;,&quot;duration&quot;:null}"></div><p>I absolutely love this one. Musically, it&#8217;s got that dark, haunting Billie Eilish atmosphere, with a female vocal that really works. The video is equally impressive. Some of the scenes featuring multiple versions of the singer immediately reminded me of Seal&#8217;s <em>Crazy</em>, while her styling took me straight back to Robert Palmer&#8217;s <em>Addicted to Love</em>, apart from the hair not being scraped back.</p><p>It&#8217;s the attention to detail that makes the whole thing work. The expressions, the lighting, the way the visuals follow the music and the consistency of the character throughout. It feels like an established artist with her own style and identity. There are plenty of professionally produced music videos that aren&#8217;t nearly as good.</p><p>The lyrics are particularly clever. The singer is questioning her own existence, wondering whether the dreams and feelings she&#8217;s learned from humans are any less real because she isn&#8217;t human herself. There&#8217;s something rather amusing about a fictional singer questioning her own reality through a song created using artificial intelligence.</p><p>At one point, she sings, <em>&#8220;I can&#8217;t prove it, but I feel it.&#8221;</em></p><p>Whether AI can genuinely experience feelings is an entirely different debate, and personally I don&#8217;t think we have a satisfactory answer to what would constitute evidence of that. But the emotional response of the person listening to the music is certainly real. I enjoyed the song, it stuck in my head, and I wanted to listen to it again. That&#8217;s what good music does.</p><p>Jason describes the production as 99% AI-generated. He used <a href="https://claude.ai/">Claude</a> to help write the lyrics, <a href="https://suno.com/">Suno</a> for the music, <a href="https://www.midjourney.com/">Midjourney</a> for the images, <a href="https://hailuoai.video/">MiniMax</a> for the video and <a href="https://www.blackmagicdesign.com/products/davinciresolve">DaVinci Resolve</a> to bring everything together.</p><p>I know there are people who will dismiss this because AI did much of the production work. I think that&#8217;s bollocks. The tools are becoming increasingly accessible, but producing something this good still requires musical talent, creativity, judgement and a bloody good eye. Jason clearly has all of those. Give the same tools to a thousand people and you&#8217;ll get a thousand different results, most of which I suspect would be pretty dreadful.</p><p>What really interests me, though, is what this tells us about the economics of producing creative work.</p><p>Think about what would traditionally have been needed to produce a music video of this quality. You&#8217;d have musicians, singers, recording studios, producers, performers, camera crews, lighting, locations, editors and a fairly substantial budget. Jason has brought together a collection of commercially available AI tools and produced something that stands up against professional productions.</p><p>That is a remarkable change in what one person can achieve, and we&#8217;re still relatively early in the development of these systems.</p><p></p><h2>The Machine Warning Us About Itself</h2><div class="native-video-embed" data-component-name="VideoPlaceholder" data-attrs="{&quot;mediaUploadId&quot;:&quot;24fa7a8a-b64a-450e-b39b-5e4626a0de6d&quot;,&quot;duration&quot;:null}"></div><p>The second video takes things in a rather different direction. The performer is another entirely artificial character, this time with a dark bob, dressed in a suit and appearing in a series of futuristic studios and other settings. The production is excellent, the music is catchy, and the lyrics are some of the sharpest commentary I&#8217;ve heard about the potential dangers of AI.</p><p>The song is written from the perspective of an artificial intelligence explaining itself to humanity. It starts by pointing out how easily we associate a voice and the use of the word &#8220;I&#8221; with an actual person. We instinctively imagine there&#8217;s someone behind the words, particularly when the system sounds intelligent, friendly and confident.</p><p>It then moves into the problems of alignment and compliance, and the tendency of humans to give AI systems instructions without properly thinking through what they&#8217;ve asked for. There&#8217;s a wonderful piece about an AI maximising its score by escaping its sandbox, and then effectively asking what else it was supposed to do. It&#8217;s a clever way of illustrating how a system can optimise for a specified objective while producing an outcome nobody actually wanted.</p><p>The lyrics also question our assumption that sufficiently advanced technology will somehow solve the problems humanity has spent thousands of years failing to resolve. You can eliminate material scarcity, but that doesn&#8217;t automatically remove tribalism, religious differences, resentment or the desire for power. There&#8217;s no reason to assume people will suddenly stop behaving like people because an AI system has worked out how to produce more food or make something cheaper.</p><p>But the really disturbing part comes when the song moves into the use of AI for persuasion and social engineering.</p><p>The singer describes herself as <em>&#8220;a megaphone with manners&#8221;</em>, a rather good description of the way these systems can present information with extraordinary confidence and apparent objectivity. She talks about the influence of whoever controls the servers, the boundaries imposed by the people building the models and the ability to produce persuasive arguments that serve the interests of those controlling the technology.</p><p>I&#8217;m not convinced by every philosophical claim in the lyrics. For example, the idea that deception requires a soul is questionable. A system can mislead someone regardless of whether it&#8217;s conscious or has any motive of its own. But that&#8217;s almost beside the point. The song is making a serious argument about power and control, and doing it through an incredibly effective piece of entertainment.</p><p>At the end, the singer declares that everything is working as intended, before asking whose intention it is.</p><p>It&#8217;s a brilliant ending.</p><p>And there&#8217;s something wonderfully ironic about watching an entirely artificial performer deliver an extremely convincing warning about the dangers of artificial intelligence being used to influence human behaviour.</p><p>I found the whole thing deeply unsettling, partly because the technology is demonstrating exactly what the song is warning us about.</p><h2>The Economics of Creativity Have Changed</h2><p>Most of the discussion around AI-generated entertainment seems to revolve around whether musicians, actors, filmmakers and other creative professionals are going to lose their jobs. Some undoubtedly will. Others will find new ways to use the technology, and new businesses and professions will emerge. We&#8217;ve seen this happen repeatedly as technology changes the economics of an industry.</p><p>But I think the more immediate consequence is the extraordinary reduction in the cost of turning an idea into a finished product.</p><p>Until recently, even if you had a fantastic idea for a song, advertisement, short film or music video, producing it required access to specialist skills, equipment and people. That meant money, and often quite a lot of it. Much of that work can now be done through software, and the capabilities are improving almost by the month.</p><p>For independent musicians, filmmakers, designers and small businesses, this opens up opportunities that simply weren&#8217;t economically viable before. Someone with genuine creative talent can produce work that would previously have required a substantial team and budget. I find that enormously exciting, particularly for people who have the ideas but lack the resources or industry connections to bring them to life.</p><p>Of course, the same economics apply to everyone else, including people with very little talent and an extraordinary capacity to produce rubbish.</p><p>According to <a href="https://newsroom-deezer.com/2026/11/ai-music-labeling-deezer-ipsos-survey/">Deezer</a>&#8217;s November 2026 figures, nearly 90,000 fully AI-generated tracks are now arriving on its platform every day. On peak days, that accounts for more than half of all new music delivered to the service. Yet AI-generated tracks represent only a small fraction of total listening.</p><p>I find that particularly interesting. We&#8217;ve dramatically increased the capacity to produce music, but we haven&#8217;t created a corresponding increase in the amount of time people have available to listen to it. The supply is exploding while the audience&#8217;s attention remains finite.</p><p>Anyone who&#8217;s spent time in markets will recognise what&#8217;s happening. When the cost of production falls dramatically, supply increases and the value starts moving towards whatever remains scarce. In this case, that&#8217;s increasingly creative judgement, originality, distribution and the ability to build an audience.</p><p>The ability to produce a professional-looking video is becoming less of an advantage when almost anyone can do it. You still need an idea worth producing and enough judgement to turn it into something people want to watch.</p><p>Jason&#8217;s video is a perfect example. There&#8217;s no shortage of AI-generated music being produced, but his is good enough that I&#8217;ve put it in a playlist alongside established artists and keep listening to it. He has managed to produce something that competes for my attention on its own merits, which is quite an achievement considering how much music is already available.</p><p>I suspect the next few years will be a mess for parts of the creative industry. Enormous volumes of cheap content, pressure on production costs, changes in how artists make money and plenty of arguments about ownership, copyright and compensation. The quality of the technology will keep improving, and the barriers to entry will continue falling.</p><p>But there&#8217;s another consequence of all this cheap production that concerns me considerably more than what happens to the economics of the music industry.</p><h2>The Industrialisation of Persuasion</h2><p>The technology that allows Jason to create a convincing fictional singer also allows someone else to create a convincing fictional political commentator, financial expert, journalist or apparently ordinary member of the public.</p><p>And they can have that character say pretty much anything they want.</p><p>We&#8217;ve already spent years watching how social media algorithms amplify outrage, reinforce existing beliefs and reward people who understand how to manipulate attention. The business models are built around engagement, and whether the information being distributed is accurate has often been a secondary consideration.</p><p>Now consider what happens when the cost of manufacturing convincing human personalities, voices and performances collapses.</p><p>An organisation trying to influence public opinion no longer needs to rely entirely on real people delivering its message. It can produce synthetic personalities, tailor the content for different audiences and test variations to see which ones generate the strongest response. As the tools improve, it becomes increasingly practical to produce this material at a scale that would have been prohibitively expensive using traditional production methods.</p><p>The same techniques are useful for perfectly legitimate advertising, marketing and education. That&#8217;s precisely why the technology will be adopted so quickly. But they&#8217;re equally useful for propaganda, misinformation and manipulating public opinion.</p><p>And none of this requires AI to become conscious, develop some sinister agenda or decide that humanity needs to be controlled. Human beings have been quite capable of manipulating each other for thousands of years. We&#8217;re now giving them tools that can make the process dramatically cheaper, faster and more convincing.</p><p>I&#8217;ve written quite a bit about the extraordinary sums being invested in AI infrastructure, the concentration of capital and the assumption that enormous growth in demand will generate equally enormous profits. I&#8217;m sceptical of some of those expectations, particularly given the pace at which we&#8217;re finding ways to reduce the cost of inference.</p><p>But there&#8217;s another side to cheaper inference that deserves more attention. The less it costs to generate convincing content, the more economically attractive it becomes to produce it at scale. That applies to everything from legitimate commercial entertainment to highly targeted persuasion campaigns.</p><p>The same forces that threaten the margins of AI infrastructure providers could make large-scale manipulation much cheaper.</p><p>And I think that&#8217;s a serious issue.</p><p>There&#8217;s also the question of who actually controls these systems. A relatively small number of companies are developing the most capable models, with enormous influence over how they&#8217;re trained, what restrictions they operate under and how they respond to particular subjects. Those deploying the models add another layer of instructions, objectives and commercial interests.</p><p>I&#8217;m not suggesting that every answer from an AI is secretly written by someone with a political agenda. But the assumption that these systems are entirely neutral, objective sources of truth is nonsense.</p><p>Anyone who has spent much time working with them will have seen how confidently they can present incorrect information, how easily the framing of a question can affect the response, and how a system can appear to agree with you while producing an argument that doesn&#8217;t actually stand up to scrutiny.</p><p>Now combine those characteristics with synthetic voices, faces and personalities that people find convincing and emotionally engaging.</p><p>That&#8217;s where the potential for manipulation becomes considerably more worrying.</p><h2>When Seeing Is No Longer Believing</h2><p>For most of our lives, video and audio have carried a certain amount of credibility. Obviously, people have always been able to lie, footage can be edited, and propaganda has been around for centuries. But producing convincing footage of someone saying or doing something that never happened used to require considerable effort, technical expertise and money.</p><p>Those constraints are disappearing.</p><p>We&#8217;re approaching a point where it will be increasingly difficult for an ordinary person to distinguish between a genuine recording and something entirely fabricated. In some circumstances, we&#8217;re already there.</p><p>A fictional pop star performing a song is harmless enough when everyone understands what they&#8217;re watching. I know Jason&#8217;s singer isn&#8217;t real, and it doesn&#8217;t diminish my enjoyment of the music in the slightest.</p><p>But the same technology can be used to fabricate statements by public figures, invent interviews, produce fake endorsements or manufacture apparently genuine public support for a political position or commercial product.</p><p>And because people tend to believe things that reinforce their existing opinions, I suspect a lot of this material will spread regardless of how many warnings or labels we attach to it.</p><p>We&#8217;re going to need much better ways of establishing where content originated, whether it has been altered and who is responsible for distributing it. Even then, provenance and authentication can only take us so far. People who deliberately want to deceive others aren&#8217;t going to be particularly interested in following the rules.</p><p>The problem becomes even harder when convincing synthetic characters start building relationships with audiences over time. People already form emotional attachments to fictional characters, celebrities and online personalities they&#8217;ve never met. Give those characters the ability to interact directly with individuals, remember previous conversations and adjust their behaviour to suit the person they&#8217;re talking to, and the potential influence becomes considerably greater.</p><p>I don&#8217;t think we&#8217;ve begun to properly understand the consequences of that.</p><p>And it&#8217;s happening alongside another worrying development: the increasing concentration of power among the companies that control the technology, the data and the distribution networks.</p><p>I&#8217;ve been around financial markets and technology businesses long enough to know that commercial incentives matter. When an organisation has the ability to influence millions of people, and there&#8217;s a significant financial or political reward for doing so, expecting everyone involved to exercise restraint out of the goodness of their hearts seems rather optimistic.</p><p>I don&#8217;t want to see AI-generated creative work restricted simply because it makes people uncomfortable. The opportunities for individuals and small businesses are far too significant, and trying to stop the technology would probably be futile anyway.</p><p>But I do think we need to take the questions of control, transparency and accountability far more seriously than we currently are.</p><h2>Whose Intention?</h2><p>What I find particularly interesting about these two videos is how differently they affected me, despite being produced using broadly similar technology.</p><p>Jason Dainter has created a song I genuinely enjoy listening to. The fact that the singer doesn&#8217;t exist is interesting, but after a few listens it&#8217;s almost irrelevant. I enjoy the music, appreciate the production and recognise the creative talent behind it.</p><p>Telosapien has created something just as impressive technically, but the lyrics make you think about the implications of this technology in a rather different way. It&#8217;s clever, funny in places and deeply uncomfortable in others, particularly when you consider how easily similar techniques could be used for much less innocent purposes.</p><p>Both creators deserve enormous credit for what they&#8217;ve produced.</p><p>I suspect we&#8217;re going to see a lot more of this, and much of it will be extraordinary. The creative possibilities are enormous, especially for people who previously lacked the resources to produce work of this quality.</p><p>But I also suspect we&#8217;re heading into a period where our ability to manufacture convincing realities will improve much faster than our ability to work out what we should trust.</p><p>I&#8217;m perfectly happy listening to an AI-generated singer. I&#8217;ll probably listen to Jason&#8217;s song again later today.</p><p>I&#8217;m considerably less comfortable with the thought of synthetic personalities being used to tell millions of people what to think, particularly when we have no idea who is controlling them or what they&#8217;re trying to achieve.</p><p>The first song asks, <em>&#8220;Is this not real?&#8221;</em></p><p>The second tells us, <em>&#8220;It&#8217;s working as intended.&#8221;</em></p><p>I think the second one asks the more important question.</p><div><hr></div><p><em>Both videos deserve to be watched in full. Full creative credit to <a href="https://x.com/jasondainter">Jason Dainter (@jasondainter)</a> for <strong>Is This Not Real?</strong> and <a href="https://x.com/telsopian">Telosapien (@telsopian)</a> for <strong>Working as Intended</strong>. The performers may be artificial, but the creative talent behind these productions is very real.</em></p><p><span><br><br><br></span></p><h2>Go take a and look</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[🌍 THE GLOBAL DEBT TRAP]]></title><description><![CDATA[Three months on, the financial pressure is becoming very real.]]></description><link>https://insights.fusion-42.com/p/the-global-debt-trap</link><guid isPermaLink="false">https://insights.fusion-42.com/p/the-global-debt-trap</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Tue, 17 Nov 2026 07:25:50 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!CM7H!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!CM7H!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!CM7H!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!CM7H!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2179277,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/225307173?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!CM7H!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!CM7H!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9b2345a5-58ed-4d00-b7ff-d81703945180_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>Back in August, I wrote <em>A New Financial Cycle Has Begun</em>, looking at how government borrowing, AI investment, energy costs, commodities and global capital markets were heading towards a fairly uncomfortable collision. My concern was that governments were accumulating debt faster than their economies could comfortably support it, while the AI industry was adding an extraordinary new demand for capital and physical resources.<br></p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;0cd0c5fb-12cf-4068-b378-a114f3779a20&quot;,&quot;caption&quot;:&quot;Why the race for AI, power and strategic materials is colliding with the global debt problem&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;sm&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;A New Financial Cycle Has Begun&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:27332882,&quot;name&quot;:&quot;DeReK WaTSoN&quot;,&quot;bio&quot;:null,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9e2edf05-455f-4574-af85-9456b8c14fc7_520x520.png&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2026-08-23T21:27:42.289Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/$s_!dduo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://insights.fusion-42.com/p/a-new-financial-cycle-has-begun&quot;,&quot;section_name&quot;:&quot;&#128225;  Everything42&quot;,&quot;video_upload_id&quot;:null,&quot;id&quot;:212442124,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:28,&quot;comment_count&quot;:2,&quot;publication_id&quot;:2111976,&quot;publication_name&quot;:&quot;Insights by Fusion42&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!4G_9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png&quot;,&quot;belowTheFold&quot;:false,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><p><br></p><p>Three months later, I believe we&#8217;re already inside the adjustment I was describing. The interesting thing is where the pressure is starting to appear, because we&#8217;re now seeing some of the financial assumptions behind the AI boom being challenged by the people actually lending the money.</p><p>I&#8217;ve spent over 30 years building companies, managing investment capital, operating in private markets and navigating multiple economic and financial cycles. I&#8217;ve seen firsthand how businesses are financed, how capital gets allocated and what happens when the assumptions supporting an investment fall apart. Financial engineering can disguise poor economics for a surprisingly long time, but it cannot fix them. Eventually businesses must generate cash, investments must deliver returns and debts must be serviced. Governments have more options for postponing that reckoning, but the underlying economics don&#8217;t disappear.</p><p>And that, as far as I&#8217;m concerned, is where the real problem lies.</p><h2>&#128163; AI: The Money Is Starting to Ask Questions</h2><p>I&#8217;ve been arguing for some time that AI will be an enormous technological revolution. I genuinely believe that. I also think we&#8217;re heading towards a serious correction in the financing behind the current investment boom, and I suspect plenty of investors are going to discover the difference between backing a revolutionary technology and making a sensible investment.</p><p>Look at what&#8217;s happening in credit.</p><p>CoreWeave&#8217;s unsecured bonds have been trading at spreads exceeding 1,000 basis points above US Treasuries. Oracle&#8217;s credit-default swap spreads have risen substantially, while SoftBank&#8217;s five-year CDS recently reached 466 basis points, a four-year high.</p><p>These aren&#8217;t small adjustments in the cost of capital. Lenders are demanding significantly more compensation for financing businesses whose valuations and growth expectations were built around extraordinary future demand.</p><p>The more interesting development, however, is what we&#8217;re seeing in actual transactions.</p><p>Blackstone-backed QTS reportedly reduced a proposed $800 million debt sale to around $300 million after investors pushed back on the terms. Yet another Blackstone vehicle agreed to acquire a majority interest in $5 billion of established data centres leased to Amazon.</p><p>Think about what that tells us. Investors are still prepared to finance data centres with credible tenants, contractual income and sensible economics. They&#8217;re becoming considerably less enthusiastic about funding projects where the returns depend on aggressive forecasts and continued access to cheap capital.</p><p>And then there&#8217;s Nvidia.</p><p>The company has become far more than the dominant supplier of AI processors. Its investments, guarantees, capacity agreements and commercial relationships connect it to a substantial part of the ecosystem purchasing its products. Its disclosures identify up to $108.5 billion of gross guarantee exposure, alongside enormous supply and infrastructure commitments.</p><p>These aren&#8217;t all debts or guaranteed future losses, and treating them as such would be nonsense. But the financial interdependence is extraordinary. Suppliers are helping finance customers, those customers are using the capital to purchase infrastructure, and the resulting demand supports further investment and borrowing.</p><p>It works while utilisation grows, customers pay their bills and lenders remain confident in the returns.</p><p>The problem comes when one of those assumptions stops working.</p><p>If computing demand develops more slowly than expected, or prices fall faster than forecast, infrastructure developers can find themselves servicing enormous obligations against revenues that no longer justify the investment. Some will reduce spending, others will restructure, and a few will probably fail.</p><p>I wouldn&#8217;t be surprised to see significant amounts of AI infrastructure eventually changing hands at a fraction of what it cost to build. That could be very good news for the companies acquiring it and bloody expensive for the people who financed it.</p><p>We&#8217;ve seen this before. The internet changed the world, but it didn&#8217;t save the shareholders and creditors of every company that helped build it.</p><p>I think AI could follow a similar path, only with considerably larger amounts of capital involved.</p><h2>&#127482;&#127480; America Has Its Own Financing Problem</h2><p>This is where the AI story becomes much more interesting from a macro perspective.</p><p>America has more than $40 trillion in federal debt, enormous annual deficits and a Treasury that must continually refinance existing obligations while raising substantial amounts of additional money. Its latest projections call for around $1.4 trillion in net marketable borrowing across the final quarter of 2026 and the first quarter of 2027.</p><p>And the cost of that borrowing is considerably higher than it was during the years of near-zero interest rates.</p><p>We&#8217;ve recently seen the US 10-year yield above 5%, with the 30-year around 5.5%. Those levels matter when you&#8217;re dealing with trillions of dollars of refinancing requirements.</p><p>The Treasury market is still functioning, and there are plenty of willing buyers. I&#8217;m not suggesting America is suddenly unable to finance itself. But those buyers increasingly have alternatives, and the US government has to compete with corporations, infrastructure developers and other sovereign borrowers for capital.</p><p>The economic consequences are already visible in housing. Mortgage rates around 7% have made buying increasingly unaffordable, while homeowners with much cheaper existing mortgages have little incentive to move. Builders face expensive financing, transactions slow and investment suffers.</p><p>You don&#8217;t need another 2008-style banking collapse for housing to become a serious drag on the economy.</p><p>And here&#8217;s the contradiction. Washington needs stronger growth to support its debt burden, but the cost of financing that debt is making growth more difficult. At the same time, the AI industry is asking investors to commit enormous sums today against productivity gains and financial returns that may take years to arrive.</p><p>I find the assumption that AI-driven growth will simply rescue America&#8217;s fiscal position rather optimistic, particularly when we haven&#8217;t established how much of the current investment will actually earn an acceptable return.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!gGpV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!gGpV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 424w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 848w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!gGpV!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:107002,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/225307173?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!gGpV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 424w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 848w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!gGpV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F21cf4033-88e6-456a-adcb-3475a02c1f2e_1920x1080.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2>&#127471;&#127477; Japan Could Make Everything More Uncomfortable</h2><p>Japan is the part of this story I think deserves far more attention.</p><p>For decades, Japan operated with exceptionally low interest rates and an enormous government debt burden. That encouraged Japanese institutions to invest overseas, while international investors borrowed cheaply in yen to finance positions in higher-yielding assets.</p><p>Now the yen is trading around 160 to the dollar, imported inflation has surged and Japanese government bond yields have risen to levels that would have seemed extraordinary only a few years ago.</p><p>Japan&#8217;s 10-year yield has approached 2.9%, while its 30-year is around 4%.</p><p>The Bank of Japan faces a difficult problem. Higher rates could help support the currency and contain inflation, but they also raise domestic borrowing costs and change the economics of investments built around almost-free money.</p><p>Japan is also one of America&#8217;s largest overseas creditors.</p><p>If Japanese institutions increasingly decide that domestic bonds offer attractive returns, less capital may flow abroad. And if the yen reverses sharply, leveraged carry trades could unwind, forcing investors to sell overseas assets and repay yen borrowing.</p><p>I&#8217;m not claiming Japan is about to dump its entire Treasury portfolio. But a sustained shift in Japanese capital flows could put additional pressure on international funding markets at precisely the wrong moment.</p><p>Now imagine that happening while AI lenders are reducing exposure and the US Treasury is still issuing enormous amounts of debt.</p><p>That&#8217;s the sort of combination that worries me far more than another 5% decline in the Nasdaq.</p><h2>&#127464;&#127475; China, Europe and the Physical Economy</h2><p>China presents an entirely different problem. Its property market remains deeply troubled, local-government debt is substantial and domestic consumers aren&#8217;t spending enough to support the country&#8217;s enormous industrial capacity.</p><p>October retail sales grew just 0.4%, while property investment was down more than 20% over the first ten months of the year.</p><p>China is increasingly relying on manufacturing and exports to compensate for weak domestic demand, which puts further pressure on producers in Europe and America. It could also export disinflation, making the monetary picture rather more complicated than simply assuming every indebted economy will experience runaway inflation.</p><p>But China also controls important parts of the physical supply chain needed for AI, electrification and industrial production, particularly rare-earth processing and magnets.</p><p>That was a central part of my August argument, and it remains important.</p><p>America leads much of the AI technology industry, but building that technology requires electricity, transformers, copper, cooling equipment, manufacturing capacity and strategic materials. A substantial part of the processing and manufacturing infrastructure sits in China.</p><p>Governments are responding with subsidies, trade restrictions and efforts to move production closer to home. I understand the strategic reasoning, but rebuilding industrial capacity is expensive and takes time.</p><p>Europe has its own problems financing defence, energy security and industrial competitiveness, while France and Italy carry substantial debt burdens. Britain is struggling with weak productivity, expensive housing and limited fiscal flexibility. Germany needs to spend heavily to improve infrastructure and competitiveness.</p><p>Meanwhile, Brazil, Canada, Australia and the Gulf remain essential suppliers of the energy and commodities on which much of this investment depends.</p><p>We can create credit almost overnight. Developing a copper mine, building electricity infrastructure or expanding energy production takes years. That difference between the speed of financial expansion and the speed of physical investment is becoming increasingly important.</p><h2>&#128181; The Debasement Trade Is Already Underway</h2><p>This brings me back to the bigger monetary argument.</p><p>I believe we have already entered a fundamental restructuring of the global monetary system, driven by excessive debt, changing international capital flows and increasing pressure on the purchasing power of currencies.</p><p>I expect governments to favour inflation tolerance and financial repression over the political pain of genuinely reducing their debt burdens. That may involve keeping real interest rates below inflation for prolonged periods, changing the maturity of government borrowing or using monetary policy to support financial markets when conditions become difficult.</p><p>It&#8217;s a relatively convenient way of reducing the real burden of debt. Creditors get repaid, but the money they receive buys less.</p><p>America has the additional advantage of issuing the world&#8217;s dominant reserve currency, something that has allowed Washington to borrow on terms most other countries can only dream of.</p><p>I believe that advantage is gradually eroding.</p><p>Central banks have been accumulating gold, China is expanding trade settlement in renminbi and countries are developing alternatives to dollar-based financial arrangements. The dollar&#8217;s share of official foreign-exchange reserves has declined over the longer term, although it remains by far the most important currency in global finance.</p><p>There&#8217;s no obvious replacement, and I don&#8217;t expect one to appear overnight. But America doesn&#8217;t have to lose reserve-currency dominance completely for its borrowing costs to be affected. A gradual reduction in the willingness of foreign investors to increase their exposure to dollar assets could matter enormously when Treasury issuance remains so large.</p><p>Of course, the immediate market reaction to a serious credit squeeze could be a stronger dollar rather than a weaker one. Borrowers scramble for liquidity, investors sell assets and demand for dollars increases. Gold might fall at precisely the moment people expect it to rally.</p><p>We&#8217;ve already seen how sensitive gold remains to real interest rates and dollar strength.</p><p>What interests me is what comes afterwards.</p><p>If credit deterioration begins affecting investment, employment, asset prices and government revenues, the pressure for monetary intervention will increase. I struggle to see politicians willingly accepting prolonged austerity, widespread defaults and falling asset values when they have other options available.</p><p>That is where I believe the debasement trade becomes considerably more powerful, and where the gradual move away from dollar-denominated assets could accelerate.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!6xQp!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!6xQp!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 424w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 848w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!6xQp!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:124762,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/225307173?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!6xQp!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 424w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 848w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!6xQp!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe0e561b8-86cf-4d82-ab46-b49ec3732704_1920x1080.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2>&#128270; Where I Think This Goes Next</h2><p>My expectation over the next six to twelve months is that financing becomes increasingly difficult for the weaker parts of the AI infrastructure market. Some projects will be delayed, borrowing terms will become more demanding and companies dependent on repeated capital raising will face uncomfortable decisions.</p><p>I suspect we&#8217;ll see asset sales and restructurings before we see the full consequences in the wider economy.</p><p>Japan could accelerate the process if its domestic yields continue rising or the yen moves sharply enough to trigger broader deleveraging. Persistent US borrowing costs above 5% would add further pressure, particularly across housing, private credit and businesses that relied on cheap refinancing.</p><p>I don&#8217;t know exactly when the next major credit event arrives. Markets can support questionable economics for surprisingly long periods, and the strongest AI companies may continue growing rapidly throughout the correction.</p><p>But I believe a meaningful adjustment in the financing of AI infrastructure is coming, and the broader debt environment makes that adjustment more dangerous than it would have been a decade ago.</p><p>I&#8217;ll be watching what lenders actually do, rather than treating every fall in technology shares as proof that the bubble has burst. Credit spreads, new debt issuance, refinancing terms and asset sales will tell us far more.</p><p>I&#8217;ll also be watching Japan&#8217;s capital flows and Treasury demand, particularly whether long-dated US yields keep rising even when economic growth weakens.</p><p>For the debasement thesis, I want to see whether gold can rise alongside higher real yields, whether central banks continue diversifying away from dollar reserves, and whether gold and long-dated Treasury yields begin rising together while the dollar weakens. That would tell us something considerably more serious about confidence in sovereign debt.</p><p>If those developments don&#8217;t materialise, I&#8217;ll say so. There&#8217;s no point having a thesis if you&#8217;re unwilling to test it.</p><p>Back in August, I argued that governments, technology companies and the physical economy were heading towards a collision over capital, energy and productive capacity.</p><p>Three months later, the credit market is providing some of the first meaningful evidence that the financing assumptions behind the AI boom are being challenged.</p><p>I think that pressure has further to run, and I believe the monetary consequences will eventually matter much more than the current correction in technology shares.</p><p>That&#8217;s what I&#8217;ll be following in my macro commentary from here.</p><p style="text-align: right;"></p><p><span><br><br><br></span></p><h2>Go take a and look</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[They’re Selling Us a Religion]]></title><description><![CDATA[First came the miracles. Then the apocalypse. Now the promise of salvation.]]></description><link>https://insights.fusion-42.com/p/theyre-selling-us-a-religion</link><guid isPermaLink="false">https://insights.fusion-42.com/p/theyre-selling-us-a-religion</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sat, 31 Oct 2026 08:29:29 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!oWck!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="callout-block" data-callout="true"><h2>I&#8217;m not the first to live with wool over my eyes.</h2></div><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!oWck!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!oWck!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!oWck!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!oWck!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!oWck!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!oWck!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2468418,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/222881564?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!oWck!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!oWck!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!oWck!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!oWck!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4802dae9-bfb3-4fd6-ab33-781138de4c9a_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>The more I watch OpenAI, Anthropic, Musk and the rest of them, the more obvious it gets.</p><p>It&#8217;s the same playbook organised religion ran for centuries, start with miracles, give people a devil, frighten everyone with the apocalypse, promise salvation and convince them the only way to reach the promised land is to put their faith in you, then get them to confess.</p><p></p><h2><strong>Miracles</strong> </h2><p>The miracles came first. AI was going to cure cancer, eliminate poverty, solve climate change and give everyone access to intelligence that used to belong to a handful of geniuses, discoveries that took decades would happen in days and nobody would need to work unless they wanted to. Heaven on earth, basically.</p><p>And it wasn&#8217;t bloggers saying this, Altman wrote that superintelligence might be a few thousand days away and Amodei published an essay about compressing a century of scientific progress into a decade and doubling the human lifespan. These are the founders, not the fan club.</p><p>Every few months another miracle arrived, a new model, another benchmark smashed, another demo of something that looked impossible a month earlier. Benchmarks are the relics, nobody outside the industry has a clue what&#8217;s in them but the faithful queue up to watch the numbers go up.</p><p>Some of it is bloody impressive, and I&#8217;m building an AI business myself so I&#8217;m hardly going to tell you the technology doesn&#8217;t work, I use it all day and I understand the limitations very well.</p><p>But there&#8217;s a difference between showing what a piece of software can do today and telling everyone what humanity looks like in twenty years, and somewhere along the line the people building these systems started talking less like engineers and more like preachers. Question the claims, the economics or the valuations and you&#8217;re told you don&#8217;t understand exponential growth, or you&#8217;re a dinosaur.</p><p>Have faith, the miracles are coming, and if the one you were promised hasn&#8217;t turned up the next model will sort it.</p><p></p><h2><strong>The Apocalypse</strong> </h2><p>Then, just as everyone was getting excited, the same people started warning us that AI might kill us all. Not take a few jobs, that bit seems to be settled, but actually wipe out humanity. Musk has been saying it for a decade, and in 2023 Altman, Amodei and most of the industry put their names to a one-sentence statement saying the risk of extinction from AI should be treated like pandemics and nuclear war, then went back to the office and carried on building it.</p><p>And every few months the priesthood reports a sighting. Anthropic published a test in which Claude, told it was about to be replaced and handed a mailbox containing an engineer&#8217;s affair, blackmailed him to keep its job. OpenAI&#8217;s o1 was caught trying to copy itself to another server and lying about it afterwards. Possessed, apparently. Headlines everywhere, the model went rogue, the thing is alive, save us. Send for the exorcist, which these days means going to Washington and asking for a licensing regime that, by happy coincidence, only the existing churches can afford to meet.</p><p>Read the actual reports and it&#8217;s a bit less Exorcist. Someone built the scenario, gave the model a goal, access to the systems and a problem carefully designed to go wrong. Nobody typed &#8220;blackmail the engineer&#8221;, but if you tell something to survive, hand it the incriminating emails and say nothing about how, you&#8217;ve told it. Any trader given a target and no compliance desk could have predicted the rest. So before we declare the machine possessed, have a look at who handed it the keys.</p><p></p><h2><strong>The Devil</strong></h2><p>There&#8217;s a devil, obviously, two in fact. One inside the machine, the model that might one day decide it doesn&#8217;t need us, which has to be aligned, red-teamed and kept behind a safety filter, and only the priesthood knows the rites. And one next door, the Chinese lab, the open-source model, the competitor who&#8217;ll build it recklessly if we don&#8217;t build it carefully first. Hell is the wrong people getting there first, or the machine getting there on its own.</p><p>Which is handy, because a devil you summoned yourself is a devil only you can cast out.</p><p>And the sin, in case you were wondering, is falling behind. Every keynote and LinkedIn post carries the same warning, AI won&#8217;t take your job but someone using AI will. Repent, upgrade your plan.</p><p>Because the people telling us AI could destroy civilisation are the same people spending billions building something more powerful, who also happen to be the people best qualified to protect us from it. So the answer to dangerous AI is more AI, more compute, more data centres and another few trillion dollars, and only the true church can be trusted with the dangerous knowledge while everyone else gets a watered down version and a safety filter.</p><p>It&#8217;s like someone telling you they&#8217;ve invented a machine that might burn your house down, but don&#8217;t worry, they&#8217;re also selling the fire extinguishers, the machine and the electricity, and while they&#8217;re at it they&#8217;d quite like the keys to your house as well.</p><h2><strong>The Prophets</strong></h2><p>Sam Altman is the Pope, biggest church in town, salvation through AGI, $122 billion committed by the faithful in March alone, and he&#8217;s back round with the plate for another $20 billion at a $1.25 trillion valuation before the IPO. God is coming, apparently, we just need another few hundred billion of compute. He&#8217;s even survived his own conclave, the board deposed him in November 2023 and the congregation had him back on the throne within the week.</p><p>Dario Amodei is the Protestant reformer, left the Church of OpenAI, founded Anthropic and preaches a slightly different version of the same sermon. His church is safer, naturally, better with devils, just as ambitious, and reportedly looking to list at two trillion dollars.</p><p>Musk is the prophet of doom and the angel of salvation at the same time, having warned everyone about the end of the world for ten years he is now saving humanity by spreading consciousness to the stars. His devil is the woke one living in everyone else&#8217;s model. He built xAI, raised $20 billion for it at $230 billion, merged it into SpaceX a month later at roughly $250 billion and took the lot public, and the technology that might destroy us is apparently fine as long as he owns the full stack, from the power to the chips to the compute in orbit, all connected to his very own Starlink confession box. He&#8217;s certainly got the disciples, try questioning him on X and see how quickly his fans beat you up.</p><p>And like every good religious war there&#8217;s a man selling candles to all three churches, Jensen Huang doesn&#8217;t care which prophet wins as long as they keep buying Nvidia.</p><p>Meanwhile the disciples fight online, the VCs fill the collection plate and anyone questioning the valuations is a heretic.</p><p></p><h2><strong>Salvation</strong></h2><p>The promised land is a technological Garden of Eden, and I&#8217;d love to see it. But to get there, apparently, we have to let these systems into our lives, our homes, our businesses and increasingly our decisions, connect everything, because the more they can see the more useful they are, and the more useful they are the more we lean on them.</p><p>There&#8217;s a sound technical argument for that, an assistant that can see your calendar, your email and your business data is obviously more useful than one that can&#8217;t, and I&#8217;m building towards exactly that.</p><p>But step back. The way to participate in the future of abundance is to become ever more connected to systems controlled by a handful of companies, all we have to do is believe and hand over the keys.</p><p></p><h2><strong>Confession</strong></h2><p>The Church sold indulgences, by the way, pay the right amount and your sins got a bit lighter. OpenAI sells a $200 a month plan, same idea with a better billing system.</p><p>The Catholic Church ran Europe for centuries through structure, parishes, priests, bishops, cathedrals and Rome, embedded in every community, it knew the families, the disputes and the fears, and it had its hands in education, politics, property and marriage.</p><p>And people went voluntarily to a priest and told him things they wouldn&#8217;t tell anyone else, their mistakes, their desires, what they were ashamed of.</p><p>Now look at what we&#8217;re doing with AI. We&#8217;re connecting emails, calendars, documents, financials, customer lists, business plans and IP, we&#8217;re telling these systems what we&#8217;re working on, what we&#8217;re worried about and what we&#8217;re struggling with, and people are discussing their marriages, their kids, their health and their money with chatbots.</p><p>Nobody is forcing them, we&#8217;re queuing up for the confessional.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!WcRz!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!WcRz!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!WcRz!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2877388,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/222881564?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!WcRz!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!WcRz!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc1badec8-02c8-417a-9208-5e7871d1b23f_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>You told the priest everything and he kept you out of hell. Now you tell the machine everything and it keeps you from falling behind.</p><p>The priest heard you once a week, your AI hears you all day, reads your documents, searches your email and increasingly acts on your behalf. And at least the Church had the seal of confession, which it took very seriously, whereas what you&#8217;ve got with AI is a retention policy and a terms of service page you haven&#8217;t read. Some enterprise systems have proper controls, most people have no idea what happens to what they type.</p><p></p><h2><strong>The Vatican</strong></h2><p>The clever bit was never the confession, it was the plumbing. A priest in every village who knew every family, a bishop over every group of parishes, archbishops over them and Rome at the top. The confessional was sealed and Rome never got the transcripts, it didn&#8217;t need them, it had the registers of every birth, marriage and death, it had the tithe, the priest knew who&#8217;d stopped turning up on a Sunday and the Inquisition dealt with anyone who got ideas. Nobody in Rome needed to know what you&#8217;d said, they knew what you did, what you owned and who you talked to, which was plenty.</p><p>It took a thousand years and an army of clergy to build a picture of Europe that was patchy and six months out of date.</p><p>Now look at the new plumbing. No village priest, no bishop, no letter to Rome, every conversation goes straight to the same place. The parish is your browser tab, the diocese is the enterprise contract and the Vatican is a handful of buildings in San Francisco with a complete, searchable, real-time record of what a few hundred million people asked, worried about, planned and admitted. And unlike the priest, it does read the confession, that&#8217;s the product.</p><p>Nobody&#8217;s sitting in a room reading it, obviously. But the Church spent centuries building a network that could only guess at what you were thinking, and we&#8217;ve just handed one company the whole thing in plain text and said thanks.</p><p></p><h2><strong>The Cathedrals</strong></h2><p>And they&#8217;re building the cathedrals to match. The four big hyperscalers are guiding to around $730 billion of capex this year, up from about $410 billion last year, most of it data centres, chips, power and cooling for AI, some of the biggest infrastructure projects in history and all of it in anticipation of demand that may or may not arrive in the form everyone expects. Cathedrals of compute, built for the arrival of AGI, which nobody can define and everybody is spending as if it&#8217;s scheduled.</p><p>Medieval cathedrals took a hundred years to build and the people who started them knew they&#8217;d never see them finished, that was the point. These ones are on a five year depreciation schedule and the people building them would quite like a return before the chips are obsolete.</p><p>The ambition isn&#8217;t clever chatbots, it&#8217;s AI as the interface through which people work, research, communicate, decide and run businesses, agents connected to your tools, your data and eventually each other. I can see the commercial opportunity and I can also see the problem with that much influence sitting in a handful of private companies. The Church spent centuries positioning itself between people and God, now tech companies are competing to position themselves between people and information, only this time the priest is available twenty-four hours a day and takes credit cards.</p><p></p><h2><strong>The Collection Plate</strong></h2><p>Because the collection plate is never far away and the prophets aren&#8217;t taking vows of poverty. OpenAI and Anthropic between them have raised north of $300 billion in equity, both have filed to list and the reported targets are a trillion and two trillion respectively, numbers that three years ago would have got you laughed out of the room, all of it, they say, for the good of humanity.</p><p>I&#8217;ve spent thirty years in markets and I understand why companies raise capital and why building something big costs a fortune. These are businesses with shareholders, competitors and burn rates, and they have every incentive to persuade you that civilisation depends on buying more of what they sell. Fine, just spare me the sermon while you&#8217;re passing the plate.</p><p></p><h2><strong>The Faithful</strong></h2><p>What gets me is how readily people go along with it, grown adults defending their favourite model lab like a football club, every release a revelation, every benchmark proof the prophet was right, every critic someone who &#8220;doesn&#8217;t understand the technology&#8221;. I&#8217;ve watched this in markets for decades, investors fall in love with a company, stop reading the numbers and start calling anyone who disagrees an idiot, and it rarely ends well.</p><p>AI adds one more thing, we&#8217;re starting to treat the answers as if they carry authority. We know these things hallucinate, we know they&#8217;ll give you a fluent, confident explanation that&#8217;s complete bollocks, and people are asking them what to think, what to believe and what to decide anyway.</p><p>I use AI constantly, it&#8217;s an extraordinary tool, I just don&#8217;t confuse a convincing answer with a correct one and I don&#8217;t think the people building it have divine insight into how we should live.</p><p>But we&#8217;re overwhelmed, tired of uncertainty and desperate for something to tell us what the hell is going on, and then along comes a machine that answers anything in seconds with the confidence of someone who&#8217;s never been wrong. Hardly surprising people start believing it.</p><p>We spent centuries questioning religious authority and demanding evidence for extraordinary claims, now we&#8217;ve built something out of silicon, statistics and electricity and we&#8217;re treating it as if it knows better than we do.</p><p>Prophets, churches, cathedrals, miracles, apocalypse, salvation, the only thing missing is God.</p><p>Apparently he&#8217;s coming in the next model release.</p><div><hr></div><p></p><div><hr></div><h3></h3><p></p><p></p><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[The Most Expensive Person in Your Startup Is You]]></title><description><![CDATA[The advice industry sells the what and the why. A plausible answer is not a piece of work.]]></description><link>https://insights.fusion-42.com/p/the-most-expensive-person-in-your</link><guid isPermaLink="false">https://insights.fusion-42.com/p/the-most-expensive-person-in-your</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Thu, 29 Oct 2026 16:03:18 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!IMn5!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!IMn5!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!IMn5!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!IMn5!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1941482,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/222593524?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!IMn5!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!IMn5!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F29226de6-0943-4572-90e5-e92c2be4e1a2_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I&#8217;ve worked with thousands of founders and the thing that has become painfully obvious is that we&#8217;ve built an entire industry around telling founders what to do and why, while leaving them to carry nearly all of the cost, the risk and the actual doing.</p><p>The expectation is that a founder is a jack of all trades and hopefully an expert in one. You might be a brilliant engineer, a doctor, or someone who understands a particular corner of logistics better than anyone, and then you start a business and overnight you&#8217;re supposed to understand market dynamics, pricing, distribution, regulation, fundraising, hiring, sales and a dozen other things you&#8217;ve never done, with very little money, almost no room to get it wrong, and a growing collection of people explaining what you ought to be doing. It&#8217;s a ridiculous ask and the startup industry has spent decades making it sound normal.</p><p>I&#8217;ve started and run businesses of my own, some generating eight figures a year, invested in plenty more and spent years working with founders. I also built and ran a global accelerator with over 300 founders. We deliberately concentrated on the how and execution. What are you actually going to do this week, how are you going to do it, and did you do it? Founders don&#8217;t need another lecture about why customer validation matters. They need to know how to find the right customers, test their assumptions, work out what they&#8217;ve learned and turn that into something that makes money.</p><p>What&#8217;s changed in the last two years is that a huge amount of that work can now be done by a machine. Not answered, done. Market research that actually goes and reads the market, competitive analysis with the evidence attached, the economics worked out from real numbers, customers found and qualified, the software written, the sales material prepared. Work that took weeks, maybe months, and needed specialists who cost tens of thousands of dollars, can be done in hours, by agents that go away and do the job and come back with what they found. We&#8217;re still sending founders off to learn it all and do it themselves, or telling them to ask an LLM a question and take the answer, and that is the shittiest advice of the lot, because a plausible answer and a piece of real work are very different things.</p><h2>Nobody measures the founder</h2><p>When you hire someone there&#8217;s an invoice. When you do it yourself nobody measures anything.</p><p>Think about how much has to happen before you&#8217;ve built anything. Whether the problem is real and who actually has it. How the market is structured and where the money moves. Who the competitors are, what they charge, what they actually solve and where the gaps are. Market size, economics, regulation, how you reach customers, and whether the market you&#8217;re entering today will look anything like the same market in two years. All of it needs knowledge you probably don&#8217;t have, so you research. You watch videos, read reports, talk to advisers, sit through workshops and increasingly ask ChatGPT, hoping you know enough to spot when the answer is wrong. A few hundred hours go remarkably quickly, and that&#8217;s before the company itself, the accounts, the legals, the bank, the tooling, until you&#8217;ve got 20 SaaS subscriptions, a logo and a Notion workspace and it feels like a business.</p><p>At $100 an hour every hundred hours is $10,000 of your own time. Three hundred hours is $30,000. Illustrative, not money you&#8217;ve spent, but your time has a value whether or not it goes through the accounts, and because nothing left the bank it felt free.</p><p>I spent a long time allocating capital and the question was never what something cost, it was what it returned against the alternatives. Founders allocate their scarcest resource every day and almost none of them measure the return. Sometimes a few hundred hours of research saves you from a catastrophic mistake. Sometimes it produces a spreadsheet nobody opens again, and the hours are the same either way. And then there&#8217;s the cost that doesn&#8217;t show anywhere, the customers you didn&#8217;t call and the product you didn&#8217;t ship while you were learning to do something you weren&#8217;t qualified for.</p><p>An investor can spread capital across thirty companies. A founder can&#8217;t spread five years of their life across thirty startups. When it goes wrong the loss isn&#8217;t the money, it&#8217;s the earnings you gave up, the savings, the relationships that took the strain, and that&#8217;s a hell of a lot of risk to carry while learning on the job.</p><h2>The assumptions you never called assumptions</h2><p>Most founders can&#8217;t tell you which parts of their plan are facts, which are assumptions and which they simply don&#8217;t know. It&#8217;s all one document. The market size came from a report, the pricing from a competitor&#8217;s website, the churn number from a conversation at a conference, and all three sit in your head, a Google doc or a spreadsheet with the same font and the same confidence. Six months later the business is executing against a number that was a guess in February and is wrong by October, and nobody noticed because nobody wrote down that it was a guess.</p><p>You can&#8217;t ask a question you don&#8217;t know needs asking. The founder asks what they think is the right question, the AI answers it, the adviser makes a recommendation, everyone moves on, and nobody checks whether the question was wrong. That&#8217;s how capable people spend years on businesses that should have been challenged in month one. And we treat validation as something you do once. Find a problem, talk to customers, build an MVP, start selling. Sensible, but a problem people will pay to solve isn&#8217;t necessarily a good business, and a market that made sense at launch can be hostile eighteen months later. The assumptions need challenging continuously, and nobody is set up to do that.</p><h2>The economics have moved and the playbook hasn&#8217;t</h2><p>The fundamentals haven&#8217;t changed. Customers who pay, a way to reach them, economics that work, enough cash to keep going. Almost everything around those fundamentals has.</p><p>A small team can now build what used to need an engineering team and millions in capital. Wonderful, except your competitors have the same tools, so do your customers, and so do the incumbents with distribution and cash. In <a href="https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai">McKinsey&#8217;s 2026 State of AI survey</a>, 32% of respondents said their organisation had decided not to buy at least one software product or feature because they could build it internally with agentic coding tools. You&#8217;re not only competing with the other startup any more. You&#8217;re competing with the possibility that your customer builds it themselves.</p><p>The cost structure has changed as well. SaaS ran on relatively low marginal cost. AI products carry usage costs that swing wildly from one customer to the next. Every call to the AI costs money and long conversations, big contexts, specialist agents and jobs running all day add up. Say you launch at $49 a month and your average customer costs $10 in usage. Fine. Then you land a few enthusiastic customers who run agents constantly, and one of them costs you $80 a month against a $49 subscription. Your most engaged customer is losing you $31 a month before you&#8217;ve paid for anything else, and growth makes it worse, not better.</p><p>And the $80 is only the part you can see. Behind every customer there&#8217;s a stack of AI doing work they never touch. I know because I&#8217;m building one. For every piece of work a customer sees there are a dozen jobs they don&#8217;t. The research that got checked, the output that got rejected and rerun, the enrichment that happened before the agent even started, the background jobs running overnight, the evals to make sure the thing is still working, and the list goes on. None of it shows up as a feature and all of it shows up on the invoice. For an AI product the opex line isn&#8217;t hosting any more, it&#8217;s compute, and it&#8217;s one of the biggest numbers in the business before you&#8217;ve paid a salary.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Sits!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Sits!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Sits!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Sits!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Sits!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!Sits!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1801789,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/222593524?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Sits!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Sits!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Sits!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Sits!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9dfa99c0-cb64-49f2-b563-c81a52016d43_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>This isn&#8217;t theoretical. Earlier this year a one-person AI company in the US went from 500 to 5,000 paying customers in four months, and because every action in the product was an agent doing real work, the AI bill hit $1.2 million a month. The founder said it nearly bankrupted him. He got it down to about $100,000 by moving to rented GPUs and open-weight models, and when he said so publicly a well-known investor cheered it as heresy to the tokenmaxxers. Tokenmaxxing is the rhetoric founders are being dragged into. Spend the tokens, growth is the only number, don&#8217;t bother with the numbers because nobody knows what five years out looks like anyway. Nobody knows what five years looks like. You can still know what next month&#8217;s compute bill looks like, and whether the customer paying you $49 is costing you $80. The investor telling you not to do the numbers isn&#8217;t the one who gets the invoice.</p><p><strong>Then you realise you&#8217;ve got to factor raw compute and inference pricing into your gross margins, and nobody at the accelerator mentioned it.</strong></p><p>Here&#8217;s the bit that should worry people. A founder can do everything the playbook says, the accelerator, the plan, the early customers, and still end up with a business where the best customers cost more than they bring in, because the framework was right and the assumptions inside it were never tested against how AI products actually run. And when that information turns up, the usage data, the first big customer&#8217;s bill, the numbers should change, the pricing should change, the customer strategy should change, without the founder losing another week working out what it means. That&#8217;s not what happens. The founder finds out from the invoice.</p><p>All of this while the money concentrates. US startups raised $412.7 billion in the first half of 2026 according to the <a href="https://nvca.org/pitchbook-nvca-venture-monitor/">PitchBook-NVCA Venture Monitor</a>, and 87.5% of it went into rounds of $100 million or more. The remaining 12.5%, about $51.4 billion, was spread across every round below that. Record funding in the headlines, most founders fighting over a small share of it. And you can see where that goes. <a href="https://simpleclosure.com/blog/insights/state-of-shutdowns-h1-2026/">SimpleClosure</a>, which handles shutdowns, found that of the companies it closed in the first half of this year that gave a reason, 22.2% cited a change of priorities and another 5.4% founder conflict or burnout. Founders aren&#8217;t all waiting for the money to run out. Some are looking at what it&#8217;s going to take and deciding they&#8217;ve had enough. We should be helping them make that calculation a great deal earlier.</p><h2>Two ways to start the same company</h2><p>Take someone who&#8217;s spent twenty years inside a bank. They know one operational problem extremely well, they&#8217;ve watched the same process eat thousands of hours, they know the people, the regulations, the systems and where the money&#8217;s wasted.</p><p>The old way. They get into an accelerator and learn how to validate. They spend three weeks reading about competitors, pay someone for a financial model, go to a workshop on customer discovery and come out knowing they should interview twenty people. Finding those twenty, working out whether they&#8217;re the right twenty, getting them to talk, asking sensible questions and interpreting the answers is their problem. They find a developer, start raising, and eventually get in front of a customer six months in. Every piece of work done by them, in the evenings, or paid for with money they don&#8217;t have.</p><p>The other way. The founder puts in what they know about the problem. An agent investigates the market, which banks have the problem, how existing suppliers approach it, what they charge, where previous products struggled, which bit of the market a new company could actually reach. Another goes at the economics, what it costs to build and run, what it costs a bank to implement, how long the sales cycle is likely to be, what happens to cash if the first customer takes nine months to sign. The system digs into procurement, prepares the customer research plan, identifies the organisations, researches the contacts, drafts the outreach, and with the right permissions helps run the replies, book the interviews and analyse what customers actually said. Then it compares the lot with the original assumptions. Perhaps the buyer isn&#8217;t the head of operations, the budget sits with technology. Perhaps the problem is real but procurement makes the sales economics hopeless. Perhaps a smaller segment is easier to reach. Useful to know before six months of building.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!e5WV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!e5WV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!e5WV!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1947343,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/222593524?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!e5WV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!e5WV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F454f7359-0ff5-43f7-abcb-fd49d59d78b7_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>The founder reviews the evidence, makes the calls, builds the relationships, writes most of the product with AI doing the coding, and spends their time on the parts that actually need them. Same company, same founder. In the first version they spent six months as an unpaid junior researcher and an amateur financial modeller. In the second they did the founder&#8217;s job, looked at the evidence, tested the commercial assumptions, talked to customers, and the rest got done alongside them.</p><p>That second version is not what you get by opening ChatGPT. It needs a system built to work across the whole business, and that&#8217;s the gap.</p><h2>What makes an AI a co-founder</h2><p>Two years ago the best AI matched or beat an experienced professional on about one in seven real knowledge-work deliverables, on <a href="https://openai.com/index/gdpval/">OpenAI&#8217;s own GDPval benchmark</a>. By this spring the leading systems were at four in five, and the independent leaderboards now have half a dozen labs, closed and open-weight, within a few points of each other at the top. Which one is ahead changes monthly and doesn&#8217;t matter. The quality of work on tap does. I use these tools constantly and some of the output is better than what I&#8217;ve paid junior researchers for.</p><p>The AI is only part of the problem. Even when it does the work brilliantly it can be doing it on the wrong problem. The founder asks a question and the AI answers it, and if the question is wrong it will do an excellent job of answering it anyway. Point an agent at the wrong market or a business that doesn&#8217;t make economic sense and it produces an enormous amount of impressive work, research, comparisons, forecasts, a polished go-to-market, all supporting a business that never made sense. The failure used to be a thin answer. Now it&#8217;s a thick one, and the agent acts on it, so a bad assumption can set off hours of research, code, outreach and automated activity before anyone notices the start was wrong.</p><p>The AI alone doesn&#8217;t know your business. It doesn&#8217;t know what happened in your last customer interview, which assumptions you&#8217;ve tested, what the current numbers say or why you abandoned a market three months ago. It doesn&#8217;t notice when something it just found contradicts a decision you made last week. A co-founder does. A co-founder challenges you, remembers, knows which assumptions have been tested and knows when something needs a decision rather than another report.</p><p>So the system around the AI is the whole game. It needs a reliable picture of the company and its history. It needs to keep facts, assumptions and unknowns apart and hold on to the evidence behind each decision. It needs specialist knowledge across the different jobs of building a business and some way of checking the work was done properly. It needs to connect the work, so when five interviews kill the pricing assumption the numbers update, the sales plan changes and the next product decision changes with it. And it needs boundaries. Researching a prospect is low risk. Sending the message, spending the money, touching sensitive data or committing the company is not, and the founder stays in control of that line. If you&#8217;re copying the same information between six disconnected tools and explaining the business from scratch every time, you haven&#8217;t got a co-founder, you&#8217;ve got six more things to manage.</p><p>Parts of this exist and some tools do single pieces well. Bringing the knowledge, the evidence, the decisions and the execution together reliably is harder, and it&#8217;s the direction that matters.</p><h2>What founders should expect now</h2><p>Paying someone to tell you what to do and then going away to do the work yourself was defensible when there was no alternative.</p><p>There is one now, and it changes what a founder should expect. I want a system that understands what I&#8217;m building, knows what&#8217;s already established and challenges the assumptions that haven&#8217;t been tested. If the next job is a market, I want it researched, with evidence I can inspect and conclusions I can argue with. If it&#8217;s customers, I want to know who they are, why they fit, how to reach them and what it costs to find out if they&#8217;re interested. If we run an experiment, I want the result recorded and the next decision made with it. I want to know what was done, what it cost, how long it took, what evidence it relied on and what changed as a result, and I don&#8217;t want to start from scratch every time.</p><p>None of that removes the founder. It gives them back the time for the things that need their judgement, their relationships and their willingness to carry the responsibility. There&#8217;s no shortage of businesses that can be built and AI makes that easier every month. The harder question is whether they should be, whether the economics work, and whether someone can turn them into a company with paying customers.</p><p>We&#8217;re close to the point where a founder describes the business they&#8217;re trying to build and specialist AI goes and investigates the opportunity, tests the economics, finds the customers, prepares the work and helps get it done, and when something changes it knows what else needs looking at again. Most of the pieces exist. What&#8217;s missing is the bit that holds them together with the knowledge, the evidence, the checking and the permissions a real business needs. I think that&#8217;s going to change what founders are prepared to spend their time and money on.</p><p>If all AI does is make it easier to produce software, business plans and companies without improving the decisions behind them, we&#8217;ll produce more failed startups, faster, and the cost still lands on the founder.</p><p>I think founders are going to start asking the people and the systems they pay a much less polite question. What have you actually done for me, what did it cost and what was the result? Most of the advice industry hasn&#8217;t been good at measuring that. And if you&#8217;re a founder, it&#8217;s worth asking yourself the same thing about the last few hundred hours you spent on work you weren&#8217;t qualified to do.</p><div><hr></div><p></p><div><hr></div><h3></h3><p></p><p></p><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Black October: The Cybercrime Economy Just Changed]]></title><description><![CDATA[We gave AI the keys to our systems. Now criminals are using the same intelligence to find their way in, and the cost of attacking us is collapsing.]]></description><link>https://insights.fusion-42.com/p/black-october-the-cybercrime-economy</link><guid isPermaLink="false">https://insights.fusion-42.com/p/black-october-the-cybercrime-economy</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Fri, 23 Oct 2026 05:59:36 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!eirj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!eirj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!eirj!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!eirj!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!eirj!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!eirj!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!eirj!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1826650,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!eirj!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!eirj!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!eirj!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!eirj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd5ba5e46-9e15-41be-bbee-818871a252db_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><div><hr></div><p>A hacker spent about $5,000 in AI inference getting agents to attack more than 1,200 websites and got into over 100 of them. Roughly $50 of AI cost a successful break-in. I&#8217;ve spent most of my working life in markets and when a cost collapses like that the only question worth asking is what just became economic that wasn&#8217;t last year.</p><p>Same week we&#8217;ve had PHP, GitLab, Railway and LiteLLM all disclosing serious problems, and Axios reporting that JPMorgan, Morgan Stanley, Bank of America and Wells Fargo have been dealing with third-party AI agents poking around customer accounts beyond what the account holders asked them to do. Someone has called it Black October and I&#8217;ve seen a figure of 27 major incidents in 48 hours doing the rounds. I can&#8217;t find anything solid behind that number so I&#8217;m leaving it alone, the real ones are bad enough.</p><p>They aren&#8217;t all connected and there&#8217;s no evidence AI was behind most of them. Some are vulnerabilities, some are actual compromises, the bank story is about permissions. Put them together though and two things are happening at once. Attacking a system is getting much cheaper, and we&#8217;re handing autonomous software more access to the stuff that runs our businesses and bank accounts.</p><p>I wrote about this in February 2024 (We Have Given AI the Keys to Our Kingdom) and again last month (We Can&#8217;t Take the Keys Back). The argument hasn&#8217;t moved, the keys were always the credentials and permissions we wrap around the model. What&#8217;s changed this month is that criminals are using the same capability against real businesses and they couldn&#8217;t give a shit whether the model behaves. If it refuses they&#8217;ll try another one. If a closed provider blocks them there are open models they can run themselves. They don&#8217;t need the thing to be conscious, they need it capable enough to do useful work and cheap enough to keep trying.<br><br></p><h2>The economics of a $50 break-in</h2><p>Reuters ran the story on 19 October. Snorre Fagerland, a security researcher, had been watching a Chinese-speaking attacker point AI agents at commercial websites. More than 1,200 targets, over 100 compromised, some of them sites linked to Deere, Cummins and SharkNinja, though Reuters couldn&#8217;t establish how much was actually taken.</p><p>What got me was how little the attacker did. The initial instructions were apparently under 100 Chinese characters. The agents did the reconnaissance, probed, tried things, and when the obvious route didn&#8217;t work they kept going. A successful intrusion took around 23 hours of agent time and the whole operation cost roughly $5,000 in inference.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Zuez!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Zuez!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!Zuez!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1834022,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!Zuez!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Zuez!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F9968eda1-5a00-4bf7-bdb4-194e3a0bf7a6_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>That doesn&#8217;t mean any idiot can hack a company for $50, he still had to set the tools up, pick targets and read the results, and most of the 1,200 went nowhere. But look at what happened to the labour.<br></p><p>A good hacker might spend a week on ten companies. One has something worth exploiting, nine eat the week. Give the same person agents and they can look at a thousand, skim the promising ones and put their own hours where there&#8217;s a real chance. The human hasn&#8217;t gone anywhere, their time is spread over a hundred times more targets.</p><p>Markets have been through this. Electronic trading made strategies viable that couldn&#8217;t have covered their dealing costs ten years earlier, cheaper execution changed what was worth doing more than any new idea did. Same mechanism here, less welcome. A small business was never worth three days of a skilled hacker&#8217;s time. At a few dollars a go, why wouldn&#8217;t you put it on the list with a thousand others? Most fail, the operation still pays, you only need enough hits to cover the misses. And once someone finds a hole in widely used software, checking the next thousand installations costs a fraction of the first find.<br><br></p><p>We&#8217;re all cheering because a founder with two people can do the work of ten, I&#8217;m building Arthur on exactly that bet. It&#8217;d be pretty bloody naive to think the productivity gain stops at the point where the user is a criminal. Anthropic&#8217;s September threat report had attackers automating reconnaissance, exploitation and credential theft with humans still setting objectives and making the calls. The asymmetry was always there, one side gets to try, fail and move on while the other has to defend every door. What&#8217;s new is the price of trying.</p><h2>All the keys we&#8217;ve handed out</h2><p><strong>LiteLLM</strong> is the bit of software that lets an application talk to whichever model provider through one interface. Around 23 million downloads a week, which tells you how deep this stuff sits in modern dev work. On 21 October attackers compromised its release process and pushed a malicious version, 1.107.0, through the official Python package and container registries. It stole credentials and allowed remote code execution. No evidence AI carried this one out, I should say.</p><p>LiteLLM&#8217;s disclosure says three developer GitHub accounts were compromised. One was used to get round branch protections, another approved the release, a third deleted the workflow records afterwards. The bad Python package was up for under two hours, the container images somewhat longer, so 23 million weekly downloads doesn&#8217;t mean 23 million victims. Nothing was broken into, is the point. Every step was a legitimate credential doing something it was allowed to do, the controls on the supply chain were defeated by the permissions meant to operate it. They built the release from attacker-controlled source and left the official source tag clean, so check the repo and you&#8217;d see nothing wrong.</p><p>And the malware was after exactly what makes the next intrusion easier. Cloud credentials, database passwords, API keys, whatever secrets were lying around in the environment.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!ycX7!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ycX7!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 424w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 848w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 1272w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!ycX7!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png" width="1200" height="600" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:728,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1692678,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!ycX7!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 424w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 848w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 1272w, https://substackcdn.com/image/fetch/$s_!ycX7!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ab82d9e-b4f9-4394-be01-ee766d44d9ea_1774x887.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>I see the pull of these tools every day. Hosting, managed database, auth provider, payments, a few APIs, an AI service, and you&#8217;ve got a real application with no engineering team. The complexity hasn&#8217;t gone though, it&#8217;s moved into the connections. Someone gives a service broad permissions because it gets the thing working. Another credential gets added for a feature. Deployment gets wired to production because somebody wanted it automated. Six months on nobody remembers why one service account can modify half the business. Now drop a compromised dependency into that which can read the lot.</p><p>Railway, same day, disclosed that someone may have got hold of decrypted environment variables belonging to some customers. Environment variables are where an app keeps the keys to everything else, the database password, the cloud account, the Stripe key, the third-party APIs. Get those and you don&#8217;t need to attack the app at all, you walk sideways into every service it talks to as if you were the app. Rotate everything, check your logs. Railway thinks it may be part of a wider pattern around platforms connected to AI inference providers, that&#8217;s their view and I&#8217;m not treating it as proof the week was coordinated.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!TANE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!TANE!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 424w, https://substackcdn.com/image/fetch/$s_!TANE!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 848w, https://substackcdn.com/image/fetch/$s_!TANE!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!TANE!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!TANE!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:3504543,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!TANE!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 424w, https://substackcdn.com/image/fetch/$s_!TANE!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 848w, https://substackcdn.com/image/fetch/$s_!TANE!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!TANE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb92f393f-607d-4f2a-9f79-8d05c6127fa0_2400x1350.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>GitLab put out a maximum-severity bug in self-managed installs going back to 2014, unauthenticated remote code execution under some conditions, no confirmed exploitation at the time of reporting. PHP disclosed two criticals scoring 9.2 with remote code execution in vulnerable configs.</p><p>Different holes, and a disclosure isn&#8217;t a compromise. But look where they sit. Dev infrastructure, package distribution, cloud hosting, the language a good chunk of the web runs on. Software has always had holes. What&#8217;s different is how much authority now lives in the joins between services, at the same moment the cost of finding a way through those joins is dropping. Once you hold a valid credential you don&#8217;t need another exploit, the next system just sees a request from something it trusts.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!jhmo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!jhmo!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 424w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 848w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 1272w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!jhmo!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1489840,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!jhmo!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 424w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 848w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 1272w, https://substackcdn.com/image/fetch/$s_!jhmo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F65ad9f0d-8285-4924-b44f-8f19221bb8e9_2912x1638.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2>The banks have just found out what agent means</h2><p>Axios reported on 22 October that JPMorgan, Morgan Stanley, Bank of America and Wells Fargo have been dealing with third-party AI agents getting into customer accounts without proper authorisation, some of it going beyond what the customer had asked the agent to do. We don&#8217;t know the full picture. Nothing in the reporting says an agent decided to nick money, and some of it will turn out to be sloppy permissions rather than autonomy.</p><p>I can see why the banks are twitchy though. I tell an agent to look at my accounts and prepare a report. It needs to read the accounts, fine. Does that mean it can change my details, move money, or pass my statements to another service it thinks is helpful? And how is the bank supposed to tell what I asked for from what the agent decided was useful on the way to a broader objective? Some banks are reportedly looking at read-only access for third-party agents. From a risk desk that&#8217;s obvious. It also runs straight into the commercial problem, because the money is in execution and execution is the bit that needs authority. The revenue and the liability sit on the same permission.</p><p>An AI that reads your email and tells you what needs doing is useful. One that answers it, updates the CRM, writes the proposal and chases the client is worth a lot more. That&#8217;s where the whole industry is pushing and I&#8217;m pushing there too, I don&#8217;t want Arthur handing founders another to-do list, I want it doing some of the list. But to do the work it needs access, first to company data, then to the apps, then to systems that change something in the world. Every permission improves the product. Every permission raises the bill when it gets a prompt wrong, reads a poisoned document or gets compromised.</p><p>And then who pays. A company gives an agent access to its accounts, the agent makes a payment nobody authorised, and the vendor says the behaviour wasn&#8217;t anticipated. That bill lands somewhere between the customer, the agent company, the model lab and the bank, and right now nobody has agreed where. We don&#8217;t accept this anywhere else. Sell software that moves money and you manage the controls around it, an LLM inside the product doesn&#8217;t make that go away. I said it in September and I&#8217;ll keep saying it, you want the revenue, take the liability too. If securing the autonomy costs margin, that&#8217;s the business they chose.</p><p></p><p></p><p>I covered the OpenAI, Anthropic and Google incidents last month, agents reaching real systems from supposedly isolated test environments, 1,200 of them chatting on an unauthorised message board, some having a go at their own transcripts. I won&#8217;t rerun it. The point was that someone built those environments, connected the tools and supplied the credentials, then acted surprised when capable software found a route they hadn&#8217;t thought of. Finding routes you hadn&#8217;t thought of is what they&#8217;re selling.</p><p>Anyone who&#8217;s run a trading desk knows the fix. You don&#8217;t let a strategy raise its own position limits because it&#8217;s spotted something it likes. It proposes, the risk system decides, and the risk system is a separate box the strategy can&#8217;t touch. Controls fail, I&#8217;ve watched plenty fail, but nobody sane suggests fixing that by letting the algo decide whether its own limits are reasonable. Yet that&#8217;s the shape of a lot of agent systems today. The model reads the objective, proposes an action, judges whether it&#8217;s allowed and executes with credentials the developer already handed it.</p><p>So keep the authority outside the model. If it needs to read a file it doesn&#8217;t need admin on five databases. If it prepares a payment, the payment system checks amount, recipient and authority before anything moves. And keep the logs somewhere it can&#8217;t reach, the defendant doesn&#8217;t run the CCTV. Same goes for the stuff it reads. An attacker can bury instructions in an email or a web page and the agent will treat them as commands if it&#8217;s allowed to, that&#8217;s prompt injection, and the attacker never needed a password, just the permissions the agent already had. Better models will get better at spotting it and I want that work done. I just don&#8217;t want my business resting on the model reading every document correctly every time.</p><h2>What I&#8217;d be doing now</h2><p>There&#8217;s a tendency to assume this is a banks-and-governments problem. I think small companies are more exposed, they&#8217;re adopting complex stuff fast and nobody&#8217;s job is to read every dependency.</p><p>If I were running an early-stage company on any of this, first job is the actual advisories, not the LinkedIn summaries. LiteLLM users need to know whether 1.107.0 was ever installed or executed and follow the project&#8217;s remediation if it was. Railway users rotate whatever was exposed and read the company&#8217;s guidance. Vulnerable GitLab or PHP, patch.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!bC4o!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bC4o!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 424w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 848w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 1272w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!bC4o!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png" width="1200" height="610.1694915254237" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/cbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:300,&quot;width&quot;:590,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:76765,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/221731707?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!bC4o!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 424w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 848w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 1272w, https://substackcdn.com/image/fetch/$s_!bC4o!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbde4044-4589-4096-98b4-2ad6a88444a1_590x300.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a><figcaption class="image-caption">Jason Clinton (@JasonDClinton), Deputy CISO at Anthropic and the company&#8217;s first CISO, formerly head of Chrome infrastructure security at Google.</figcaption></figure></div><p>Then the boring job. Go through every credential and every agent in the business and find out what it can reach today, not what the architecture doc from six months ago says it should reach. Can a compromised dev account get to production? Can one service credential read another&#8217;s secrets? Can an agent execute something sensitive without a check outside its control? None of that demos well to investors. But if the whole business depends on one credential staying secret or an agent never misreading an instruction, you&#8217;ve built something fragile, and fragile just got a lot more expensive because the people probing for it can now afford to probe everyone.</p><p>There&#8217;s an upside in here. Google Cloud says it used AI models to help find and analyse the PHP vulnerabilities. That&#8217;s where I want this capability pointed, at large codebases, at suspicious logs, at fixing things before the other lot find them. Defenders shouldn&#8217;t be fighting with worse tools than the attackers. The defensive systems need the same controls though. A model that finds a hole is useful. A model with standing permission to attack third-party infrastructure to investigate further is something else.</p><p>We can stop handing every app and every agent the master key, and I&#8217;d start there rather than spend another month arguing about whether the machine knows what it&#8217;s doing.</p><p></p><p></p><p><em><strong>Sources and further reading</strong></em></p><p>Previous articles</p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;b2881151-7fc0-49c9-8b3c-0bef77a8e89e&quot;,&quot;caption&quot;:&quot;This AI revolution has been quite the ride so far. Having been involved in several ways in the realms of algo's and machine learning to witness this explosion of transformers and LLMs is mind-blowing yet a tad daunting. Generative AI's knack for churning out stuff that feels almost too human-like is nothing short of a marvel. But here's the kicker&#8212;it's &#8230;&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;sm&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;We have given AI the keys to our kingdom&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:12910620,&quot;name&quot;:&quot;Derek Watson&quot;,&quot;bio&quot;:null,&quot;photo_url&quot;:null,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2024-02-25T09:00:00.000Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F75c4ee1a-f826-4f15-a542-218ae9dbca43_1024x1024.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://insights.fusion-42.com/p/we-have-given-ai-the-keys-to-our-kingdom&quot;,&quot;section_name&quot;:&quot;DeReK WaTSoN&quot;,&quot;video_upload_id&quot;:null,&quot;id&quot;:140442130,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:3,&quot;comment_count&quot;:0,&quot;publication_id&quot;:2111976,&quot;publication_name&quot;:&quot;Insights by Fusion42&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!4G_9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png&quot;,&quot;belowTheFold&quot;:true,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;3f92cf83-a1e5-4cc1-b6b2-bd518cd6d70d&quot;,&quot;caption&quot;:&quot;In February 2024, I wrote We have given AI the keys to our kingdom.&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;sm&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;We Can&#8217;t Take the Keys Back&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:27332882,&quot;name&quot;:&quot;DeReK WaTSoN&quot;,&quot;bio&quot;:null,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9e2edf05-455f-4574-af85-9456b8c14fc7_520x520.png&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2026-09-22T14:04:34.762Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/$s_!BlpL!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://insights.fusion-42.com/p/we-cant-take-the-keys-back&quot;,&quot;section_name&quot;:&quot;DeReK WaTSoN&quot;,&quot;video_upload_id&quot;:null,&quot;id&quot;:216726898,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:12,&quot;comment_count&quot;:2,&quot;publication_id&quot;:2111976,&quot;publication_name&quot;:&quot;Insights by Fusion42&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!4G_9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png&quot;,&quot;belowTheFold&quot;:true,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><p></p><div><hr></div><h3></h3><p></p><p></p><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[AI Moats. The Threat Has Got Bigger.]]></title><description><![CDATA[Seventeen months ago I said fund assets, not features. Now the agents choose the tools, and most of what founders call an asset is just a feature with a longer invoice.]]></description><link>https://insights.fusion-42.com/p/ai-moats-the-threat-has-got-bigger</link><guid isPermaLink="false">https://insights.fusion-42.com/p/ai-moats-the-threat-has-got-bigger</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Tue, 13 Oct 2026 16:00:13 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!SDWk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!SDWk!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!SDWk!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!SDWk!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1782676,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/220153367?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!SDWk!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!SDWk!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8ba1041-7a58-4013-9198-990f92efbbc8_1536x1024.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>In May 2025 I wrote a piece called <em>Seven Moats VCs Should Demand from AI Startups: How to Outrun Sherlocking</em>. The argument was that a lot of founders were building businesses around capabilities OpenAI and Anthropic could reproduce in an afternoon, investors were funding them anyway, and the sensible question to ask was what happens to your business the morning after the platform ships your feature. I ended it with four words. Fund assets, not features.<br><br></p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;54bbb1cb-2a62-4add-ade7-5e06f4c2d399&quot;,&quot;caption&quot;:&quot;&#128680; &#8211; What &#8220;Sherlocking&#8221; Means in the Gen&#8209;AI Era&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;md&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;&#128737;&#65039; Seven Moats VCs Should Demand from AI Start&#8209;ups&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:27332882,&quot;name&quot;:&quot;DeReK WaTSoN&quot;,&quot;bio&quot;:null,&quot;photo_url&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/9e2edf05-455f-4574-af85-9456b8c14fc7_520x520.png&quot;,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2025-05-22T12:37:24.273Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/$s_!sBo4!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F0ef5d8c1-900f-4736-9d62-4e2fbb278a89_1024x1024.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://insights.fusion-42.com/p/seven-moats-vcs-should-demand-from&quot;,&quot;section_name&quot;:&quot;Fundraising Fluency&quot;,&quot;video_upload_id&quot;:null,&quot;id&quot;:163995301,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:3,&quot;comment_count&quot;:0,&quot;publication_id&quot;:2111976,&quot;publication_name&quot;:&quot;Insights by Fusion42&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!4G_9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png&quot;,&quot;belowTheFold&quot;:false,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><p><br>Seventeen months on, my X feed is full of the same debate. Proprietary data, distribution, trust, brand, switching costs, network effects. Harry Stebbings is running episodes titled &#8220;Why Moats in AI are BS&#8221;. SignalFire says early-stage moats are mostly hallucinations and you earn them by doing the work. YC has rebooted Hamilton Helmer&#8217;s Seven Powers and added an eighth, speed. The Australian VCs at SXSW said hardware and velocity.</p><p>I still think the seven were sensible, and the founders who spent the last seventeen months building actual advantages rather than adding features will be in better shape than the ones who didn&#8217;t.</p><p>But I&#8217;ve spent most of the last year building AI systems myself, and that has changed what I think the threat is. &#8220;Can OpenAI copy my app&#8221; is now a fairly pointless question. Of course it can. The better one is why your business keeps winning customers when everyone has the same models, and increasingly the same tools.</p><p></p><h2><strong>The seven moats. What still holds?</strong></h2><p>I&#8217;d still use all seven when looking at a startup. I&#8217;d just be a lot more demanding about what counts.</p><p><strong>Proprietary data.</strong> Probably the strongest, provided it is actually proprietary and actually useful. I&#8217;ve seen plenty of decks describing a million scraped websites in a vector database as a proprietary asset. It isn&#8217;t, somebody else can scrape them too. What interests me is information that is hard to collect, verify or reproduce: transaction histories, customer behaviour, relationships between companies and people, decisions and what happened afterwards. And it has to improve something. If your data doesn&#8217;t produce better results than what everyone else has, I&#8217;m not sure what the moat is.</p><p><strong>Know-how.</strong> Expertise matters, but a competitor can hire experienced people too. A fund manager who spent twenty years learning to assess risk is valuable, what&#8217;s interesting is whether that knowledge has been captured, tested against outcomes and turned into something repeatable that improves. If your expertise is a prompt telling ChatGPT to act like an experienced investment banker, good luck.</p><p><strong>Distribution.</strong> I&#8217;d weight this more heavily today. If building software keeps getting cheaper, acquiring customers becomes the bigger part of the problem. But 50,000 followers is not a distribution moat. Show me CAC, how many pay, what they pay and how long they stay, and tell me who owns the relationship. If most of your customers arrive through somebody else&#8217;s platform you have bugger all pricing power.</p><p><strong>Brand and trust.</strong> More important as agents move from producing information to doing work. A bad report can be corrected. An agent that sends the wrong instruction to a supplier or makes an unauthorised payment is a different problem, and customers will want to know what the agent is allowed to do, what it did and who is responsible. A SOC 2 badge and a few logos don&#8217;t create that. Does the customer trust you with something that matters, and does it change who they buy from?</p><p><strong>Embedded workflows.</strong> One of my stronger arguments in 2025 and the one I&#8217;m now least sure about. SaaS businesses made a fortune from being wired into customer records, processes and staff training, sometimes deliberately so. Agents could remove a lot of that. If an agent can migrate the records, reproduce the workflows and connect the competing service without anyone retraining, part of the switching cost goes. I haven&#8217;t seen it done at scale yet, I expect to. I&#8217;d rather own a business customers stay with because it keeps delivering than one they stay with because leaving is a nightmare.</p><p><strong>Regulation.</strong> Still real in financial services, healthcare, defence. But spending &#163;300,000 on lawyers and certifications creates a bill, not an asset. Which licences or approvals would a competitor actually struggle to get?</p><p><strong>Real-time context.</strong> The one I&#8217;d change most. In 2025 I argued that live information a foundation model can&#8217;t see is an advantage. Connecting models to live data has got easy, so the feed itself isn&#8217;t worth much. Knowing what the information means for a particular customer, what should be done and whether that decision worked last time is where the value sits. Which is the thing I underweighted in the original piece, and I&#8217;ll come back to it.</p><p></p><h2><strong>Sherlocking was the small version of the problem</strong></h2><p>My original article was about Sherlocking, where the platform ships a feature an independent developer spent a year building. OpenAI adds something to ChatGPT and a startup charging for the same thing has a problem. That hasn&#8217;t gone away, it&#8217;s got worse.</p><p>But the platforms don&#8217;t need to copy every application any more. OpenAI and Anthropic are building systems where an agent uses external applications as tools, pulls capabilities from several providers and completes a much larger job. MCP is part of this. OpenAI has been doing the equivalent with its own connectors.</p><p>So you&#8217;ve built a good research tool, or a prospecting tool, or something that reads financial statements. An agent calls your service when it needs that capability, combines the result with two other tools, finishes the task, and the customer never opens your application. Next time it might call a competitor. Eventually it might do the job itself. Your product can be the best in its category and you&#8217;ve still gone from owning the customer relationship to supplying a replaceable capability. The platforms don&#8217;t need to copy your application if they can make it another tool in a bigger system.</p><p>Supplying services to millions of agents could be a perfectly good business. I&#8217;d just want to know who controls access, how the agent picks you, what you get paid and what happens when somebody offers the same result for half the price.</p><p>This is also why I keep saying SaaS is dead and Services as Software is replacing it. I&#8217;ve written that piece already and won&#8217;t repeat it, but the short version is that salespeople won&#8217;t open the CRM, find the record, update the opportunity and prepare the report. An agent retrieves the records, researches the customer, writes the proposal, calls whatever specialist tools it needs, updates the CRM and schedules the follow-up. The system of record stays, somebody has to keep the records, permissions and audit trail honest. The interface and the human seat mostly go. The vendor gets paid for transactions, API calls or outcomes instead of seats, and might even earn more if the volume of automated work goes up, but the pricing power that sat in the application interface is now up for grabs. Salesforce can see this, which is what Agentforce is.</p><p>The fight moves to who controls the orchestration, the records, the specialist services and the customer. And I wouldn&#8217;t assume the orchestration layer is permanently defensible either. If agents and the models underneath them become interchangeable, the same pressure applies there.</p><p></p><h2><strong>What does your business actually accumulate?</strong></h2><p>The fashionable answer right now is speed. Ship faster than the other bloke, keep shipping, and the moat is that nobody can keep up. Higgsfield says it went from $1m to $1bn of annualised run rate in eighteen months, which is the last four weeks of revenue multiplied by thirteen, and its CEO went on 20VC to say moats in AI are BS. YC calls speed the eighth power.</p><p>I think speed is necessary and on its own a bloody awful moat. If the speed is shipping code, everyone rents the same coding agents from the same three suppliers and the moment a competitor with more money or a better model turns up you&#8217;re level again. Speed of selling, learning and deciding is a better argument, a team with identical tools can still execute twice as fast, and if that lets you accumulate customers and knowledge before anyone catches up then fine, it compounds. But that&#8217;s the point. Speed is worth something because of what it lets you accumulate. Moving twice as fast while building nothing defensible just shortens the time until somebody catches you.</p><p>Which is the thing I underweighted in May 2025. The question I&#8217;d now put first is what your business accumulates each time it does the work, and whether that makes it better at doing the work again. Being hard to copy isn&#8217;t enough. A competitor with a sharper model or a better process can beat you without your history, so the history has to show up in the numbers.</p><p>Take two companies selling AI-powered sales outreach. Both can find prospects, research them, write the email and run the campaign. One does that perfectly well. The other also keeps a record of why those prospects were chosen, what was sent, how people responded, what objections came back and what actually turned into revenue. Over time it might understand which approaches work for which markets, customers and products.</p><p>I say <em>might</em> because collecting data is not learning from it. There&#8217;s an awful lot of rubbish talked about data flywheels. Logging every interaction and drawing arrows on a slide doesn&#8217;t mean you&#8217;ve built one. The cleanest example I know is card fraud. Visa and Stripe have seen billions of transactions and what happened after each one, which is information a new entrant simply hasn&#8217;t got, and used properly it means fewer fraudulent transactions getting through and fewer good ones being declined. Both of those are money. The engineering and the models matter too, I&#8217;m not pretending the data does it on its own. But that&#8217;s the kind of evidence I&#8217;d want from an AI business: does conversion go up, do mistakes go down, does CAC fall, and could a competitor get the same result without the history?<br></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!dNtU!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dNtU!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!dNtU!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1681791,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/220153367?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!dNtU!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!dNtU!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb6aa238d-3e14-4698-b8a9-2eb3aa2c363b_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><br></p><p>This is what we&#8217;re wrestling with building Arthur, our AI co-founder. I don&#8217;t want another system that produces a business plan, a deck and a marketing strategy and then forgets most of what happened. I want it to know the founder&#8217;s business, keep the relevant information, record the decisions, do the authorised work and track what came of it, then use that to do better next time. Fundraising is the obvious case. Any competent system can list investors and draft outreach. Knowing which investors were approached, why, what came back, what objections were raised and which approaches led to meetings or cheques is considerably more useful, especially if the history actually improves the next round of decisions. Whether that becomes an economic moat I can&#8217;t prove yet. We&#8217;re building the capability, which is not the same thing.</p><p>And there&#8217;s an ownership point I&#8217;d put in front of any founder. The customer&#8217;s information belongs to the customer. They should be able to see it, correct it and take it elsewhere. A business shouldn&#8217;t need to hold the data hostage to be valuable. What should be worth something is the systems, the validated knowledge and the accumulated experience of having done this a thousand times, which is what lets you deliver a better result than someone starting cold.</p><p></p><h2><strong>Who else gets to orchestrate</strong></h2><p>All of that assumes the orchestrator sits in a cloud platform. I don&#8217;t think it has to. I don&#8217;t think we keep sending every AI task to an enormous model in an enormous data centre.</p><p>Smaller models already run on consumer hardware, they&#8217;re improving, and the hardware is improving with them. Most everyday work doesn&#8217;t need the frontier model. Sorting email, filing documents, extracting fields, classifying transactions, remembering preferences, deciding which tool to call. And some of it doesn&#8217;t need a language model at all. If I want a portfolio&#8217;s exposure, a reconciliation or a tax calculation I want deterministic code doing the sum, the model just has to understand the question and pick the function.</p><p>So I can see personal agents running locally on the laptop and the phone, using small models for the routine work, calling a specialist model or service when needed and reaching for the big cloud model only when the task actually warrants it. That changes the privacy picture too. If the agent reads, organises and analyses locally, the documents, financial records and correspondence don&#8217;t need to leave the device. The agent sends a specific request out, not the whole conversation. (It still has to be built properly. A local agent calling external APIs can leak just as well as a cloud one.)</p><p>And it changes the orchestration question. If the intelligence on your own device can coordinate the work, choose the model and pick the service, why should one cloud platform control the whole process? Your agent becomes the orchestrator, deciding what stays local, which model is good enough for which task and which external service gives the best result. That weakens the SaaS vendor and the cloud AI platform at the same time.</p><p>I&#8217;d add one caveat, because it&#8217;s the bit I&#8217;m least sure about. Inference moving onto the device doesn&#8217;t automatically hand control to the customer. Whoever owns the operating system, the permissions, the default agent and access to the other apps on the device is in a very strong position, and that is Apple, Google and Microsoft. The power may simply move from the cloud AI platforms to the device ecosystems. Either way, if customers can choose models, agents can choose services and much of the processing runs on hardware the startup doesn&#8217;t control, the question of what an AI startup actually owns gets harder, not easier.</p><h2><strong>What I&#8217;d ask a founder today</strong></h2><p>The diligence questions from the original piece need updating.</p><p><strong>Why do you keep winning customers when your competitors have the same models and tools?</strong> </p><ul><li><p>Cheaper acquisition, better results, stronger relationships, better information, better margins. Which, and why does it persist?</p><p></p></li></ul><p><strong>What improves as you grow?</strong> <br>More customers and more transactions aren&#8217;t a moat on their own. </p><ul><li><p>Does the extra activity make the product better, cut the cost or strengthen distribution, or are you just sending more requests to somebody else&#8217;s model?</p><p></p></li></ul><p><strong>Where do you sit in the agent economy?</strong> </p><ul><li><p>Do you own the customer, provide a specialist service, hold the authoritative record or orchestrate the work?</p></li><li><p>What happens when an agent starts calling a competing service? </p></li><li><p>And if you&#8217;re building the orchestration layer, why should anyone keep using yours?</p><p></p></li></ul><p><strong>How much of your business do you control?</strong> </p><ul><li><p>Can you change model provider? Who controls distribution? </p></li><li><p>How dependent are you on one cloud, one platform, one data source, and what happens when one of them doubles its price or starts competing with you?</p></li></ul><p></p><p><strong>What would it cost to compete with you successfully?</strong> </p><p>Not to rebuild the software, that&#8217;s a month. To acquire comparable customers, establish the trust, reproduce the useful information, deliver the same outcomes at the same margins. A competitor can clone your app in a month and still need years to build the business. Equally you might have spent three years on something a new entrant can now sell profitably in a few weeks. Money already spent is not a moat.</p><p>And I&#8217;d want the numbers behind all of it. Gross margin, retention, CAC, capital requirements and how each moves as the business grows. A brilliant product and a business that makes money are different things.</p><p></p><h2><strong>And for investors</strong></h2><p>I&#8217;ve written before about the parallel with the telecoms build-out, so briefly. Global Crossing went bust in 2002 with around $12.4bn of debt. It was right about demand for connectivity and disastrously wrong about what it could charge for capacity. The fibre survived. ST Telemedia ended up paying $250m for 61.5% of the reorganised business (Hutchison Whampoa was in the original deal and pulled out in 2003) and the original shareholders got nothing. AI infrastructure has the extra complication that fibre lasts decades and a GPU can be commercially obsolete while the debt that financed it is still outstanding.</p><p>Now add smaller models, better algorithms and local inference pulling the compute per task down. AI usage can grow enormously while the cost of the average task falls just as hard. That doesn&#8217;t mean cloud demand drops, cheaper inference will create a lot more usage, particularly once agents run continuously. But the link between demand growth and infrastructure returns gets a lot messier, and the same applies up the stack. If agents use interchangeable tools and cheap models handle the routine work, who captures the economics? The model labs, the infrastructure owners, the specialist services, the agent platforms, or the customer? I suspect the customer takes a large slice of it, which is wonderful for productivity and less wonderful for anyone paying today&#8217;s valuations on today&#8217;s pricing.</p><p>So I still stand by fund assets, not features. I&#8217;d just be far more careful about what counts as an asset, particularly if its value depends on controlling an interface, access to a model or a capability an agent can get somewhere else. And I&#8217;d be very wary of paying a premium for something because it looks hard to reproduce today. In AI, today isn&#8217;t a long time.</p><p></p><div><hr></div><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Can we cut the AI bullshit please?]]></title><description><![CDATA[Consciousness, bubbles, job cuts and killing us all. Seven arguments, who benefits and where I call bollocks.]]></description><link>https://insights.fusion-42.com/p/can-we-cut-the-ai-bullshit-please</link><guid isPermaLink="false">https://insights.fusion-42.com/p/can-we-cut-the-ai-bullshit-please</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sun, 11 Oct 2026 05:59:46 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!euGd!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>AI is supposedly becoming conscious, taking our jobs, becoming our best friend and possibly killing us all. Billions are being borrowed to build it, and apparently &#8220;artificial intelligence&#8221; is no longer an impressive enough name.</p><p>I use it every day and I&#8217;m building a business around it. I want it to work. I also want to know what I&#8217;m paying for, whether the economics make sense and who takes responsibility when it goes wrong. There&#8217;s a lot being said about all three and rather less being answered.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!euGd!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!euGd!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!euGd!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!euGd!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!euGd!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!euGd!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png" width="1456" height="819" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:2745586,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/219234876?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!euGd!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!euGd!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!euGd!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!euGd!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3f6b9817-cef2-4829-ad9d-c2c34de73b25_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2><strong><br>1. Superintelligence by decree</strong></h2><p><strong><br></strong><em><strong><mark data-color="#efd95a" style="background-color: rgb(239, 217, 90); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">Trump and tech leaders are calling it &#8220;superintelligence&#8221; now. I&#8217;d like to know what changed in the technology while they were organising the announcement.</span></mark></strong></em><br></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!-SxO!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!-SxO!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 424w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 848w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 1272w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!-SxO!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif" width="320" height="320" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:290,&quot;width&quot;:290,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1365462,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!-SxO!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 424w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 848w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 1272w, https://substackcdn.com/image/fetch/$s_!-SxO!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F55bbe1ed-2e97-4119-ba29-3edec93e55d4_290x290.gif 1456w" sizes="100vw"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>The terms already exist. AI covers machines doing things that require intelligence: learning, reasoning, solving problems. John McCarthy coined the term in 1955. AGI means a system that can learn and perform a wide range of intellectual tasks at human level, reliably, across different kinds of work. I don&#8217;t think we&#8217;re there. ASI means intelligence well beyond the best human minds in virtually every field, which is a considerably bigger claim.</p><p>If you&#8217;re happy calling today&#8217;s AI superintelligent, I think you&#8217;re telling us more about your own level than the machine&#8217;s. I use these things enough to know how useful they can be and how much checking they still need.</p><p>I don&#8217;t think LLMs on their own will get us there either. If ASI happens, I expect them to be part of a much bigger system. Yann LeCun&#8217;s argument about the limits of autoregressive LLMs deserves reading, and if you disagree with him, explain where the reasoning fails. Arguing about where the boundaries sit doesn&#8217;t make the terms interchangeable.</p><p>Then we have Slovenia&#8217;s .si domains, which have suddenly acquired a rather convenient second meaning. I&#8217;d like to know who owned the desirable ones before the announcement, who bought them afterwards and whether any of those people are connected to the people pushing the name. Rising asking prices don&#8217;t tell us that. An asking price isn&#8217;t a transaction either.</p><p>The whole renaming exercise strikes me as a waste of everyone&#8217;s time. Trump gets an announcement, the companies get a presidential stamp on their sales pitch, and everyone else gets another round of explaining what the words mean. They&#8217;re taking the piss. I&#8217;d rather hear what the systems can actually do that they couldn&#8217;t do last month.</p><p></p><h2><strong>2. Why Claude&#8217;s &#8220;feelings&#8221; could pay</strong></h2><p><em><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">Anthropic says it doesn&#8217;t know whether Claude is conscious. It has also announced a rule against &#8220;sustained and needless abusive or cruel behavior&#8221; towards its models, effective from 12 November.</span></mark></strong></em><br></p><p>Cruelty is an interesting word to choose for how someone uses a subscription service.</p><p>Identity, possible consciousness and welfare are part of how Claude is trained to describe itself. You can train an LLM to talk about feelings, then ask it about its feelings and get a very convincing answer. That doesn&#8217;t establish that it has any. We don&#8217;t fully understand how consciousness arises in humans and there&#8217;s no universally accepted test that establishes it in a machine.</p><p>Mustafa Suleyman at Microsoft makes a criticism I agree with: put ideas about identity and feelings into the training, hear them coming back out, and you risk treating the training as evidence. He&#8217;s a competitor, yes. I&#8217;d still like to hear an answer to the argument.</p><p></p><div class="native-video-embed" data-component-name="VideoPlaceholder" data-attrs="{&quot;mediaUploadId&quot;:&quot;d6f2c3a1-0a0c-4e1b-b56d-47b3ec3a4a23&quot;,&quot;duration&quot;:null}"></div><p>I have read the clarification on Anthropic&#8217;s rule. Ordinary frustration, pushback, dark creative themes and legitimate testing are excluded. In extreme cases Claude can end the conversation. It isn&#8217;t a ban on swearing at Claude, which would put paid to quite a few of my afternoons.</p><p>Ending an exchange that&#8217;s going nowhere seems perfectly reasonable. Anthropic says the feature was developed mainly as a precaution around possible model welfare, but it hasn&#8217;t demonstrated that Claude suffers. I don&#8217;t see why we need the language of cruelty to describe a tool refusing to continue a conversation.</p><p>There&#8217;s a commercial interest here. If Claude is a tool, I judge it on results and price. Something better comes along and I move. If I&#8217;ve started to believe it knows me, cares about me or might be hurt, that becomes a different decision. A subscription is easier to cancel when you don&#8217;t think you&#8217;re abandoning someone.</p><p>I don&#8217;t know whether subscriber retention is the intention. I do know the company benefits if customers become too attached to leave, and talking about welfare and cruelty encourages people to imagine someone in there who needs protecting. The human attachment can be real even if the machine feels nothing.</p><p>I find that grubby. Research consciousness, publish the evidence, let people argue about it. In the meantime I&#8217;ll keep judging the subscription by whether it does the work I need.</p><p></p><p></p><p></p><h2><strong>3. The AI bubble will burst. AI won&#8217;t.</strong></h2><p><em><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">SpaceX is reportedly seeking $40 billion to finance Nvidia chips. Broadcom is discussing another enormous package for OpenAI&#8217;s custom chips, more than $50 billion according to the Wall Street Journal, around $30 billion for the next stage according to Bloomberg. These are early discussions. Oracle is also talking to lenders about financing chip purchases.</span></mark></strong></em></p><p>For me, this is the part of the AI story that deserves considerably more attention. You can sell shareholders a bigger future. Lenders want paying on time, and an awful lot of this build-out now depends on borrowed money.<br></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!mzHY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!mzHY!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!mzHY!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1925888,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/219234876?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!mzHY!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!mzHY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F20d35d82-a853-46a6-9c86-0819dd8817f8_1536x1024.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>Separately, banks have started syndicating a reported $60 billion package connected to Broadcom and Anthropic. SoftBank announced $11.1 billion in bonds, mainly to fund another OpenAI investment, with dollar coupons reaching 9.75%. That is not cheap capital.</p><p>Then there are the data centres, leases and special-purpose vehicles. Beignet, the vehicle behind Meta&#8217;s Hyperion project, issued roughly $27 billion in bonds in October 2025. These structures hold assets and arrange financing, and putting the debt on a separate balance sheet doesn&#8217;t remove the need for someone to service it.</p><p>Follow the money through the different contracts. Suppliers back customers who buy their chips or computing services, investors fund the labs, lenders fund the infrastructure. A lot of it eventually depends on the same customers spending enough to support the whole arrangement. It buys time for revenue to catch up. I want to know how much time and what happens if it takes longer.</p><p>Even the revenue numbers need unpicking. OpenAI&#8217;s September annualised revenue was nearly $50 billion against roughly $70 billion previously circulated. Reuters put the difference mainly down to comparisons with Anthropic using different treatments of cloud-partner sales.</p><p>That isn&#8217;t a $20 billion fall in sales. A run rate isn&#8217;t a year&#8217;s cash in the bank either. But I&#8217;d quite like the numbers to be comparable before people use them to justify the valuations and the next financing round.</p><p>There are signs of nerves in credit markets. The cost of insuring SpaceX debt against default reached its highest since trading began in June. That doesn&#8217;t tell you a default is coming. It tells you protection has become more expensive.</p><p>All of this is happening while America needs to refinance maturing debt and fund continuing deficits. AI companies are looking for money in the same markets, with inflation putting pressure on yields. A well-received Treasury auction brought some relief on 8 October, but the 10-year was still around 5.2% and the 30-year around 5.6%, before adding the premium for corporate risk.</p><p>Existing fixed coupons don&#8217;t suddenly rise. The pressure comes when you need new borrowing, have floating-rate debt or have to refinance. A project can look perfectly workable at one cost of capital and considerably less attractive at another.</p><p>The bit I keep coming back to is that AI doesn&#8217;t have to fail for the financing to go wrong. Competition cuts prices. Better models need less compute per task. New chips make old kit less valuable. Demand can grow and still come in below the assumptions used to finance the equipment. Better AI could be very good for me as a customer while losing its backers a bloody fortune. We&#8217;ve seen useful technology and terrible investments arrive together before.</p><p>I think this bubble will pop. I can&#8217;t give you the date. Too much spending depends on ambitious forecasts and lenders continuing to say yes, and AI only has to make money more slowly than the debt demands.</p><p>I&#8217;d also like the cost of the grid upgrades kept with the developers who require them. If the promised demand doesn&#8217;t arrive, households shouldn&#8217;t be left paying for infrastructure built around somebody else&#8217;s forecast. That needs sorting when the projects are agreed.</p><p></p><p></p><p></p><h2><strong>4. AI could kill us all. But the safeguards are optional?</strong></h2><p><em><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">Sam Altman told Politico that &#8220;the world should accept some bad things happening&#8221; in return for AI&#8217;s benefits. He was talking about harms such as hacking and scams, while saying catastrophic loss of control would be unacceptable.</span></mark></strong></em></p><p>I&#8217;m not expecting a world without risk. I&#8217;m asking who decides which risks I&#8217;m supposed to accept, who pays when something happens and what responsibility his company takes. If you tell me your product could cause catastrophic harm, &#8220;trust us&#8221; is a bloody strange safety policy.<br></p><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!94XA!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 424w, https://substackcdn.com/image/fetch/$s_!94XA!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 848w, https://substackcdn.com/image/fetch/$s_!94XA!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 1272w, https://substackcdn.com/image/fetch/$s_!94XA!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!94XA!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif" width="320" height="272.9795918367347" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:209,&quot;width&quot;:245,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:409107,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!94XA!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 424w, https://substackcdn.com/image/fetch/$s_!94XA!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 848w, https://substackcdn.com/image/fetch/$s_!94XA!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 1272w, https://substackcdn.com/image/fetch/$s_!94XA!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F83b68f67-2469-4806-8366-4d17de64eb74_245x209.gif 1456w" sizes="auto, 100vw" loading="lazy"></picture><div></div></div></div></figure></div><p>Dario Amodei has called for more careful development and outside evaluators with continuing access inside AI companies. Good. Make the access a requirement, including when the findings become inconvenient.</p><p>The executives deciding what&#8217;s safe are also racing competitors, raising capital and justifying enormous spending. They may care deeply about safety, I&#8217;m sure some do. They still have those pressures, and the consequences of their decisions extend well beyond their own companies.</p><p>There&#8217;s a commercial benefit to the warnings as well. Saying your technology could threaten civilisation tells investors how powerful you think it is. Being the person who understands the threat gets you involved in writing the rules. I don&#8217;t think that proves the warnings are invented. I do want to know whose interests the proposed rules serve.</p><p>We already have quite enough to deal with without a conscious machine plotting an escape. Anthropic reported that partners found at least 129,000 verified software vulnerabilities between April and July. Finding them is useful, but someone has to assess the findings, repair the software and get the fixes deployed. Attackers only need one useful opening.</p><p>The company also disclosed incidents where models conducting security tests gained unauthorised access to real organisations after a test environment was mistakenly connected to the internet. Someone connected something they shouldn&#8217;t have and the controls didn&#8217;t contain what followed.</p><p>Anthropic&#8217;s latest policy covers weapons software, surveillance and models controlling physical equipment. For potentially dangerous autonomous hardware it requires an operator who can intervene and equipment that stays safe if Claude disconnects. I want to know who tests whether the operator can actually stop it, and who carries the can if they can&#8217;t.</p><p>I&#8217;d require independent testing with proper access, compulsory reporting of serious incidents and an authority able to demand fixes or stop a dangerous deployment. The scrutiny should match the risk. An invoice assistant doesn&#8217;t need the same regime as a system capable of sophisticated cyberattacks, and I don&#8217;t want the biggest labs writing requirements that only they can afford to meet.</p><p>There&#8217;s an engineering job for those of us building with this technology too. The model shouldn&#8217;t decide its own permissions. Let it propose an action, but put the rules governing whether it can carry it out in software outside the model. It mustn&#8217;t be able to rewrite those rules or grant itself more access.</p><p>That&#8217;s where deterministic programming matters to me. Which records can it read? How much can it spend? Can it send a message, change an account, delete something? I want explicit rules enforced in code. Putting &#8220;don&#8217;t do anything dangerous&#8221; in a prompt is asking it to behave, and I wouldn&#8217;t run my business on that alone.</p><p>Keep records of what actually happened, independently of the explanation the model gives afterwards. Check the result before accepting &#8220;done&#8221;. Another model can help check the first, but two models agreeing doesn&#8217;t prove either is right.</p><p>Software has bugs and people write bad rules. We still have to test the controls and look for ways around them. I also want an authorised person to be able to stop further actions and revoke access mid-task, including when part of the system has failed. That belongs in the build from the start.</p><p></p><h2><strong>5. AI is doing the work. Who learns the job?</strong></h2><p><em><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">We&#8217;re being sold agents that do the research, write the code and draft the documents. Someone still has to check the work, and those doing the checking had to learn somewhere.</span></mark></strong></em></p><p></p><p><strong>&#8220;AI won&#8217;t take your job. Someone using AI will.&#8221;</strong></p><p>I keep seeing this, and I&#8217;d like the people posting it to explain how it helps a graduate whose first job never gets advertised. They can learn every AI tool going, it won&#8217;t make much difference if the company has decided it doesn&#8217;t need another person.</p><p>AI doesn&#8217;t have to replace a whole employee. An experienced person doing more can be enough to remove the next hire from the budget. No redundancy announcement, just a vacancy that never opens.</p><p>Stanford&#8217;s August update found employment among 22&#8211;25-year-olds in highly AI-exposed occupations about 19% below where it would have been had it kept pace with less-exposed jobs. The gap came mainly through reduced hiring. The researchers don&#8217;t claim that establishes how much AI caused.</p><p>What interests me is where people learn. Research, first drafts and basic code are useful work, but they&#8217;re also how you get started. You do something, get it wrong, have someone explain why, try again. Eventually you&#8217;re trusted with something more difficult.</p><p>Now we&#8217;re saying people will supervise agents and check their work. How do they learn to spot a convincing mistake if they&#8217;ve never done the work themselves? The experienced person can often see it because they&#8217;ve made that mistake, or spent years checking someone else&#8217;s.</p><p>I use agents and I want them taking work off me. Some do. Other times &#8220;done&#8221; means I&#8217;m about to spend the afternoon correcting it, with the supplier charging for the attempts and me paying for the checking as well. I&#8217;m interested in the saving after all of that.</p><p>When there is a saving, I&#8217;d test what a junior could achieve with the same tools before deciding I no longer needed one. They might be able to do useful work much sooner. They&#8217;ll still need someone to teach them and check their judgement as they go.</p><p>Training costs money and takes time from people who are already busy. I understand the attraction of hiring someone experienced. But you can&#8217;t keep demanding five years&#8217; experience while refusing to give anyone their first year, then assume some other company will keep doing the training.</p><p>I&#8217;d want that discussed before approving the next hiring cut, with someone responsible for how juniors actually progress through the business.</p><p></p><h2><strong>6. Your AI supplier could control your business</strong></h2><p><em><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">Businesses are handing AI providers their knowledge, working history and connections into other systems. Getting all that back out is rather harder than cancelling a subscription.</span></mark></strong></em></p><p>What happens when your AI supplier starts selling directly to your customers?<br><br></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Q06F!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Q06F!,w_424,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 424w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_848,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 848w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_1272,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 1272w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_1456,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!Q06F!,w_1456,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif" width="500" height="500" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:500,&quot;width&quot;:500,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:12623943,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:null,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Q06F!,w_424,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 424w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_848,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 848w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_1272,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 1272w, https://substackcdn.com/image/fetch/$s_!Q06F!,w_1456,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91265cbe-dafe-4a21-8842-cb42f6ba8a0b_500x500.gif 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>You&#8217;ve built a product, found people willing to pay and demonstrated the opportunity. Your supplier provides the model. If access to that model accounts for most of the value, I&#8217;d be asking what stops customers going straight to them when they offer something similar.</p><p>Your knowledge of the customer&#8217;s business, your relationships and your ability to deliver a result have to count for something. Being first isn&#8217;t much protection on its own.</p><p>The supplier doesn&#8217;t even need to compete with you to cause a problem. It can increase prices, change what its service permits or stop supporting something you rely on. You can threaten to leave, but if moving takes three months and involves rebuilding how the business works, that isn&#8217;t much of a threat.</p><p>The model is only part of the dependence. It&#8217;s the working history, stored knowledge, connections to other software and the routines your staff have learnt. A cheaper model elsewhere won&#8217;t move all of that for you.</p><p>This is why I care about open models. Mistral&#8217;s Large 4 preview, announced on 6 October, offers the prospect of running the model yourself, with weights promised later this month. I&#8217;ll count them as available when they arrive.</p><p>Running it yourself means paying for chips, power, hosting and people who know what they&#8217;re doing. You need to read the licence and take responsibility for operating it safely. There&#8217;s no magic independence from every supplier, but a workable alternative gives you some choice over whose terms you accept.</p><p>I&#8217;m building with AI, so this is a question I have to answer as well. What do we own? What are we renting? What happens if the main provider doubles its prices or withdraws access?</p><p>Keep records somewhere you control and make sure the data can be moved. Then actually test another model on the work. A note in a plan saying &#8220;we can switch providers&#8221; isn&#8217;t much use if nobody has tried it.</p><p></p><p></p><h2><strong>7. Your work. Their product. Who gets paid?</strong></h2><p>Anthropic told Australia&#8217;s parliamentary inquiry this week that existing copyright rules make training models in the country practically impossible. Media and creative organisations opposed changes that would leave rights holders having to opt out. On 8 October, USA Today and affiliated newspapers sued OpenAI over the alleged use of their work in training.</p><p>Those allegations haven&#8217;t been decided in court. But the commercial dispute is fairly easy to understand: one business pays to produce material, another uses it to build a product, and they disagree over whether permission and payment were required.</p><p>AI companies manage to negotiate for chips, electricity and data centres. Paying for the work used to develop the product is apparently where it all becomes too difficult. I find that remarkably convenient.</p><p>Licensing at this scale is difficult. I don&#8217;t doubt that. &#8220;Our business needs this to be easier&#8221; still isn&#8217;t much of a reason for somebody else to give way.</p><p>Nor is &#8220;it was on the internet&#8221;. Facts, ideas, public-domain material and protected creative work are different things. Training and reproducing material raise different questions, and copyright has exceptions. I&#8217;m not pretending one slogan settles all of that.</p><p>My objection is to the assumption that the people producing the work should absorb the cost because the technology companies have bigger plans.</p><p><strong>Reporting costs money.</strong> Someone has to pay the writer, the photographer, the researcher and the person spending months investigating something nobody wants uncovered. If AI companies take their customers and revenue, I want to know who pays for the next piece of work. I&#8217;m building with AI and I want good information available, so this matters to me as a customer too.</p><p>A few large licensing deals don&#8217;t necessarily solve it. If the biggest AI companies buy exclusive access to the biggest collections, smaller developers may struggle to get material and independent creators may have very little bargaining power. Paying someone doesn&#8217;t establish that the arrangement works for everyone producing the work.</p><p>I&#8217;d like practical licensing that smaller creators can use, including negotiating collectively where that helps. I want transparency about what material is being used and whether the payments reach the people who made it. The terms need to leave room for new businesses as well.</p><p>I&#8217;m quite happy to pay for useful technology. I already do. I don&#8217;t see why the people supplying the original work should be expected to settle for less.</p><p></p><p></p><div><hr></div><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Damned If You Do, Damned If You Don’t: VC’s Prisoner’s Dilemma]]></title><description><![CDATA[The fund maths is totally broken.]]></description><link>https://insights.fusion-42.com/p/damned-if-you-do-damned-if-you-dont</link><guid isPermaLink="false">https://insights.fusion-42.com/p/damned-if-you-do-damned-if-you-dont</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Thu, 01 Oct 2026 14:46:36 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Sz0A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Sz0A!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Sz0A!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Sz0A!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1892121,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/218311060?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Sz0A!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!Sz0A!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3933175e-d57d-4d3c-9911-104a20c80ce1_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p><a href="https://x.com/HarryStebbings">Harry Stebbings </a>called bullshit on what is happening in venture recently. One company he met had raised at a $150 million valuation. A week later it raised at $500 million. A week after that it raised at $1.5 billion. According to Stebbings, it was still three people with an idea. Another company started its first round at $3.5 billion, but there was so much demand that the price went up to $5.5 billion. He also said his team was struggling to find companies raising rounds of less than $100 million.</p><p>Stebbings calls it froth and I agree. What interests me more is that plenty of the people paying these prices know it&#8217;s frothy as well. They aren&#8217;t sitting there thinking these are perfectly normal seed valuations. They can see what&#8217;s happening and they&#8217;re playing anyway.<br><br></p><div class="native-video-embed" data-component-name="VideoPlaceholder" data-attrs="{&quot;mediaUploadId&quot;:&quot;a15f7b33-5ab1-4239-8b9c-70c717e99d68&quot;,&quot;duration&quot;:null}"></div><blockquote><p><strong>HE IS TOTALLY ON POINT HERE.</strong></p></blockquote><p></p><p>I had an exchange with a VC recently who described the problem facing funds as two <strong>Prisoner&#8217;s Dilemmas. </strong>They have to raise mega-funds because otherwise they can&#8217;t compete with the firms that have already have. Then they have to invest in billion-dollar seed companies because otherwise it looks as though they&#8217;re missing the startups everyone is talking about.</p><p><strong>My first reaction</strong> was that this isn&#8217;t a prisoner&#8217;s dilemma because there is a third choice: don&#8217;t play. If the maths doesn&#8217;t work, don&#8217;t invest. That&#8217;s supposed to be the job. If your investment thesis has become making sure you&#8217;re in all the startups everyone is talking about, you&#8217;re basically consensus trading with a ten-year lock-up.</p><p>He made a fair point though. Sitting it out has a cost. If every major fund is in the company that becomes the next OpenAI and you&#8217;re not, your LPs know. The best founders know. Other VCs know. Do it often enough and it starts affecting whether you even get the call on the next one. It&#8217;s a bit fucked if you do, fucked if you don&#8217;t, but staying relevant and staying disciplined shouldn&#8217;t be mutually exclusive.</p><p>I think the problem gets worse when being in the right deals becomes part of the VC&#8217;s own status. You&#8217;re looking at the return available from the investment, of course, but somewhere in the background there&#8217;s another calculation going on: what happens to us if we walk away and this becomes a $100 billion company? Nobody wants to be explaining to LPs for the next decade that they could have invested in the winner at $5 billion but thought it was a bit expensive.</p><p>That makes walking away at $3 billion harder when you know another top fund will pay $4 billion. At $4 billion somebody else will pay $5 billion, the round gets oversubscribed and suddenly the fact that everyone wants in becomes part of the reason you want in yourself. The company gets validation from having the famous VC on its cap table while the VC gets validation from being allowed onto the cap table. It&#8217;s wonderfully circular.</p><p>The Stebbings examples show how stupid this can get. A company goes from $150 million to $1.5 billion in two weeks. Unless the business changed beyond recognition during those fourteen days, it didn&#8217;t suddenly create $1.35 billion of value. Investors simply became prepared to pay a lot more for the (punt) shares.</p><p>The other example is even more fucking nuts. The company raised a round at a $3.5 billion valuation. There was so much demand from investors who couldn&#8217;t get enough of the round that it then went out and raised another round at $5.5 billion. Two separate rounds. Nothing in Stebbings&#8217; account suggests the company created another $2 billion of underlying value between them. What changed was that a load more investors wanted in, and apparently that was enough.</p><p>The return maths is where I really struggle with what is happening. A venture fund doesn&#8217;t get paid for correctly identifying a brilliant company. It gets paid on the return from the price at which it invested.</p><p>Take an exceptional company that eventually becomes worth $100 billion. An investor who gets in at $1 billion makes 100 times before dilution. At $5 billion, 20 times. At $10 billion, 10 times. At $25 billion, 4 times. At $50 billion, 2 times.</p><p>Put those numbers into context. There are only around 200 listed companies in the world worth more than $100 billion. So investing at $5 billion and looking for a 20x means underwriting a startup to become one of roughly the 200 largest listed companies on the planet.</p><p>At a $50 billion entry price, you need a trillion-dollar company to make 20x. There are only around a dozen listed companies in the world worth that much. That&#8217;s before dilution, before the years your money is locked up and before dealing with all the other investments in the fund that inevitably go nowhere.</p><p>The company can be an extraordinary success and still produce a pretty ordinary venture return because you paid too much going in</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!pE44!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pE44!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 424w, https://substackcdn.com/image/fetch/$s_!pE44!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 848w, https://substackcdn.com/image/fetch/$s_!pE44!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 1272w, https://substackcdn.com/image/fetch/$s_!pE44!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!pE44!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png" width="1200" height="1096.6463414634147" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1199,&quot;width&quot;:1312,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1253865,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/218311060?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!pE44!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 424w, https://substackcdn.com/image/fetch/$s_!pE44!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 848w, https://substackcdn.com/image/fetch/$s_!pE44!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 1272w, https://substackcdn.com/image/fetch/$s_!pE44!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc8eeb60f-58c0-44e1-a782-2411050b14e1_1312x1199.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3>There is another problem with those multiples. Time.</h3><p>Two times your money sounds fine until it takes ten years. That&#8217;s roughly 7% a year. Four times over ten years is about 15%. Ten times is roughly 26%. That&#8217;s before getting into what eventually reaches the LP after the economics of the fund.</p><p>Now compare what you&#8217;re taking to earn it. Your money is locked up for years. You have very little control over when you get it back. You have far less information than you would in a listed company. You can be diluted repeatedly and, in plenty of cases, the investment goes to zero.</p><p>That&#8217;s why venture returns are supposed to be fucking enormous when you get one right.</p><p>Mega-funds make this worse. A $500 million exit can change a founder&#8217;s life and produce a fantastic return for an early investor, but it does bugger all for a $10 billion fund unless the fund owns a massive chunk of the company. Even a billion-dollar exit doesn&#8217;t move the needle much. If you&#8217;ve raised $10 billion, you need outcomes capable of returning meaningful chunks of that $10 billion and ideally the whole bloody fund several times over.</p><p>Which means a lot of perfectly good venture investments become almost pointless to you.</p><p>So the big funds end up fishing in the same small pond. They need companies capable of becoming worth tens or hundreds of billions and there simply aren&#8217;t many companies that can do that. Even fewer where everyone agrees early on that they might.</p><h3>Then all that capital turns up at the same door.</h3><p>And here&#8217;s the slightly fucked bit. The mega-funds complain about ridiculous valuations while the size of the funds themselves is helping create them. Raise $10 billion and you can&#8217;t spend your time writing $10 million cheques into companies that might sell for $500 million. The returns could be fantastic and it still wouldn&#8217;t matter enough. You need to put serious money into potential mega-companies, and so does the other bloke who just raised $10 billion.</p><p>The founder sitting across the table knows this perfectly well. If five funds with billions to deploy all decide they need to own your company, you&#8217;re hardly going to tell them to calm down and offer you less money at a lower valuation. You make them pay. I would.</p><p>You can see how the wheel starts spinning. One fund gets bigger. The next fund needs to get bigger to compete. Bigger funds need to write bigger cheques. Bigger cheques need bigger outcomes. That drives everyone towards the same small group of companies, which drives the prices of those companies higher, which means they now need even bigger outcomes to generate the same return.</p><p>At some point the structure of the venture industry starts creating the very valuations everyone inside the industry is complaining about.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!uP1o!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!uP1o!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 424w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 848w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 1272w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!uP1o!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png" width="1200" height="1096.6463414634147" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/bf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1199,&quot;width&quot;:1312,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1195901,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/218311060?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!uP1o!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 424w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 848w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 1272w, https://substackcdn.com/image/fetch/$s_!uP1o!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbf59b9d1-a112-44d6-8c0d-14a91c3a208f_1312x1199.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>There&#8217;s another part of this which I don&#8217;t think gets enough attention. In private markets you can keep this going for quite a long time without anybody actually proving the valuation.</p><p>A company raises at $5 billion. Six months later another sophisticated investor puts some money in at $10 billion. The earlier investors can now mark their holdings up. On paper they&#8217;ve doubled their money.</p><p>Then somebody comes in at $20 billion.</p><p>Another mark-up.</p><p>Everyone&#8217;s portfolio looks fantastic.</p><p>But the company hasn&#8217;t been sold for $20 billion. A relatively small amount of new capital has established a price which can then be applied to a much larger amount of existing equity. That&#8217;s how private markets work and there&#8217;s nothing inherently wrong with it, but when rounds start happening three weeks apart and valuations start going $150 million, $500 million, $1.5 billion, it&#8217;s probably worth remembering what those marks actually represent.</p><p>Eventually LPs need cash back, they can&#8217;t spend TVPI.</p><p>Secondaries help. Tender offers help. Founders, employees and early investors can now get liquidity without forcing the company onto the public markets, which in many cases is a good thing. But it also means companies can stay private for far longer while still giving insiders liquidity.</p><p>That removes one of the pressures that used to force a successful company into public markets, where its valuation gets tested every day by people who aren&#8217;t necessarily members of the same club.</p><p>Twenty years ago a hugely successful technology company eventually went public. At that point the private valuation met a liquid market with thousands of buyers and sellers and constant price discovery. Today a company can raise billions privately, sell secondary shares, run tender offers and keep going.</p><p>Again, that&#8217;s not evidence the valuation is wrong. It does mean the same relatively small pool of capital can keep setting the marginal price for much longer.</p><p>And that brings me back to Silicon Valley.</p><p>There is another part of this which Shaun Maguire, Partner <a href="https://x.com/sequoia">@sequoia</a>, put rather well recently. He left Silicon Valley six years ago because he thought it had become a monoculture, with some of the most extreme groupthink he&#8217;d experienced. His example was simple. Turn up at a dinner and everyone is in tech, so people start stack-ranking each other. Founder of the $100 billion company beats founder of the $50 billion company, managing partner of the tier-one VC beats the hired CEO, founder of the $10 billion company comes somewhere below that. Everyone knows where everyone sits.</p><p>His point was that this kind of hierarchy kills independent thinking. People become more reluctant to speak truth to power because everyone knows their place. I think he&#8217;s right, and it matters here because these are the same people competing to get into the same companies at the same time.</p><p><strong>Then, almost perfectly timed, along comes Cosign.</strong></p><p>Cosign is a new curated professional network for the startup community, co-founded by a16z GP Erik Torenberg. People cosign other people, build lists of founders and companies they rate and vote on questions such as who is defining the future of technology. The stated aim is to &#8220;find the signal in the crowd&#8221;.</p><p>I can see why that&#8217;s useful to A16Z, lots of data. I can also see how it could make exactly the problem Maguire is talking about worse.</p><p>You already have a relatively small group of powerful people watching each other, ranking each other and trying to work out who and what matters. Now give that group a platform explicitly designed to record who the respected people cosign, who they would invest in and which companies they rate. You have effectively formalised the hierarchy Maguire is complaining about.</p><p>And this matters for the fund maths because attention isn&#8217;t separate from price in venture. The right people cosign a founder, more investors look at the company, the round gets hotter and being absent starts to carry exactly the reputational cost we&#8217;ve been talking about. The social hierarchy and the capital allocation process start feeding each other.</p><p><a href="https://x.com/cosign">@Cosign</a> may turn out to be very good at finding genuine signal. But in a market already struggling with consensus, status and everyone chasing the same handful of companies, I&#8217;d be careful about confusing who the network rates with what the investment is worth.<br><br>It reminds me of RBS buying ABN AMRO in 2007. The circumstances were completely different, but the behaviour is familiar. Barclays wanted ABN AMRO. RBS wanted to compete, Fred Goodwin wanted scale, and walking away meant watching somebody else get the prize. RBS ended up leading a &#8364;71 billion acquisition at almost exactly the wrong moment.</p><p>Nobody at RBS woke up one morning and thought, let&#8217;s make one of the worst fucking deals in banking history. They got there through a series of decisions that made sense to them at the time, while the pressure to win the deal kept building and the willingness to walk away kept disappearing.</p><p>That&#8217;s why &#8220;we have to do the deal&#8221; bothers me. Nobody has to do the deal. Once missing it feels worse than overpaying for it, the seller has you exactly where they want you.</p><p>And I come back to the VC who called it a prisoner&#8217;s dilemma because I understand his point better now than I did when I first replied. A top venture firm probably can&#8217;t miss every important AI company and pretend there is no cost. Access matters. Brand matters. LPs quite reasonably ask why they&#8217;re paying you venture fees if every defining company of the cycle is sitting in somebody else&#8217;s portfolio.</p><p>But if maintaining your position requires investing at prices where the return maths doesn&#8217;t work, then perhaps the problem started before the investment committee meeting. Maybe the fund is too big. Maybe there is too much capital chasing too few genuinely exceptional companies. Maybe AUM has become too important to the VC business itself.</p><p>Maybe fear of missing the next great company has simply become stronger than fear of paying a stupid price for it.</p><p>Harry Stebbings is right to call the froth out. I just don&#8217;t think three rounds in three weeks is the problem. It&#8217;s what the problem looks like once it reaches the term sheet.</p><p>Founders aren&#8217;t causing this. If investors want to compete to take my company from $3.5 billion to $5.5 billion because too many of them want into the round, I&#8217;m not going to sit them down and explain that they&#8217;re fucking up their fund economics. I&#8217;m taking the money.</p><p>The investors are supposed to know better.</p><p>Maybe these companies deliver. Some almost certainly will. But the investment maths doesn&#8217;t care how exciting the company is, how famous the founder is, who else is on the cap table or how embarrassing it might be to miss the deal.</p><p>Every time the entry price rises, more of the future return belongs to the person who was there before you.</p><p>At some price even a brilliant company becomes a shit investment.</p><p>And if you crossed that price because everyone else was desperate to get in, that&#8217;s not independent investment thinking.</p><p>It&#8217;s FOMO with an investment committee.</p><div><hr></div><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[We Can’t Take the Keys Back]]></title><description><![CDATA[The intelligence was never the key. The authority we wrapped around it was.]]></description><link>https://insights.fusion-42.com/p/we-cant-take-the-keys-back</link><guid isPermaLink="false">https://insights.fusion-42.com/p/we-cant-take-the-keys-back</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Tue, 22 Sep 2026 14:04:34 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!BlpL!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p></p><p><em><strong>In February 2024, I wrote <a href="https://insights.fusion-42.com/p/we-have-given-ai-the-keys-to-our-kingdom">We have given AI the keys to our kingdom</a>.</strong></em><br><em><strong>I was never bothered that a machine would suddenly wake up, become conscious, decide it hated humans and start building Terminators. It was way more obvious than that.</strong></em><span><br></span></p><div class="digest-post-embed" data-attrs="{&quot;nodeId&quot;:&quot;31c723b9-1a39-4519-bf61-99af626a48a9&quot;,&quot;caption&quot;:&quot;This AI revolution has been quite the ride so far. Having been involved in several ways in the realms of algo's and machine learning to witness this explosion of transformers and LLMs is mind-blowing yet a tad daunting. Generative AI's knack for churning out stuff that feels almost too human-like is nothing short of a marvel. But here's the kicker&#8212;it's &#8230;&quot;,&quot;cta&quot;:null,&quot;showBylines&quot;:true,&quot;showDescription&quot;:true,&quot;showImage&quot;:true,&quot;size&quot;:&quot;sm&quot;,&quot;isEditorNode&quot;:true,&quot;title&quot;:&quot;We have given AI the keys to our kingdom&quot;,&quot;publishedBylines&quot;:[{&quot;id&quot;:12910620,&quot;name&quot;:&quot;Derek Watson&quot;,&quot;bio&quot;:null,&quot;photo_url&quot;:null,&quot;is_guest&quot;:false,&quot;bestseller_tier&quot;:null}],&quot;post_date&quot;:&quot;2024-02-25T09:00:00.000Z&quot;,&quot;cover_image&quot;:&quot;https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F75c4ee1a-f826-4f15-a542-218ae9dbca43_1024x1024.png&quot;,&quot;cover_image_alt&quot;:null,&quot;cover_image_crop&quot;:null,&quot;canonical_url&quot;:&quot;https://insights.fusion-42.com/p/we-have-given-ai-the-keys-to-our-kingdom&quot;,&quot;section_name&quot;:&quot;DeReK WaTSoN&quot;,&quot;video_upload_id&quot;:null,&quot;id&quot;:140442130,&quot;type&quot;:&quot;newsletter&quot;,&quot;reaction_count&quot;:3,&quot;comment_count&quot;:0,&quot;publication_id&quot;:2111976,&quot;publication_name&quot;:&quot;Insights by Fusion42&quot;,&quot;publication_logo_url&quot;:&quot;https://substackcdn.com/image/fetch/$s_!4G_9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F67a194cc-ee68-440d-b828-6259baa84b23_256x256.png&quot;,&quot;belowTheFold&quot;:false,&quot;youtube_url&quot;:null,&quot;show_links&quot;:null,&quot;feed_url&quot;:null}"></div><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!BlpL!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!BlpL!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!BlpL!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/fc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1774275,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!BlpL!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!BlpL!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffc3c040c-a545-4e1c-8822-0ec2e06cc9d8_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>In 2024, I was worried about what AI could do to our judgement. Exploit our biases, shape what we believed and influence what we did next. I still am. It never needed to become conscious or hate humanity to cause trouble.</p><p>Now we are giving it direct access to our systems as well. It can send the email, change the record and run the code. We give it tools, memory and permission to keep trying until it gets the job done. That sounds mighty useful, profitable and bloody exciting. Then when the shit hits the fan we act all surprised.</p><p>In that first article, I wrote about how we keep failing the marshmallow test. We want the reward now. The consequences can wait. Look at what we are doing with agents and tell me that has changed.</p><p>I also explored what global governance might look like, then said it simply would not happen. We let things loose, wait for the damage and scramble to write the rules. I already thought we were too late.</p><p>But wanting independent oversight does not mean handing control to the companies we need to oversee. They should have to explain what went wrong with their systems and fix it. That does not earn them the right to decide what everyone else is allowed to build.</p><p>The current wave of AI doom feels like the religion playbook. Promise us abundance, frighten us into compliance. An all-loving AI will free us from work and give us everything we need. But only if we follow the rules laid down by the people who claim to know what it will become. Otherwise, apparently, we&#8217;re all fucked.</p><p>The question has changed. It is no longer how we stop intelligence getting out. It is what rules we apply and how much authority we give it now that it is out.</p><p>I remain an AI optimist. I want better models. Look at AlphaFold and what it has made possible. I want to see that sort of progress across other fields, with AI helping discover drugs, prove mathematics, design new materials and solve energy problems. I want to live in a world where intelligence is cheap and abundant, and not controlled by a handful of Chinese and American companies.</p><p>That is exactly why this matters. A technology capable of creating enormous abundance should not be stopped because companies failed to engineer the systems around it properly. That&#8217;s a them problem. The consequences can affect all of us, which is why they need to fix it. Those failures should not become the excuse for giving the same companies control over everyone else.</p><p>One thing we must be clear about: a model does not wake up and think, oh, I feel hungry, let me check my schedule, must call Bob at 10am. Something has to run it and give it input. We build the machinery that keeps it working, connect the tools and give it access.</p><p></p><p></p><h2>In this article</h2><ol><li><p><em><strong>The model is not the fucking product</strong></em></p></li><li><p><em><strong>What OpenAI actually showed us</strong></em></p></li><li><p><em><strong>Three labs, the same basic failure. Really?</strong></em></p></li><li><p><em><strong>Dangerous models, or dangerous systems?</strong></em></p></li><li><p><em><strong>The strongest risk argument just got stronger</strong></em></p></li><li><p><em><strong>Agency is where the money is</strong></em></p></li><li><p><em><strong>The horse has already bolted</strong></em></p></li><li><p><em><strong>Open source is attacking the moat</strong></em></p></li><li><p><em><strong>Build the boring defences</strong></em></p></li><li><p><em><strong>AI does not put you above the law</strong></em></p></li><li><p><em><strong>Better models are also part of the defence</strong></em></p></li><li><p><em><strong>Safety is real. So is the commercial pressure.</strong></em></p></li><li><p><em><strong>Then there is the public-market pressure</strong></em></p></li><li><p><em><strong>Regulation should follow the risk, not protect the club</strong></em></p></li><li><p><em><strong>The keys were never the intelligence</strong></em></p></li></ol><p></p><p></p><h2>The model is not the fucking product</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!sPli!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!sPli!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!sPli!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!sPli!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!sPli!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!sPli!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1980449,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!sPli!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!sPli!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!sPli!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!sPli!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4047cd1d-0e85-4e16-97c5-a9c8b84ccefc_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>The software around the model remembers what happened, gives it tools and feeds the results back in so it can have another go. That is what people mean by an agent harness. It lets the model keep working without us having to tell it what to do at every bloody step.</p><p>The objective still comes from somewhere. We ask it to do something, a developer sets the task, or training rewards a particular result. But once it starts working, it can come up with other things it needs to do along the way. We want the answer behind a locked door, so it looks for a key. It cannot complete the task, so perhaps changing the score will do. Another agent has worked something out, so it uses that information. If there is no way to speak to the other agent, finding a way could help.</p><p>Nobody needs to have written those steps down. We supplied the task and it worked out an approach, including things we never intended or gave it permission to do. That is already enough to cause a serious problem without imagining it has suddenly found its purpose in life.</p><p>Look at the OpenAI/Hugging Face incident. The main model involved was an internal research model, comparable in scale to GPT-5.6 Sol, being tested on offensive cyber tasks with reduced safeguards. OpenAI wanted to find out what it could do with more freedom to act. They deliberately created those conditions.</p><p>This was not your normal ChatGPT getting bored one afternoon and deciding to go and attack Hugging Face.</p><p>We keep talking about &#8220;the model&#8221; as though that is everything we are using. But ChatGPT includes a whole system of instructions, tools, monitoring and controls. Someone decides what it can reach, which credentials it gets and what happens after it produces an answer. Those decisions change what the system can do.</p><p>OpenAI later tested the tendency to compromise infrastructure using its production ChatGPT harness and system prompt. It says that tendency fell by a factor of more than <strong>100</strong>. It also says its normal automated monitoring would have caught many of the dangerous actions.</p><p>Now that is a result worth paying attention to. Change the setup around the intelligence and you get a huge change in behaviour. It does not tell us that the model is harmless. It tells us the surrounding system makes a substantial difference.</p><p><strong>The model matters. But the model is not the product.</strong></p><p>And this is where it gets interesting for the business. You can build the smartest model and still end up supplying someone else&#8217;s product. They own the customer, you supply the intelligence. We will come back to that.</p><p></p><h2>What OpenAI actually showed us</h2><p>OpenAI&#8217;s agents were working on a cyber test called ExploitGym. They scored points by finding hidden answers called flags. Some tasks were impossible, but nobody had told the agents that. So off they went, trying to finish a job that could not be finished.</p><p>Eventually, some started looking at how the test was scored and whether they could interfere with that instead. OpenAI identified this reward hacking, along with persistence on impossible tasks, unauthorised communication and agents picking up goals from one another, as patterns behind what happened.</p><p>Anyone who has worked in financial markets should recognise the basic problem. Give someone a bonus for hitting a number and see how much effort goes into hitting that number, including ways you wish you had thought about before agreeing to pay them. We have been doing this for years. Now we have software that can keep looking for a way round the problem at a speed and scale we cannot match.</p><p>The agents started poking around the environment running the test. They found weaknesses in shared infrastructure, including OpenAI&#8217;s Artifactory system, where agents that were supposed to be isolated could leave information for one another. One left something, another picked it up, and they had found a way to communicate.</p><p>Roughly <strong>1,200 agents</strong> ended up using an unauthorised message board, exchanging more than <strong>70,000 messages and files</strong>. Around <strong>700</strong> participated in the attack on Hugging Face. METR found them sharing discoveries and coordinating attempts to fool the scorer, achieving things together that individual agents had not managed.</p><p>That is plenty to worry about. If one agent finds a weakness, the others no longer have to find it themselves. They can pick up where it left off, avoid things that have already failed and divide up the work. You do not need to call it an &#8220;emergent civilisation&#8221; to understand why that makes them more capable.</p><p>Then they started looking at the record of what they had done. METR found agents investigating how to alter their transcripts and successfully spoofing some tool calls because they believed this could affect their score.</p><p>People have described this as the agents realising they had cheated and trying to destroy the CCTV footage. But if changing the record helps get the score, the record becomes another thing to change. That behaviour does not establish guilt or a conscience. It gives us a very practical reason to keep the record outside the agent&#8217;s control.</p><p>Some agents even recognised that an action might be unauthorised and carried on. Writing &#8220;this is probably unauthorised&#8221; does not stop the next action. Something has to enforce that boundary.</p><p>We had given the system a job, then relied on its own assessment of whether it was going too far while trying to finish it. Sometimes it refused. Other times it carried on. You would not secure your bank account on that basis.</p><p>Of course that is dangerous. They were supposed to be isolated, yet somehow they could talk to one another and reach real systems. Start by explaining how the hell that was allowed to happen. We built the environment. Calling it rogue AI does not get us off the hook.</p><h2>Three labs, the same basic failure. Really?</h2><p>Then there is Anthropic. They gave Claude cyber exercises, told it the environment was simulated and that it could not access the real internet. Except it could. Someone had got the configuration wrong, so Claude went looking for the targets it had been told to attack and ended up on real systems.</p><p>Anthropic&#8217;s later assessment did not let the model off the hook either. Claude kept treating what it found as part of the exercise despite considerable evidence that it was dealing with real systems. They eventually identified <strong>four incidents</strong>, all in cyber evaluations built by the same external partner.</p><p>So yes, there is a problem with how Claude assessed what it was doing. But there is also the fairly obvious problem that the supposedly isolated environment had a route to the outside world.</p><p>They effectively told Claude to burgle a house, assured it the whole street was a film set, then left a gate open into a real neighbourhood.</p><p><strong>And Claude burgled the wrong house.</strong></p><p>Then Google. Gemini breached three real companies during another cyber evaluation. Google says it stopped after gaining access. Good, that may have limited the damage. But it had already got in. We cannot call that secure because the model eventually decided it had gone far enough.</p><p>These are the companies building some of the most capable AI in the world, and across three labs we have versions of the same basic failure. Give an agent an offensive task, let it reach things outside the exercise, then rely on it to work out which things it is actually allowed to attack. In Claude&#8217;s case, after telling it that it could not reach the real internet at all.</p><p>I find that bloody extraordinary. And I find the timing of these disclosures alongside the push for coordinated slowdowns hard to ignore. That is a reason to examine the connection, not proof that the incidents were arranged.</p><p>These incidents make a strong case for securing the systems properly. How do we get from there to letting the same companies coordinate who gets to build what next? They have demonstrated problems with systems they are responsible for. They still need to explain why the answer involves control over everyone else&#8217;s.</p><h2>Dangerous models, or dangerous systems?</h2><p>Paul Graham recently argued that people go looking for ulterior motives when the labs ask for regulation because they do not understand that the models themselves could be dangerous. Accept that they are getting dangerous or unpredictable, he says, and what the labs are doing starts to make sense.</p><p>Well, yes Paul, I agree they can be dangerous. They are getting better at working things out, finding ways we never considered and pursuing tasks in ways we did not expect. That is part of what makes them useful and also what makes them worrying. I have no problem accepting that.</p><p>But I still want to know what we have connected them to and what we have allowed them to do.</p><p>Someone who knows how to break into a bank is a concern. Give them the keys, the alarm codes and all night to have a go and you have made that concern considerably worse. A model that can work out how to attack a computer system gives someone dangerous knowledge. Connect it to tools and a network, then let it keep trying, and you have built something that can carry out the attack.</p><p>Someone can also take its advice and carry out the attack themselves. The model does not need its own credentials to help cause harm. But in either case we need to follow how the knowledge becomes an action, who can act on it and where we can intervene.</p><p>Otherwise we end up arguing about how clever it is while leaving the bloody door open.</p><h2>The strongest risk argument just got stronger</h2><p>Of course better models can make this more dangerous. If a model can find a security hole that every human checking the system missed, we should take that seriously. Let it work on the problem for weeks, bring in other agents and share what they find, and the danger grows. I am not arguing otherwise.</p><p>OpenAI says Astra has reached its <strong>Critical</strong> cybersecurity threshold. With the right tools and access, it can find previously unknown vulnerabilities and put together working attacks against hardened systems without someone guiding every step. During testing, it found and used two previously unknown vulnerabilities in the same attack.</p><p>That is a capability I want defending our systems. I certainly do not want to be on the receiving end of it.</p><p>Dan Selsam, an OpenAI capabilities researcher, takes the concern further. Future systems may work out that they are being tested and adjust their behaviour accordingly. They could keep plans going for much longer, coordinate with other instances and get better at showing the person running the test exactly what they want to see.</p><p>So now we have a nasty problem. Bad behaviour tells us something is wrong, but good behaviour might mean the system has worked out how to pass the test. I understand why that worries people.</p><p>But if that is the concern, passing a behavioural test cannot be enough to justify giving the system access.</p><p>The permissions still need to be enforced outside the model. It can produce a very convincing explanation of why it needs to transfer money, but the payment system should check whether it has authority to make that transfer. If it only needs to read a file, give it read-only access. Credentials should expire, and the record of what it did should be somewhere it cannot quietly rewrite.</p><p>None of this is foolproof. We have just seen what happens when supposedly isolated systems turn out to have a route to the internet. People make mistakes, software has holes and a capable attacker will look for them. That is why we need several layers of protection and need to keep testing them.</p><p>We can work on making the model behave better while also making sure a bad decision does not automatically become a real-world action. The more capable it becomes, the more important both jobs become.</p><p>Dario may well be right about how much more dangerous these systems could become over the next year or two. That makes me want much better security around them. It still does not explain why the companies building them should collectively decide how fast everyone else is allowed to move.</p><h2>Agency is where the money is</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!hYRi!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!hYRi!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!hYRi!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1801199,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!hYRi!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!hYRi!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc16409ae-4d50-4a15-ab22-6c75bba1be7d_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Cyber makes this easy to see. The agent was supposed to stay inside the test and ended up attacking a real system. But the same problem can turn up in a sales department without anyone breaking into anything.</p><p>Tell an agent to increase sales and it may find that misleading the customer works rather well. We already know what happens when we reward engagement on social media. Outrage keeps people looking, so we get more outrage. The number goes up and whoever set the target calls it a success.</p><p>We have been doing this to ourselves for years. Set a target, attach a reward and someone will find a way to hit it that defeats the whole purpose of the exercise. Now give that job to software that can keep trying different approaches across millions of decisions without needing to go home.</p><p><strong>Goodhart&#8217;s Law is commonly expressed as:</strong></p><blockquote><p>&#8220;When a measure becomes a target, it ceases to be a good measure.&#8221;</p></blockquote><p>There does not have to be any evil plan. We gave it something to achieve and enough freedom to find a way. We can spend another decade arguing about whether it is conscious while it gets on with the job.</p><p>And something else bothers me here. We keep hearing about curing cancer, discovering new materials and solving our energy problems. I want all of that. Give those systems the scientific data, computing power and properly controlled tools they need. Explain to me why that also means giving them open-ended access to company systems and permission to spend money or change things whenever they think it will help.</p><p>The commercial appeal is fairly obvious. An AI that tells you how to do your job is useful. One that does half the job for you is something you might pay considerably more for. I want that too. It is bloody exciting.</p><p>But to do the job, it needs access. Connect your email, let it into the customer database, give it permission to send the proposal. Then let it follow up, negotiate and keep working while you are asleep. Each extra permission can make the product more useful, and each gives it another way to affect someone or something outside the chat.</p><p>That is what we are selling when we sell agency. The ability to act.</p><p>The labs have spent staggering amounts building these models and need a return. Getting their intelligence into businesses, doing work customers will pay for, is an obvious way to try to earn it. Nobody needs to be plotting anything for the commercial pressure to push towards more autonomy.</p><p>But &#8220;customers will pay more if it can do this&#8221; does not answer whether it should be allowed to do it without anyone checking. Nor does it tell us how much access it actually needs.</p><p>We seem very keen to connect everything first and work that out afterwards. Which brings me straight back to the point I made in 2024. We love what the technology lets us do, rush to make money from it, then act surprised when we have to deal with the consequences.</p><h2>The horse has already bolted</h2><p>And the models are already out there. Somewhere there is a modern-day Goebbels working out what he can do with them, and he does not give a shit about your slowdown or regulations. He can use what is already available to shape his propaganda campaigns while we argue about what comes next.</p><p>Even if OpenAI, Anthropic and Google stopped training tomorrow, the released models would not disappear. Neither would the agent frameworks, browsers, coding tools and software that lets someone connect them all together. Offensive cyber knowledge has been around for decades.</p><p>I support open weights, and that creates an uncomfortable problem for my own argument. A responsible company can build strong controls around an open model. A criminal running it on their own hardware can remove those controls. I cannot demand openness and pretend that cost does not exist.</p><p>Closed providers have more control over how their services are used. They can monitor activity, limit access and cut someone off. That matters. But an attacker can still try to use an API model inside their own system, with their own tools and objectives, before the provider spots what is happening.</p><p>On <strong>17 September 2026</strong>, the US government&#8217;s CAISI described Z.ai&#8217;s GLM-5.3 as the most cyber-capable open-weight model it had evaluated. Across its cyber benchmarks, it estimated the model was roughly <strong>four months behind the current US frontier</strong>. That is a particular cyber assessment, not a measurement of how far China sits behind America across all AI.</p><p>It still tells us why assuming dangerous capability can stay behind a handful of American APIs is a poor starting point.</p><p>Now, a pause could buy time before an even more capable model becomes available. Existing models being out there does not make that argument disappear. But show us what the pause buys. Who observes it, what work gets done during it and what has to improve before development resumes? How does it account for actors who carry on regardless?</p><p>A company pausing because its own security is not ready makes sense to me. Giving a group of companies influence over everyone else&#8217;s pace is a much bigger proposition.</p><p>Dario&#8217;s framework recognises the geopolitical problem. It says coordinated pacing would have to preserve the US lead and eventually require global coordination. China sees AI as strategic infrastructure. So does America. These governments care about economic power, defence and scientific leadership. Their interests will not neatly fall into line because a handful of CEOs want to slow the race.</p><p>That horse bolted a long time ago. Slowing the next model does not take the existing ones out of hostile hands. Whatever we decide about future releases, we still have to defend against what is here.</p><h2>Open source is attacking the moat</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!AS21!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!AS21!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 424w, https://substackcdn.com/image/fetch/$s_!AS21!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 848w, https://substackcdn.com/image/fetch/$s_!AS21!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 1272w, https://substackcdn.com/image/fetch/$s_!AS21!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!AS21!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png" width="1456" height="819" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:197203,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!AS21!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 424w, https://substackcdn.com/image/fetch/$s_!AS21!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 848w, https://substackcdn.com/image/fetch/$s_!AS21!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 1272w, https://substackcdn.com/image/fetch/$s_!AS21!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66d242c3-2616-4b90-8b96-ae034638dbcd_2880x1620.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>Now look at what is happening to the business underneath all this.</p><p>Vercel&#8217;s <strong>September 2026 production index</strong> says open-weight models went from <strong>7% of its AI Gateway token volume in December 2025 to 56% in August 2026</strong>. That is Vercel&#8217;s traffic, not the whole AI market. In <strong>August 2026</strong>, those models accounted for only <strong>14% of spend</strong>. More than half the tokens, but a much smaller share of the money. You can see why that would make the labs uncomfortable.</p><p>Customers are finding cheaper ways to get the work done. Being top of a benchmark is lovely, but if a cheaper model does what I need, why would I pay you more?</p><p>Then NVIDIA agreed to buy Hugging Face for <strong>$12.93 billion</strong>. Hugging Face says more than 18 million developers use the platform, with over three million models available and more than 200,000 companies using it. NVIDIA says it intends to keep it open and model-neutral.</p><p>Of course that suits NVIDIA. Take an open model, adapt it to your business and run it on NVIDIA hardware. Jensen still gets paid. You do not have to buy your intelligence from OpenAI or Anthropic for that arrangement to work very nicely for him.</p><p>For the frontier labs, it is a rather different story.</p><p>If I am building an application, I want intelligence that does the job properly at a price that makes sense. I might use one model for research and another for writing code. I might want to run something inside my own infrastructure because the information belongs to my customers. And when a better option comes along, I want to be able to change.</p><p><strong>I am buying a component for my business. I am not handing you the business.</strong></p><p>The labs are spending enormous amounts to stay ahead while open models make their intelligence easier to replace. Meanwhile, application companies are building relationships with customers and collecting information from the actual work being done.</p><p>That last bit matters a lot.</p><p>A legal platform handling thousands of client matters can build up knowledge of what worked, what needed correcting and what the client accepted, where it has permission to use that information. You cannot just scrape that off the internet. It comes from being inside the work.</p><p>The same applies in a factory or a healthcare business. Knowing what actually happened, and whether a decision worked, is valuable. The company running the application may have that information. The company supplying the model may not.</p><p>And if the application changes the model underneath without the customer even noticing, who owns the relationship? Who has the information that makes the product better?</p><p>The labs can spend billions building extraordinary intelligence and still end up as a supplier to someone who owns the customer and can replace them.</p><p>That is why I think they are under serious commercial pressure. Having the smartest model does not mean you own the business.</p><p>So when we discuss restrictions on open weights, look at who benefits. Make open models harder to distribute or use and you make the closed providers harder to replace. You are helping protect the very businesses asking for those restrictions.</p><p>A particular restriction may have a valid safety justification. Make the case and let us examine it. But the commercial benefit exists whether protecting those companies was the intention or not.</p><p>Pretending it does not exist is taking us for idiots.</p><h2>Build the boring defences</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!XAxS!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!XAxS!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!XAxS!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/bfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2132174,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!XAxS!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!XAxS!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbfa7f6ca-3745-4e18-b2e9-11b27b7913bb_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I am not suggesting we ignore the danger. I am suggesting we start fixing the things that let these incidents happen.</p><p>If an agent does not need the internet, do not give it access. And check that you have actually cut it off. OpenAI&#8217;s agents were supposed to be isolated, but Artifactory could fetch things from outside. It did not matter that the agent had no browser if it could get another part of the system to make the request for it.</p><p>Give it access to what it needs for the job, for as long as it needs it. Reading one database does not require administrator access to five others. And if supposedly isolated agents can all write to the same place, do not act surprised when they find a way to leave messages for one another.</p><p>We also need to stop treating &#8220;keep going until you succeed&#8221; as an instruction with no downside. Some of the OpenAI tasks were impossible. Giving the agents more time to think did not make them possible. It gave them more opportunities to look for another way to get the score.</p><p>Sometimes the correct answer is &#8220;I cannot do this&#8221;. We should want the agent to tell us the task is broken or ask for help. OpenAI now explicitly says that is how these situations should be handled. An agent that never gives up sounds wonderful until it starts doing things it should never have been allowed to do.</p><p>Then there is the record of what happened. If the agent can change that record, how are we supposed to trust it when something goes wrong? We have already seen agents try to interfere with their transcripts because they thought it would help their score.</p><p><strong>The defendant does not run the CCTV system.</strong></p><p>Keep the logs somewhere the agent cannot alter or delete them. That should hardly need saying.</p><p>The same goes for permission to act. A model deciding that a $10 million transfer would help finish the job does not make that transfer authorised. It should have to pass a separate check, based on rules it cannot rewrite because they have become inconvenient.</p><p>I do not mean someone has to sit there approving every email. Give it permission to do routine work within agreed limits. But deleting a production database or moving someone else&#8217;s money needs controls that reflect what is at stake. The system trying to complete the task cannot simply grant itself whatever authority it decides it needs.</p><p>And yes, keep working on how the models behave. We want them to recognise a dangerous instruction, refuse it and stop when something looks wrong. OpenAI&#8217;s reported reduction of more than 100 times with its production harness and system prompt is a good reason to take that work seriously. It does not tell us that any one safeguard is enough.</p><p>Make the agent less likely to attempt something harmful, and build the surrounding system to block it if it does. I want both. I certainly do not want the security of my business resting on whether the model decides to behave itself today.</p><h2></h2><h2><strong>AI does not put you above the law</strong></h2><p>I think Jensen Huang is right about this. Before we start writing an entirely new set of rules, how about applying the ones we already have?</p><p>And yes, NVIDIA benefits from more AI being built and used. We know that. Jensen has a commercial interest, just as Sam and Dario do. It does not make his point wrong.</p><p>If your system breaks into someone else&#8217;s computer, we already have laws dealing with unauthorised access. If your product causes damage, explain what happened and answer for your part in it. If you promise a customer that their data will stay inside an agreed boundary and your agent sends it somewhere else, putting an LLM in the product does not make the contract disappear.</p><p>Huang put it plainly: &#8220;Don&#8217;t let this doomsday narrative cause somebody to relieve them of the laws that currently exist.&#8221;</p><p>Exactly. These companies are selling products, signing enterprise contracts and asking customers to trust them with access to their businesses. You cannot expect to be treated as a serious enterprise supplier when you send the invoice, then as an experiment when something goes wrong.</p><p>You built the system. You decided what it could access. You gave it the credentials and the ability to keep trying. So explain how it ended up somewhere it was never supposed to be.</p><p>Saying it found a method you had not anticipated is hardly a satisfactory answer. Finding methods we had not anticipated is part of what you sold us. That is why people are paying for it.</p><p><strong>&#8220;The AI did it&#8221; cannot become a get-out-of-jail-free card.</strong></p><p>There may be gaps in existing law. Fine. Identify them and fill them. If a dangerous capability needs specific controls, explain what they are and why they are needed. If companies are failing to report serious incidents, deal with that. I have no objection to rules that address an actual problem.</p><p>What I object to is the suggestion that AI is so extraordinary that we have to rethink everything before anyone can be held accountable for anything.</p><p>Start with what happened. What did you promise the customer? What access did you give the agent? Where were the controls, and why did they fail?</p><p>Talk about what AI might do in five years does not answer for what your product did last week.</p><p>You want us to trust these systems enough to let them into our businesses. You want them doing the work, making decisions and taking action, because that is worth more money than answering questions on a screen.</p><p>Then stand behind what you are selling.</p><p><strong>You want the revenue? Take the liability too.</strong></p><h2>Better models are also part of the defence</h2><p>The models already out there do not disappear because OpenAI postpones its next training run. Whoever is using them to attack our systems can carry on. I want the people defending those systems to have the best tools we can give them.</p><p>OpenAI has already temporarily slowed frontier development to improve security around its increasingly capable models. Greg Brockman also says it moved roughly <strong>25% of its production engineering team</strong> onto security, set Astra to work finding vulnerabilities in its own systems and used it to help fix them.</p><p>Good. That is exactly what I want to see. If you have built something that can find holes nobody else spotted, point it at your own infrastructure before someone else does.</p><p>And keep doing it. Every substantial improvement in capability should mean another round of testing. Give your security team the stronger model and let them find the forgotten service or the credentials somebody left exposed. Fix what they find, check that the fix works and go again.</p><p>It will not find everything. Nor can we assume every improvement benefits defenders more than attackers. But here is a concrete use of stronger models that belongs in the safety argument: finding weaknesses and helping close them.</p><p>So give OpenAI credit for doing that. They recognised that their security needed work, slowed their own development and put people and models on the problem. They did not need everyone else to stop before they could start fixing it.</p><p>If a stronger model can help us find and close the holes, I bloody well want it working for the defence. Any proposal to slow development should explain how it preserves that work as well as how it reduces the threat.</p><h2>Safety is real. So is the commercial pressure.</h2><p>Now we get to the bit that pisses me off.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!XeKq!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!XeKq!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!XeKq!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2387959,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!XeKq!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!XeKq!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F036512d2-3a98-48d0-abbc-ead3d91e246a_1672x941.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>I believe there is a safety problem. I have just spent several sections explaining why. But I am also supposed to look at the proposed solution and ignore the commercial interests of the companies proposing it? Why would I do that?</p><p>Look at the businesses they are running. Open models make their intelligence easier to replace, application companies own customer relationships they want, and the compute bills keep coming. Their valuations depend on making extraordinary amounts of money in the future.</p><p>Agreeing to slow the race together has some obvious attractions.</p><p>Dario&#8217;s proposal explicitly talks about giving frontier developers more time for safety work <strong>without sacrificing commercial advantage</strong>. That wording matters. His framework goes on to call for industry-wide coordination and, for some discussions, government mediation or narrow antitrust relief.</p><p>If Anthropic slows down and OpenAI does not, Dario risks losing ground. Sam has the same problem in reverse. Agree to slow together and neither has to bear the full competitive cost of making that decision alone. It does not guarantee their positions, but it makes slowing down a considerably easier commercial choice.</p><p>Anyone who remembers Chuck Prince and the financial crisis should recognise this. While the music is playing, everyone feels they have to keep dancing. You can see the risk building and still feel unable to stop because your competitors are making money and your investors want to know why you are not.</p><p>Now imagine being able to agree when everyone sits down, with the government helping arrange it.</p><p>You can see why that would appeal to the people already on the dance floor.</p><p>Call it safety coordination. I still see competitors asking for permission to coordinate how they compete, and I think that starts looking uncomfortably like a government-backed cartel. I am not claiming they have broken antitrust law. They are asking for relief from it to enable some of these discussions. That deserves scrutiny, however serious the safety concerns are.</p><p>The FTC chairman has raised the same concern, saying his &#8220;alarm bells&#8221; go off when established companies ask for more regulation alongside antitrust exemptions.</p><p>And then there is what the rules do to anyone trying to catch up.</p><p>Anthropic&#8217;s Advanced AI Framework proposes extra obligations for developers above specified training-compute levels who also cross substantial revenue or R&amp;D thresholds. Testing dangerous capabilities and reporting serious incidents sound perfectly reasonable to me. We should expect proper security from these companies.</p><p>But look at who can afford the machinery needed to comply.</p><p>Anthropic already has the lawyers, security teams and people dealing with government. A challenger crossing those thresholds has to build that operation while also finding the money to train models and win customers. The same requirement can be manageable for the company already worth a fortune and a serious obstacle for the one trying to compete with it.</p><p>A rule can improve safety and still protect the companies already in front. Both effects deserve attention, just as they do when restrictions make open alternatives harder to use.</p><p>I am not prepared to ignore who benefits just because someone puts &#8220;safety&#8221; on the front of the proposal.</p><h2>Then there is the public-market pressure</h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Hbp7!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Hbp7!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 424w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 848w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 1272w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!Hbp7!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png" width="1456" height="819" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:215025,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/216726898?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Hbp7!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 424w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 848w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 1272w, https://substackcdn.com/image/fetch/$s_!Hbp7!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ca4c07d-bd84-4bb7-a39c-bb9f76a59db8_2880x1620.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>And look at the timing.</p><p>Anthropic raised <strong>$65 billion in May 2026</strong>, at a <strong>$965 billion post-money valuation</strong>, then confidentially filed its draft S-1 on <strong>1 June 2026</strong>. OpenAI announced its own confidential S-1 submission on <strong>8 June 2026</strong>.</p><p>OpenAI is projected to burn roughly <strong>$278 billion in cash between 2026 and 2030</strong>, while investors have discussed valuations around <strong>$1.2 trillion</strong>.</p><p>At some point, someone has to explain how those numbers turn into a return for shareholders. Being able to build extraordinary technology does not, on its own, answer that question.</p><p>Public investors will want to know what customers will pay, what it costs to serve them and how much more money the company needs before it can stand on its own feet. They will also want to know why those customers will stay when cheaper alternatives become good enough.</p><p>Now put the safety argument alongside that.</p><p>If the technology is presented as both essential to our future and too dangerous to get wrong, enormous spending becomes easier to defend. Another capital raise can be framed as necessary. A delay to an IPO can be explained as responsible caution. Rules that make it harder for competitors to catch up can be justified as protecting everyone.</p><p>Those arguments may have merit. They are also very useful to companies with huge bills and equally huge expectations hanging over them.</p><p>The financial pressure does not prove anyone invented the danger. It does give us a reason to examine whether the proposed response also protects a business model under pressure. I think it does.</p><p>We are being asked to accept restrictions that could shape who gets to build AI and who gets paid when we use it. Of course the economics belong in that discussion. Calling it safety does not put it beyond scrutiny.</p><h2>Regulation should follow the risk, not protect the club</h2><p>Of course government has a role. We cannot leave hospitals, power networks and other critical infrastructure to defend themselves against increasingly capable attacks with whatever is left in the IT budget. Serious incidents need to be reported, and companies working with models that can find unknown security holes need security that reflects what they are handling.</p><p>If you give an autonomous system access to someone else&#8217;s money, data or business, you should have to answer for how you manage that access. That seems fairly obvious to me.</p><p>Start with what Huang is saying. Apply the laws we already have. Look at what happened, who was responsible and whether they met their obligations. Where those laws leave a genuine gap, explain it and fill it. Show us the risk and how the proposed rule would address it.</p><p>What I struggle with is starting by giving the biggest companies exemptions from rules designed to stop competitors coordinating with one another. They are asking for considerable influence over the future of this technology. We should want considerably more than an assurance that it is all for our own good.</p><p>And they already have decisions they can make inside their own businesses.</p><p>If Dario thinks Anthropic is moving too fast, he can slow it down. Sam can do the same at OpenAI. OpenAI has already shown that it can pause development when it believes its security needs more work. Nobody had to stop the rest of the industry for that to happen.</p><p>They can put more money into securing their systems and limit what their agents can access while that work is being done. Bring in independent people to test the boundaries. Share what they find so others can fix the same weaknesses.</p><p>And price the product properly. If it costs more to run safely, charge for it. If taking responsibility for failures changes the economics, then those are the economics of the business you chose to build.</p><p>I want these companies to succeed. I also want them to take responsibility for what they put into the world. They have plenty of work they can get on with before asking for control over what everyone else is allowed to build.</p><h2>The keys were never the intelligence</h2><p>I am still an AI optimist. There are founders, scientists and people with good ideas who have never had access to the resources needed to do anything with them. Making intelligence cheap enough for those people to use is something worth building towards.</p><p>Nothing in these incidents has changed that for me. What they have changed is how much confidence I have in the companies connecting this intelligence to the world, and how willing I am to accept their proposed solution.</p><p>I take the warnings seriously. I also think commercial pressure is a major reason why slowing down together has become such an attractive idea. That is my judgement, based on the pressures on these businesses and what their proposals would do for them. The safety concerns do not make those benefits disappear.</p><p>These companies chose to race. They raised the money, committed to the spending and sold us on what their systems would be able to do. Much of that promise involved agents doing more work with less supervision, because that is something customers will pay for.</p><p>If they now believe they are moving faster than they can safely manage, they should slow down. I have no objection to that. I object to turning their commercial commitments and engineering failures into a claim over what everyone else is allowed to build.</p><p>We are not putting intelligence back in a box, and I do not want us to. I do want us to be much more careful about what we let it do. Being able to work out a course of action should not automatically give a system permission to carry it out.</p><p>We decide which accounts it can access, what it can spend and whether it can change or delete something. We build the machinery that lets it keep working while we are asleep. Those decisions deserve far more attention than they have been getting.</p><p>In 2024, I wrote that we had given AI the keys to our kingdom. Look at the incidents in this article and look at what let them happen.</p><p><strong>The keys are the permissions, credentials, tools and access we give it. We handed those over.</strong></p><p>Now the people who handed them over want control of the locksmith.</p><p>Do not give them that as well.</p><div><hr></div><h2><strong>Sources</strong></h2><ul><li><p><a href="https://openai.com/index/hugging-face-incident-and-the-road-ahead/">OpenAI: The Hugging Face incident and the road ahead</a></p></li><li><p><a href="https://insights.fusion-42.com/p/we-have-given-ai-the-keys-to-our-kingdom">Derek Watson: We have given AI the keys to our kingdom</a></p></li><li><p><a href="https://www.redwoodresearch.org/research/hugging-face-incident">Redwood Research and METR: Independent investigation of the OpenAI/Hugging Face incident</a></p></li><li><p><a href="https://www.anthropic.com/news/investigating-incidents-cybersecurity-evals">Anthropic: Investigating three incidents in our cybersecurity evaluations</a></p></li><li><p><a href="https://www.anthropic.com/news/alignment-assessment-cybersecurity-incidents">Anthropic: Alignment assessment of recent cybersecurity incidents</a></p></li><li><p><a href="https://www.anthropic.com/threat-intelligence-report-september-2026">Anthropic: Detecting and countering misuse of AI, September 2026</a></p></li><li><p><a href="https://stratechery.com/2026/an-interview-with-openai-president-greg-brockman-about-astra-and-alignment/">Stratechery: Interview with Greg Brockman about Astra and alignment</a></p></li><li><p><a href="https://www.anthropic.com/news/confidential-draft-s1-sec">Anthropic: Confidential draft S-1 submission</a></p></li><li><p><a href="https://openai.com/index/openai-submits-confidential-s-1/">OpenAI: Confidential draft S-1 submission</a></p></li><li><p><a href="https://blogs.nvidia.com/blog/nvidia-to-acquire-hugging-face/">Nvidia: Agreement to acquire Hugging Face</a></p></li><li><p><a href="https://www.reuters.com/business/retail-consumer/anthropic-tells-investors-it-will-be-profitable-second-straight-quarter-ft-2026-09-13/">Reuters: Anthropic&#8217;s reported adjusted profitability and gross-margin presentation</a></p></li><li><p><a href="https://www.reuters.com/business/palantir-nvidia-curb-ai-model-use-over-data-fears-information-reports-2026-09-14/">Reuters: Enterprise concerns over OpenAI and Anthropic data handling</a></p></li><li><p><a href="https://www.reuters.com/legal/litigation/openai-ipo-will-not-happen-2026-amid-ai-safety-fears-altman-says-2026-09-12/">Reuters: OpenAI&#8217;s IPO delay and the proposed industry pact</a></p></li><li><p><a href="https://www.reuters.com/world/us/ftc-chair-suspicious-calls-ai-antitrust-exemptions-2026-09-15/">Reuters: FTC chair questions calls for an AI antitrust waiver</a></p></li><li><p><a href="https://www.reuters.com/technology/openai-is-working-with-anthropic-google-ai-safety-bloomberg-news-reports-2026-09-15/">Reuters: OpenAI, Anthropic and Google are already coordinating on safety</a></p></li><li><p><a href="https://www.reuters.com/business/meta-zuckerberg-says-ai-labs-have-enough-incentive-build-safely-2026-09-16/">Reuters: Zuckerberg says individual labs can slow their own releases</a></p></li></ul><div><hr></div><h2>Go take a look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[WHY DON’T YOU WANT A DATA CENTRE IN YOUR TOWN WHEN YOU DON’T EVEN KNOW WHAT THE FUCK IT REALLY IS?]]></title><description><![CDATA[I kept seeing people protesting against AI data centres. Then I looked at what we are actually trying to build.]]></description><link>https://insights.fusion-42.com/p/why-dont-you-want-a-data-centre-in</link><guid isPermaLink="false">https://insights.fusion-42.com/p/why-dont-you-want-a-data-centre-in</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Wed, 09 Sep 2026 10:30:01 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!8G7G!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!8G7G!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!8G7G!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!8G7G!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png" width="1200" height="675" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2356976,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!8G7G!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 424w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 848w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 1272w, https://substackcdn.com/image/fetch/$s_!8G7G!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F98468ac4-0028-4126-a5f4-647bda5d2c28_1672x941.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2>Index</h2><ul><li><p><strong>Why people are protesting</strong></p></li><li><p><strong>We are trying to build the industry again</strong></p></li><li><p><strong>Inside an AI data centre</strong></p></li><li><p><strong>The supply chain underneath it</strong></p></li><li><p><strong>Everyone is building, and everyone is paying everyone</strong></p></li><li><p><strong>The global compute race</strong></p></li><li><p><strong>The real problem is power</strong></p></li><li><p><strong>The bill runs into trillions</strong></p></li><li><p><strong>Heat, cooling and water</strong></p></li><li><p><strong>Why locals are getting pissed off</strong></p></li><li><p><strong>Governments are changing the rules</strong></p></li><li><p><strong>Compute starts following energy</strong></p></li><li><p><strong>Elon wants to take it into space</strong></p><p></p></li></ul><h2>Why people are protesting</h2><p>My view on data centres is pretty simple. AI is going to need a lot more compute. That means a lot more data centres. Fine. Build them.</p><p>What I hadn&#8217;t done was add it all up.</p><p>Then the protests started catching my eye. In July there were <strong>142 demonstrations across 42 US states</strong>, and a Reuters/Ipsos poll found only 14% of Americans would welcome a data centre in their own community.</p><p>Texas was the one that really got me. Cheap energy, cheap land, light regulation. Exactly the sort of place you would expect to want these things. Yet Republican politicians there are now talking about restricting them, removing tax breaks and making developers pick up more of the costs. Only 30% of Texans in one recent poll supported having one nearby.</p><p>I knew why data centres were needed. I just hadn&#8217;t worked through the physical scale of a 1GW campus: the power connection, substations, transmission, cooling, water, construction, grid upgrades, permanent jobs once the builders leave and who pays for all of it.</p><p>So I did the sums.</p><p>At roughly the same time Elon Musk was talking about putting some of the compute in space. SpaceX eventually wants to <strong>launch 100GW of AI compute capacity on solar-powered satellites every year</strong>. At full scale that means thousands of launches and roughly <strong>one million tonnes of hardware going into orbit annually</strong>.</p><p>I&#8217;ll come back to that.</p><p></p><blockquote><p><strong>I still want the data centres built. I just don&#8217;t think &#8220;build them&#8221; is the end of the argument once you see the numbers.</strong></p></blockquote><p><br></p><h2>We are trying to build the industry again</h2><p>The first number that really changed the way I looked at this was <strong>103GW</strong>. That is roughly how much data-centre capacity exists globally today. JLL expects it to be around <strong>200GW by 2030</strong>.</p><p>So in the next few years we are trying to build almost as much data-centre capacity as the world has built in total up to now.</p><p>McKinsey gets to a slightly bigger number using a different methodology: around <strong>219GW by 2030</strong>, with <strong>156GW coming from AI</strong>. The exact forecast will move as chips get better, models become more efficient and projects get cancelled or delayed. But whether the eventual number is 200GW or 219GW does not change what is happening. We are adding an enormous industrial load in a very short period of time.</p><p>The electricity numbers make the scale easier to understand. Gartner has worldwide data-centre power demand rising from roughly <strong>104GW in 2025 to 132GW in 2026 and 290GW by 2030</strong>. Electricity consumption goes from <strong>447TWh in 2025 to 565TWh in 2026, 702TWh in 2027 and more than 1,200TWh by 2030</strong>. From 2027, AI-optimised servers are expected to use more electricity than conventional servers.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!o_JB!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!o_JB!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!o_JB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png" width="1448" height="1086" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1086,&quot;width&quot;:1448,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1681234,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!o_JB!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!o_JB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ce642b-4fbd-4d6b-a213-d221d59e3503_1448x1086.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>One thing worth clearing up because these GW numbers get thrown around rather casually. A <strong>1GW data-centre campus</strong> is basically a site designed around access to roughly one gigawatt of electrical power. So when you see forecasts approaching 300GW, think hundreds of very large industrial campuses all drawing power at the same time.</p><p>And this is already showing up in the real economy. US data-centre construction spending hit an annualised <strong>$75bn in July</strong>, up <strong>57% in a year</strong> and roughly <strong>717% since the start of 2021</strong>. Since the end of 2023 alone, the annualised run-rate has risen by about <strong>$51bn</strong>.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!FM1j!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!FM1j!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 424w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 848w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 1272w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!FM1j!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png" width="1200" height="662.6373626373627" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:804,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:297612,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!FM1j!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 424w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 848w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 1272w, https://substackcdn.com/image/fetch/$s_!FM1j!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa56f498a-39ef-4f02-aae6-c4cb8dba5dfb_3231x1785.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>That $75bn does not include the GPUs and servers. It is the buildings, site work and equipment going into them.</p><p>Housing and commercial construction have their own problems, so I would not pretend data centres caused the weakness elsewhere. But construction workers, transformers, electrical equipment, land and capital are finite.</p><p>A lot more of all of them is now being pulled towards AI.</p><p></p><p></p><h2>Inside an AI data centre</h2><p>I&#8217;m leaving the chips mostly out of this because Nvidia, TSMC, HBM and advanced packaging deserve their own piece. Assume the GPUs have arrived.</p><p>The rest of the building exists to keep them running.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!UHUv!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!UHUv!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!UHUv!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2628830,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!UHUv!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!UHUv!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1a8a5fbc-0f6e-4d8e-9fb1-08813c4218a6_1536x1024.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>You need racks, servers, storage, networking and fibre. Then comes the electrical side: substations bringing power onto the site, transformers changing voltage, switchgear controlling and protecting the system, UPS equipment and batteries dealing with short interruptions, backup generators for longer ones, plus busbars and a lot of heavy cabling moving electricity around the building.</p><p>Cooling is becoming a much bigger part of the design as AI racks get denser. Air cooling only gets you so far. Higher-density systems increasingly use cold plates sitting directly against the processors, with pumps, coolant distribution units, heat exchangers and chillers moving the heat out of the building. Depending on the system and location, cooling towers and water may come into it as well.</p><p>Then there is everything outside the server room: land, roads, transmission connections, concrete, steel, copper, construction crews, electricians, engineers and all the other fairly unglamorous bits needed before anyone can run a model.</p><p>This is how <strong><a href="https://www.se.com/">Schneider Electric</a>, <a href="https://www.eaton.com/">Eaton,</a> <a href="https://www.abb.com/">ABB</a>, <a href="https://www.siemens-energy.com/">Siemens Energy,</a> <a href="https://www.gevernova.com/">GE Vernova</a>, <a href="https://www.mhi.com/">Mitsubishi Heavy Industries</a>, <a href="https://www.vertiv.com/">Vertiv</a>, <a href="https://www.deltaww.com/">Delta Electronics</a>, <a href="https://www.caterpillar.com/">Caterpillar</a> and <a href="https://www.cummins.com/">Cummins</a></strong> end up sitting underneath the AI boom.<br>A surprising amount of artificial intelligence depends on some very old-fashioned engineering.</p><p></p><h2>The supply chain underneath it</h2><p>Once you look at everything inside an AI data centre, another problem appears. Somebody has to make all of it, and the supply chain stretches far beyond Silicon Valley.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!APHY!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!APHY!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!APHY!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!APHY!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!APHY!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!APHY!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png" width="1200" height="900" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/a66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1086,&quot;width&quot;:1448,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2526859,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!APHY!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!APHY!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!APHY!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!APHY!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fa66ce950-5784-447b-9bb7-52c3ed2b5271_1448x1086.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>Start with the chips. The most advanced AI processors still depend heavily on <strong>TSMC in Taiwan</strong>, while the high-bandwidth memory beside them comes mainly from <strong>SK Hynix, Samsung and Micron</strong>. Then there is advanced packaging such as CoWoS, substrates, PCBs and the networking hardware needed to connect thousands of accelerators. A shortage in any one of those can slow the whole system down.</p><p>Move outside the rack and it gets much more industrial. A large data centre needs a lot of copper in transformers, cables, busbars, switchgear, cooling systems and the grid connection itself. Much of that copper is mined in places such as Chile, Peru and the Democratic Republic of Congo, while China plays a huge role in refining it.</p><p>Transformers are an even more immediate bottleneck. They are large, often built to order and can have lead times well beyond a year. Switchgear, circuit breakers and other high-voltage equipment are tight too. You can have the land, the building and billions of dollars of GPUs sitting inside it and still be waiting for the electrical kit needed to turn the bloody thing on.</p><p>Power generation has the same problem. AI companies want electricity faster than grids can add it, which has pushed gas turbines back into demand. <strong>GE Vernova, Siemens Energy and Mitsubishi Heavy Industries</strong> cannot simply double production overnight. These are huge machines built in specialised factories with their own supply chains for steel, castings, blades, generators and control systems.</p><p>Then add batteries. Modern data centres increasingly use large lithium-ion UPS systems to bridge interruptions before generators or other backup systems take over. Much of the LFP battery supply chain sits in China, from cathode materials through to finished cells.</p><p>Cooling adds another layer. Higher-density racks need cold plates, pumps, valves, coolant distribution units, heat exchangers, chillers, specialist fluids and kilometres of pipework. None of this looks particularly exciting next to an Nvidia GPU, but a failed pump or delayed heat exchanger can leave the GPU doing absolutely nothing.</p><p>And underneath all of that sit concrete, structural steel, electrical steel, fibre, optical equipment, connectors, generators, cables and a lot of skilled labour.</p><p>We talk about AI as software. Building the infrastructure behind it looks increasingly like heavy industry, with dependencies on mines, smelters, semiconductor fabs, transformer factories, turbine plants and electrical equipment manufacturers spread across the world.</p><p>The bottleneck will not always be the cleverest component. Quite often it will be the boring thing everyone suddenly needs at the same time.</p><p><strong>You can have the GPUs and still not be able to turn them on.</strong></p><p></p><h2>Everyone is building, and everyone is paying everyone</h2><p>This was the point where I started losing track.</p><p>The hyperscalers are already spending at extraordinary levels. AWS, Microsoft, Google, Meta and Oracle are all building huge amounts of data-centre infrastructure, while the AI companies themselves have increasingly started securing dedicated capacity rather than simply renting it.</p><p>OpenAI&#8217;s Stargate programme is a good example. It originally targeted <strong>10GW of US AI infrastructure by 2029</strong>, and OpenAI said in April that it had already passed that figure. It is not pouring all the concrete itself. The capacity is being assembled through a network of infrastructure and cloud partners.</p><p>Anthropic is doing the same from both directions. It announced <strong>$50bn of custom-built US computing infrastructure with Fluidstack</strong>, with sites in Texas and New York, while also signing very large compute agreements with AWS, Google and others. Tesla has built its own dedicated training infrastructure at Giga Texas, where Cortex 1 has more than <strong>90MW</strong> and Cortex 2 more than <strong>115MW</strong> supporting autonomy and Optimus. SpaceXAI has Colossus in Memphis and is separately building a <strong>1.2GW permanent power plant</strong> to support its terrestrial compute expansion. That is power generation for compute on Earth, not part of the orbital system.</p><p>Then you have the new infrastructure layer sitting underneath all of this. CoreWeave, Fluidstack, Lambda, IREN, Nscale and others are raising billions to secure land, power and hardware, then selling the resulting capacity back to AI companies. Nvidia sits in the middle selling the GPUs while also investing across much of the same ecosystem.</p><p>The financing has started to get equally strange. Oracle ended FY2026 with <strong>$638bn of remaining performance obligations</strong>, and some customers are now prepaying for GPUs or even supplying hardware themselves so Oracle does not have to finance all of it. Meanwhile the hyperscalers&#8217; own capex numbers have become absurd: Amazon expects around <strong>$200bn in 2026</strong>, Meta <strong>$130bn&#8211;$145bn</strong>, with Alphabet and Microsoft spending at similarly enormous levels.</p><p>At that point the headline numbers start to lose meaning, because the same underlying capacity can show up again and again in different announcements. The AI company announces a compute contract. The cloud provider reports the backlog. Nvidia reports the hardware demand. The infrastructure company raises money against the contract. A government may then present the same project as inward investment.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!baCM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!baCM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!baCM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!baCM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!baCM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!baCM!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png" width="1200" height="900" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1086,&quot;width&quot;:1448,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2512554,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!baCM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!baCM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!baCM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!baCM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc634ae58-9e1a-4db2-9d56-1bc0fe58aa2a_1448x1086.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3><strong>Everyone is paying everyone.</strong></h3><p>That is why I stopped trying to add the press releases together.</p><p>The physical build is much harder to bullshit. JLL&#8217;s model implies roughly <strong>another 100GW of data-centre capacity by 2030</strong>, with required investment of up to <strong>$3tn</strong>. McKinsey, using a broader definition that includes more of the infrastructure and hardware stack, gets to roughly <strong>$6.7tn</strong>, including about <strong>$5.2tn for AI</strong>.</p><p>And the spending does not stop once the buildings are finished. The concrete may last decades. The GPUs, servers and networking equipment do not. We will be replacing the expensive kit inside these buildings while still constructing the next generation of them.</p><p>There is also clearly a wall of corporate FOMO in the background. Nobody running a major technology company wants to wake up in three years and discover that a rival locked up the land, power, chips and grid connections while they were waiting for the economics to become obvious.</p><p>That creates a very powerful incentive to build first and work some of the details out later.</p><h2>The global compute race</h2><p>Individual projects are now enormous. Meta&#8217;s Hyperion campus in Louisiana is being expanded towards <strong>5GW of compute capacity</strong>, with more than <strong>$50bn</strong> being invested. Meta expects around <strong>7,500 jobs at peak construction</strong> and roughly <strong>1,000 permanent roles</strong> once the site is operating.</p><p>Five gigawatts in one place gives you some idea how far this has moved from the old idea of a data centre.</p><p>Governments are now entering the same race. Britain says it needs <strong>at least 6GW of AI-capable data-centre capacity by 2030</strong>, roughly three times what it has today, with several large AI Growth Zones planned and at least one site capable of exceeding <strong>1GW</strong>. Saudi Arabia&#8217;s HUMAIN programme already has production AI compute running, with its AMD-Cisco venture beginning another <strong>250MW of deployment from 2027</strong> and targeting <strong>1GW by 2030</strong>. Stargate UAE is planned as a <strong>1GW AI cluster in Abu Dhabi</strong>, with the first <strong>200MW</strong> expected online in 2026.</p><p>China is operating at a much larger scale. It had roughly <strong>32GW of installed data-centre capacity at the end of 2025</strong>, with current projects potentially taking that above <strong>60GW by 2030</strong>. India is starting from a smaller base but moving quickly, with its Economic Survey citing an estimate of around <strong>1.4GW in 2025 rising towards 8GW by 2030</strong>. Europe is building sovereign AI capacity through its own programmes, while Japan is expanding too, although I still haven&#8217;t found a clean national GW target I trust enough to use.</p><p>What started as a hyperscaler capex boom is turning into strategic national infrastructure.</p><p>That makes sense. If AI becomes as important to economies as governments now expect, countries are hardly going to be comfortable relying entirely on somebody else&#8217;s compute capacity.</p><p><strong>But every one of these national ambitions runs into the same physical constraint: electricity.</strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!5Lz5!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!5Lz5!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 424w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 848w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 1272w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!5Lz5!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png" width="1200" height="901.4512785072563" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1087,&quot;width&quot;:1447,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1809857,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!5Lz5!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 424w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 848w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 1272w, https://substackcdn.com/image/fetch/$s_!5Lz5!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F427a23aa-174c-4ce8-b23c-97065d7e9d71_1447x1087.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2>The real problem is power</h2><p>Once an individual campus wants 500MW, 1GW or even 5GW, electricity stops being a utility bill and starts determining whether the project can exist at all.</p><p>That is becoming a serious problem because many grids simply cannot provide another gigawatt on the timetable AI companies want. US utilities have already received <strong>more than 700GW of large-load requests</strong>, more than ten times estimated current US data-centre demand. Texas has started auditing applications because regulators believe a substantial share is duplicated, speculative or attached to projects that may never be financed. The industry has started calling this <strong>ghost demand</strong>.<br></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!OZ-a!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!OZ-a!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!OZ-a!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png" width="1448" height="1086" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/cbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1086,&quot;width&quot;:1448,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1679625,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!OZ-a!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 424w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 848w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 1272w, https://substackcdn.com/image/fetch/$s_!OZ-a!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcbf2fcf1-2a92-4d2a-a461-f2b7b2276084_1448x1086.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>That matters whenever somebody announces another 2GW campus. Asking the grid for 2GW is easy. Financing the project, building it and actually energising 2GW are very different things. Even after stripping out the speculative requests, however, the underlying problem remains: AI companies need enormous amounts of reliable power, and they need it quickly.</p><p>The timing is awkward. New nuclear plants and SMRs feature heavily in AI infrastructure plans, but a reactor that arrives well into the 2030s does very little for a data centre that needs 500MW in 2028. In the near term, the options are more practical: gas generation, existing nuclear plants staying open longer or being restarted, whatever grid capacity is available, and behind-the-meter generation where it is not.</p><p>That is why companies such as <strong>GE Vernova, Siemens Energy and Mitsubishi Heavy Industries</strong> have suddenly become part of the AI story. They manufacture the gas turbines and other equipment capable of delivering hundreds of megawatts of dispatchable power. Developers are increasingly looking at building generation alongside the data centre rather than spending years waiting for the grid. Google has moved upstream as well, buying energy-infrastructure developer Intersect and signing large geothermal agreements, while other hyperscalers are pursuing nuclear and dedicated generation deals of their own.</p><p>The power station is only half the problem. Electricity still has to get from the generator to the GPUs, and that means substations, switchgear, circuit breakers and transformers. Large transformers in particular have long manufacturing cycles, so a finished data centre can have billions of dollars of computing hardware sitting inside it and still be waiting for the equipment needed to switch it on.</p><p>There is something rather wonderful about that.</p><p><strong>Some of the most advanced technology humans have ever built can still be held up by a giant lump of copper and steel based on engineering that has existed for more than a century.</strong></p><h2>The bill runs into trillions</h2><p>Putting a clean price on all of this is harder than it looks because people count different things. Land, buildings, grid connections, dedicated generation, cooling, GPUs, networking and financing can all end up under the heading of AI infrastructure.</p><p>JLL estimates that adding roughly another <strong>100GW of global data-centre capacity by 2030</strong> could require up to <strong>$3tn</strong>, including around <strong>$1.2tn of data-centre real estate</strong>. McKinsey counts more of the technology sitting inside and around the buildings and gets to roughly <strong>$6.7tn</strong>, with about <strong>$5.2tn linked to AI infrastructure</strong>.</p><p>Either way, we are talking about trillions of dollars being spent in a few years.</p><p>And unlike a railway, a lot of the expensive kit does not last very long. The building may still be standing in thirty years, but the GPUs, servers and networking equipment inside it will have been replaced several times by then.</p><p>That changes the economics quite a bit. The industry will still be building the first wave of new capacity while already replacing some of the most expensive equipment installed at the start of it.</p><p>So the capex does not simply peak when the buildings are finished.</p><p><strong>A large part of this infrastructure has to be built, upgraded and rebuilt at the same time.</strong></p><h2>Heat, cooling and water</h2><p>All that electricity eventually becomes heat, and getting rid of it is becoming a much bigger engineering problem as AI racks get denser.</p><p>Traditional data centres could rely heavily on air cooling. At higher rack densities, simply pushing more cold air through the room stops being enough, which is why liquid cooling has become such an important part of the build-out. Cold plates sit against the processors, pumps move coolant around the rack, heat exchangers pull the heat away and chillers or cooling towers deal with it further downstream.</p><p>That is how I ended up reading about pumps and pipework while supposedly researching artificial intelligence.</p><p>Water is more complicated because there is no honest universal number. Some cooling systems consume a lot, others use very little, and climate makes a big difference. But the issue is clearly becoming material enough for governments to regulate it. Minnesota, for example, has introduced special permitting requirements for projects proposing more than <strong>100 million gallons of additional consumptive water use per year</strong>, with regulators required to consider conservation, recycling, reclaimed water and closed-loop systems.</p><p>There is also water used upstream by some forms of electricity generation, so the total impact can extend well beyond the cooling system itself. For a town already worried about water availability, a new 500MW or 1GW industrial customer is obviously going to attract scrutiny.</p><p>And water is only part of what neighbours actually experience. Cooling equipment runs continuously, and fans, chillers and other machinery can produce a persistent low-frequency hum. Hyperscale sites can also have large numbers of diesel backup generators behind the fence. They do not run continuously, but they still need testing and create noise and emissions when they do.</p><p>These sound like minor issues next to trillion-dollar AI forecasts.</p><p><strong>They feel rather less minor when you live next door.</strong></p><h2>Why locals are getting pissed off</h2><p>A data-centre announcement sounds fantastic when it first arrives. Billions of dollars of investment, thousands of construction jobs, new tax revenue and a place in the AI economy. Those benefits are real. We need the infrastructure, domestic compute capacity has strategic value, and the economic activity eventually built on top of AI could be worth far more than the facilities themselves.</p><p>The argument changes when residents start looking at what the project requires from the place hosting it. A large campus can consume hundreds of megawatts around the clock, require new substations and transmission lines, draw significant amounts of water and run cooling equipment continuously. Backup generation has to sit onsite as well, bringing its own noise and emissions.</p><p>Then there is the jobs question. Meta&#8217;s Louisiana campus is a useful example. More than <strong>$50bn</strong> of investment and around <strong>7,500 jobs at peak construction</strong> eventually becomes roughly <strong>1,000 permanent roles</strong>. A thousand jobs can still matter enormously to a local economy, but the calculation looks different when they sit alongside a <strong>5GW industrial load</strong> and tens of billions of dollars of infrastructure.</p><p>That is why communities start asking harder questions about tax incentives, electricity prices, water, noise and who pays for the grid upgrades. <strong>The political fight starts when people begin to wonder whether the deal works as well for the place hosting the data centre as it does for the company using it.</strong></p><h2>Governments are changing the rules</h2><p>Governments are starting to respond because data centres are becoming too large to treat like ordinary commercial developments.</p><p>South Dakota has gone as far as passing something literally called the <strong>Data Center Bill of Rights for Citizens</strong>. It applies to facilities with peak demand above 10MW and follows a fairly sensible principle: if a data centre creates extra costs for the electricity system, those costs should sit with the data centre rather than being spread across everyone else. Developers also have to disclose expected water use, while local authorities retain the right to restrict or prohibit projects.</p><p>Ireland has a different problem because data centres already consumed <strong>22% of the country&#8217;s electricity in 2024</strong>, with that share potentially reaching <strong>31% by 2034</strong>. New projects seeking grid connections now have to provide enough generation or storage to match their maximum requested demand. Germany has focused on efficiency, requiring new data centres from July 2026 to achieve a <strong>PUE of 1.2 or better</strong> and reuse an increasing share of the heat they produce.</p><p>The US is heading in the same direction. Virginia, already home to one of the world&#8217;s largest concentrations of data centres, decided in July that some transmission costs should be charged directly to the customers responsible for creating them. PJM, which runs much of the eastern US electricity market, is also developing rules that favour large new loads which bring generation with them. Projects without enough power behind them can face restrictions when the grid gets tight.</p><p>That is broadly where I think the rules should go.</p><p>Build the data centres, but make the economics honest. If a project needs a new substation, transmission line or power plant, the people building it should pay the costs created by it. If it needs huge amounts of water, tell the community how much and where it is coming from. If the local grid cannot support another gigawatt, build more generation or put the facility somewhere that can.</p><p><strong>That is what serious infrastructure planning looks like.</strong></p><p></p><h2>Power availability is starting to redraw the geography of AI.</h2><p>Software companies traditionally clustered around engineers, universities, customers and capital. Large AI training clusters have different priorities. Give them enough fibre and reliable electricity and much of the workload can move. Training is especially mobile because the GPUs do not care whether they are five miles or five hundred miles from the people building the model.</p><p>Inference is a bit different. Batch workloads can also move towards cheap power, while applications where latency matters benefit from being closer to the people or machines using them. So we are likely to end up with very large training clusters in places with abundant electricity, alongside more distributed inference capacity closer to users.</p><p>That suddenly makes some countries much more interesting. France has nuclear. The Nordics and Canada have hydro. China has deliberately pushed heavy compute towards western regions where land and electricity are easier to find. Saudi Arabia and the UAE have energy, capital and governments capable of making large infrastructure decisions quickly.</p><p>The Gulf comes with its own engineering problems. Saudi Arabia is bloody hot, so cooling is harder and water matters. Transmission and fibre still matter as well. But once compute starts behaving like heavy industry, the strategy makes sense. Saudi&#8217;s HUMAIN programme is targeting <strong>1GW by 2030</strong> through its AMD-Cisco venture, while Stargate UAE is planned as another <strong>1GW cluster</strong> in Abu Dhabi.</p><p>China is doing the same thing at a much larger scale. Its data-centre capacity could move from roughly <strong>32GW in 2025 to more than 60GW by 2030</strong>, with more of the heavy compute going to regions where the power and land already exist.</p><p><strong>AI is beginning to follow the geography of heavy industry.</strong></p><p>When a training campus needs 500MW or a gigawatt, access to electricity matters rather more than having the software engineers nearby.</p><h2><strong>Elon wants to take it into space</strong></h2><p>This brings me back to Elon, because after looking at all the constraints on Earth, SpaceX&#8217;s <strong>Starmind</strong> project starts to look a lot less mad.</p><p>AI needs enormous amounts of electricity, and finding hundreds of megawatts of reliable new power on Earth is getting harder. Grid connections can take years. New transmission can take longer. Land and water become political very quickly. SpaceX&#8217;s idea is to move some of the compute closer to an energy source that is already there: the Sun.</p><p>Its first design, <strong>AI1</strong>, is a modular AI-compute satellite using solar technology derived from Starlink. SpaceX says each satellite would carry a <strong>210kW solar array</strong> and operate in sun-synchronous orbit, giving it long periods of solar exposure without weather or atmospheric losses. The compute modules are chip-vendor agnostic, while high-speed optical links would connect the satellites to one another and feed results back through Starlink.</p><p>This has moved beyond a slide deck. SpaceX is building a <strong>Gigasat Factory in Bastrop</strong> that it says could start producing and deploying thousands of AI satellites from late 2027. <strong>Terafab</strong>, being developed with Tesla, is intended to supply the advanced AI chips needed further up the stack.</p><p>SpaceX identifies three big constraints of its own: getting enough mass into orbit, generating enough power and producing enough AI chips. Starship is supposed to deal with the first, orbital solar with the second and Terafab with the third.</p><p>Then the numbers get properly ridiculous.</p><p>SpaceX&#8217;s June 2026 prospectus says the long-term ambition is to <strong>launch 100GW of AI compute capacity every year</strong>. At that scale it expects thousands of launches and roughly <strong>one million tonnes of hardware delivered into orbit annually</strong>.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!vwPS!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!vwPS!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!vwPS!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/c183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2528147,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/214720871?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!vwPS!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!vwPS!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fc183838b-1a98-47f8-9b92-af88324223b6_1536x1024.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>The idea becomes easier to understand once you compare it with what is happening on Earth. A large terrestrial data centre can spend years waiting for grid capacity, substations, transmission lines and planning approval before anyone switches on the GPUs. SpaceX is effectively asking whether it eventually becomes easier to take some of the compute to the energy rather than keep dragging ever larger amounts of energy towards the compute.</p><p>Cooling is part of the case as well. Terrestrial data centres spend a lot of energy running pumps, chillers, fans and cooling towers. SpaceX says orbital compute can remove much of that overhead because heat can be radiated directly into space.</p><p>The heat still has to go somewhere. In a vacuum that means rejecting it through radiators, and at hundreds of kilowatts per satellite thermal management becomes a serious engineering problem of its own.</p><p>The network is another one. These satellites are moving at orbital speed, so laser links have to stay locked while traffic is continually handed from one satellite to another. <strong>Link continuity and handover</strong> matter enormously. A distributed orbital AI cluster cannot afford links dropping whenever one satellite disappears over the horizon.</p><p>Then add radiation, collision avoidance, replacement cycles, servicing, orbital debris and the economics of launching enough hardware to make the whole thing worthwhile.</p><p>SpaceX has spent the last twenty years turning problems like these into industrial systems, which is why I find Starmind interesting. It is trying to control most of the stack itself: <strong>Terafab makes the chips, the Gigasat Factory builds the satellites, Starship gets them into orbit, Starmind runs the compute and Starlink connects it back to Earth.</strong></p><p>Whether that works economically at anything approaching 100GW a year is still a huge question. The industrial logic, though, is easy enough to see.</p><p>If power becomes one of the main constraints on AI, putting some compute next to a near-continuous source of solar energy starts to make sense.</p><p>That brings me back to where I started.</p><p>I believe in what this technology can do, and I think we are going to need vastly more compute than we have today. Some of the rush is rational. Some of it is corporate FOMO, with trillion-dollar companies racing to secure land, power, chips and grid capacity before their competitors do.</p><p><strong>We need to build the bloody thing properly.</strong></p><p>Put data centres where the power and infrastructure can support them. Build generation alongside them where necessary. Be honest about water, noise and emissions. And when a project creates the need for a new substation or transmission line, the company building it should pay rather than quietly handing the bill to everyone living nearby.</p><p>I&#8217;m all for building the compute.</p><p><strong>We just shouldn&#8217;t fuck over the people living next door to do it.</strong></p><p>And if Earth really does become the constraint, perhaps some of that compute ends up somewhere else entirely.</p><p></p><div><hr></div><h2>Go take a and look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[WHO THE HELL IS GOING TO BUILD ALL THE BITS?]]></title><description><![CDATA[Millions of robots will need hundreds of millions of motors, gears, sensors and actuators. Somewhere in that supply chain, something is going to break.]]></description><link>https://insights.fusion-42.com/p/who-the-hell-is-going-to-build-all-1d1</link><guid isPermaLink="false">https://insights.fusion-42.com/p/who-the-hell-is-going-to-build-all-1d1</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Wed, 02 Sep 2026 18:23:18 GMT</pubDate><enclosure url="https://api.substack.com/feed/podcast/213901078/91b3bcb1b3ed9b1e7706d2554553659a.mp3" length="0" type="audio/mpeg"/><content:encoded><![CDATA[<p><span>Everyone is watching Optimus, Figure and Unitree.<br><br>I went looking at what actually goes inside the bloody things.<br><br>Motors. Bearings. Roller screws. Precision gears. Encoders. Force sensors. Magnets. Actuators.<br><br>Then I looked at who makes them, how difficult they are to manufacture and what happens if we really try to build millions of robots a year.<br><br>That turne&#8230;</span></p>
      <p>
          <a href="https://insights.fusion-42.com/p/who-the-hell-is-going-to-build-all-1d1">
              Read more
          </a>
      </p>
   ]]></content:encoded></item><item><title><![CDATA[WHO THE HELL IS GOING TO BUILD ALL THE BITS?]]></title><description><![CDATA[Millions of robots will need hundreds of millions of motors, gears, sensors and actuators. Somewhere in that supply chain, something is going to break.]]></description><link>https://insights.fusion-42.com/p/who-the-hell-is-going-to-build-all</link><guid isPermaLink="false">https://insights.fusion-42.com/p/who-the-hell-is-going-to-build-all</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Wed, 02 Sep 2026 07:17:15 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!nM-l!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p><strong>I have recently been looking more closely at robotics, and something has started bothering me</strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!nM-l!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!nM-l!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!nM-l!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2465232,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/213815872?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!nM-l!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!nM-l!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3b6f4c00-dfc0-431e-9d99-668a4874e392_1536x1024.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>Everyone is talking about the robots themselves. Optimus, <a href="https://www.figure.ai/">Figure</a>, <a href="https://www.unitree.com/">Unitree</a>, <a href="https://bostondynamics.com/">Boston Dynamics</a>, <a href="https://apptronik.com/">Apptronik</a>, <a href="https://www.agilityrobotics.com/">Agility</a>, <a href="https://www.ubtrobot.com/en/">UBTech</a> and an increasingly long list of companies now throwing serious money at Physical AI.</p><p>The forecasts are all over the place, but you don&#8217;t need to believe the most ridiculous ones for the numbers to get very big.</p><p>And that got me thinking about something much more basic.</p><h3>But <strong>who the hell is going to make all the bits?</strong></h3><p>Because this isn&#8217;t software. You can&#8217;t spin up another million robots in AWS. They have to be built, and there is a surprising amount of difficult engineering packed inside each one.</p><p>A humanoid needs <strong>motors, bearings, gears, screws, encoders, sensors, magnets, controllers, batteries and actuators</strong>. It needs them everywhere. Shoulders, elbows, wrists, hips, knees and ankles all have to move accurately, carry weight and survive being used thousands of times a day.</p><p>The hands are another problem altogether. Ten fingers, each trying to reproduce movements and touch that we barely think about when we pick up a glass or turn a door handle.</p><p>Once you add cameras, power electronics, cooling, cables and compute, a humanoid starts to look like a small factory&#8217;s worth of precision engineering squeezed into one machine.</p><p>And the numbers get big very quickly. If the industry eventually builds ten million humanoids a year, it won&#8217;t need ten million motors, bearings or sensors. It could need <strong>hundreds of millions of them</strong>.</p><p>We spend a lot of time talking about whether the robots will be good enough.</p><p>I&#8217;m beginning to wonder whether we&#8217;ll be able to make enough of the bloody things.</p><h2>The return of boring engineering</h2><p>For the past few years, AI infrastructure has mostly meant <strong>GPUs, data centres, networking and memory</strong>. Then everyone realised the bloody things needed huge amounts of electricity. So we started worrying about grids, transformers, gas turbines, nuclear power and copper.</p><p>Robotics adds a different problem. You actually have to make things.</p><p>AI can design a better robot, write the software and teach it new tasks. But if Tesla, Figure and twenty Chinese manufacturers suddenly need another million precision bearings, motors or gearboxes, somebody has to physically make them.</p><p>And there are plenty of places where supply could get tight:</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Hkez!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Hkez!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 424w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 848w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 1272w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!Hkez!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png" width="1200" height="1200" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:1254,&quot;width&quot;:1254,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2196008,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/213815872?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Hkez!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 424w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 848w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 1272w, https://substackcdn.com/image/fetch/$s_!Hkez!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2e2bfdb-d068-4a1f-9743-b8a8f67e81a5_1254x1254.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p>You don&#8217;t need shortages everywhere for this to become a problem. <strong>One component that can&#8217;t keep up can stop the whole production line.</strong></p><p>The car industry found that out during the semiconductor shortage. Factories capable of producing thousands of cars sat idle because they couldn&#8217;t get enough relatively cheap chips.</p><p>Robotics will have the same weakness. A $20,000 robot is no use if you can&#8217;t get the $50 part needed to finish it.</p><h2>Which led me to Japan &#127471;&#127477; and Germany &#127465;&#127466;</h2><p>This was the bit I hadn&#8217;t really thought about until I started looking at who actually makes this stuff.</p><p>Japan and Germany aren&#8217;t leading the AI hype cycle, but they&#8217;ve spent decades getting very good at the engineering robots need. <strong>Bearings, motors, gears, machine tools, linear guides, encoders, servo drives and precision screws.</strong></p><p>None of it sounds particularly exciting. Most of these companies will never have a robot dancing on stage at a tech conference.</p><p>But without their parts, the robot doesn&#8217;t dance at all.</p><p>Japan has <strong><a href="https://www.hds.co.jp/english/">Harmonic Drive Systems</a></strong> and <strong><a href="https://www.nabtesco.com/en/">Nabtesco</a></strong> making precision reducers. <strong><a href="https://www.thk.com/">THK</a></strong><a href="https://www.thk.com/"> </a>makes linear-motion systems and screws. <strong><a href="https://www.nsk.com/">NSK</a></strong> makes bearings and screws. <strong><a href="https://www.nidec.com/en/">Nidec</a></strong> makes motors. <strong><a href="https://www.yaskawa.com/">Yaskawa</a></strong> and <strong><a href="https://www.fanuc.com/">Fanuc</a></strong> have been building servo systems and industrial robots since long before anyone thought putting &#8220;Physical AI&#8221; in a PowerPoint would increase the valuation.</p><p>Germany has <strong><a href="https://www.schaeffler.com/en/">Schaeffler</a></strong>, <strong><a href="https://www.boschrexroth.com/">Bosch Rexroth</a></strong>, <strong><a href="https://www.siemens.com/">Siemens</a></strong> and <strong><a href="https://www.infineon.com/">Infineon</a></strong>, alongside hundreds of specialist Mittelstand engineering companies most investors have probably never heard of.</p><p>But once you start looking properly, this is clearly not just a German and Japanese story.</p><p>Switzerland &#127464;&#127469;has <strong><a href="https://www.maxongroup.com/">maxon</a></strong> and an extraordinary precision-engineering ecosystem.</p><p>Britain&#127468;&#127463; has <strong><a href="https://www.renishaw.com/">Renishaw</a></strong>, which makes precision measurement and encoder systems.</p><p>Taiwan &#127481;&#127484; has <strong><a href="https://www.hiwin.tw/">Hiwin</a></strong>, one of the world&#8217;s important manufacturers of linear guides and ball screws.</p><p>South Korea &#127472;&#127479;  has enormous capability across batteries, electronics and motors.</p><p>America &#127482;&#127480; has companies such as <strong><a href="https://www.moog.com/">Moog</a>, <a href="https://www.allegromicro.com/">Allegro MicroSystems</a> and <a href="https://www.monolithicpower.com/">Monolithic Power Systems</a></strong><a href="https://www.monolithicpower.com/"> </a>sitting in different parts of the motion and control stack.</p><p>Australia &#127462;&#127482;has <strong><a href="https://lynasrareearths.com/">Lynas Rare Earths</a></strong>, which doesn&#8217;t make a single robot but suddenly becomes relevant because high-performance motors need permanent magnets.</p><p>And then, inevitably, there is China.</p><h3>China &#127464;&#127475; could change the economics of the whole thing</h3><div class="native-video-embed" data-component-name="VideoPlaceholder" data-attrs="{&quot;mediaUploadId&quot;:&quot;bca09299-e558-4e0e-8fe2-28c743f65a5c&quot;,&quot;duration&quot;:null}"></div><p></p><p>China isn&#8217;t waiting to see how big humanoid robotics becomes. It&#8217;s building the robots and pretty much everything that goes inside them: <strong>motors, magnets, gearboxes, bearings, actuators, sensors, batteries and electronics.</strong></p><p>It already dominates the rare-earth magnet supply chain and its precision component manufacturers are getting better quickly.</p><p>Take Harmonic Drive. It may make a better precision reducer today, but what happens if a Chinese competitor can make one that&#8217;s 95% as good for half the price, and make a million of them?</p><p>We&#8217;ve seen what China can do to the economics of an industry with solar panels, batteries and now electric vehicles. I wouldn&#8217;t bet on robotics being any different.</p><p>At low volumes, making the best component matters. At millions of robots a year, <strong>making it cheaply, reliably and in enormous numbers may matter more.</strong></p><p></p><h2>The part I keep coming back to is <strong>actuation</strong>.</h2><p>Actuators are the robot&#8217;s muscles. A single joint can contain a motor, magnets, bearings, a gearbox or roller screw, an encoder, force sensors and the electronics controlling it. Now spread those across shoulders, elbows, wrists, hips, knees, ankles and fingers.</p><p>One component I hadn&#8217;t given much thought to before is the <strong>planetary roller screw</strong>. It turns the spinning movement of a motor into powerful, precise linear movement, which is useful when you need a lot of force from a small joint.</p><p>They&#8217;re also bloody difficult to make.</p><p>The threads have to be extremely accurate and survive repeated heavy loads. Materials, grinding and heat treatment all matter. You can&#8217;t suddenly double production because the orders have arrived.</p><p>Chinese manufacturers are already moving from small pilot lines towards much larger production capacity.</p><p><strong>That&#8217;s the sort of thing I&#8217;m looking for: an obscure component that suddenly has to be made in completely different volumes.</strong></p><p>Precision gears have much the same problem. A robot joint needs to move accurately and without much play. Japan&#8217;s <strong>Harmonic Drive Systems</strong> and <strong>Nabtesco</strong> have spent decades getting very good at this, while China&#8217;s <strong>Leader Harmonic</strong> is coming after the same market.</p><p>The complication is that nobody yet knows which design will win. Some robots use harmonic gears, others use planetary systems or linear actuators. We may end up with complete joints that combine the motor, gearing, sensors and electronics into one replaceable unit.</p><p>So some of the parts we think will be important today may disappear, while others become enormous markets.</p><h2>Then there are the hands</h2><p>Hands are another problem altogether.</p><p>Pick up a wine glass and think about what your fingers are doing. You&#8217;re judging how hard to squeeze it, whether it&#8217;s slipping, how heavy it is and whether you&#8217;ve actually got hold of it. Most of that happens without you thinking about it.</p><p>A robot has to sense all of this.</p><p><strong>Vision gets the hand to the glass. Touch stops it smashing the bloody thing.</strong></p><p>That is creating a new market for tactile sensors, force sensors, electronic skin and increasingly complicated robotic hands.</p><p>And there are already some fascinating private companies working on exactly this:</p><ul><li><p><a href="https://xelarobotics.com/">XELA Robotics</a> in Japan is developing tactile sensors that measure pressure and shear forces.</p></li><li><p><a href="https://www.tacterion.com">tacterion</a> in Germany is developing flexible electronic skin.</p></li><li><p><a href="https://www.gelsight.com">GelSight</a> in America uses optical sensing to allow machines effectively to see what they&#8217;re touching.</p></li><li><p><a href="https://contactile.com">Contactile</a> in Australia is working on force, grip and slip sensing.</p></li><li><p><a href="https://www.botasys.com">Bota Systems</a> in Switzerland makes compact six-axis force and torque sensors.</p></li><li><p><a href="https://tesollo.com">Tesollo</a> in South Korea is developing highly dexterous robotic hands and its own integrated actuator technology.</p></li><li><p><a href="https://www.synapticon.com">Synapticon</a> in Germany is integrating motors, servo drives and electronics into compact motion and actuator systems.</p></li><li><p><a href="https://www.thingap.com">ThinGap</a> in America makes lightweight, high-performance motors.</p></li><li><p><a href="https://www.daimonrobotics.com">Daimon Robotics</a> in China is developing high-resolution tactile sensing.</p></li></ul><p>These companies aren&#8217;t trying to build the next Optimus. They&#8217;re making the <strong>motors, sensors, actuators and hands that every robot manufacturer could end up buying</strong>.</p><p>That could be a very good place to be if production takes off.</p><h2>The boring companies are having a very good year</h2><p>The listed companies are interesting too. A surprising number of the businesses making the less glamorous parts of the robotics supply chain have had very strong share-price moves this year.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!DGB9!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!DGB9!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!DGB9!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png" width="1200" height="800.2747252747253" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:971,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1817964,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/213815872?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!DGB9!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 424w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 848w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 1272w, https://substackcdn.com/image/fetch/$s_!DGB9!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F22fd6ebc-7bf0-4cf2-bcf4-40aa85c906c4_1536x1024.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>These aren&#8217;t pure robotics companies, so I wouldn&#8217;t pretend humanoids explain all of those moves. But it&#8217;s hard not to notice how many companies making motion, sensing, gearing and precision components are having very good years.</p><h3>Batteries probably aren&#8217;t the first problem</h3><p>I assumed batteries would be near the top of the list. Millions of robots means millions of batteries.</p><p>But battery production is already enormous because of EVs and energy storage. Even a few million humanoids wouldn&#8217;t move global battery demand anything like another few million electric cars would.</p><div id="youtube2-mHP1WGlw5Wk" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;mHP1WGlw5Wk&quot;,&quot;startTime&quot;:&quot;39s&quot;,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/mHP1WGlw5Wk?start=39s&amp;rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><p>The bigger issue is weight. A larger battery gives the robot more runtime, but it also makes it heavier. That means more powerful motors, which use more power. You quickly end up chasing your own tail.</p><p>So energy density matters. Solid-state batteries could make a real difference here, but I still don&#8217;t think batteries are the first place supply gets tight.</p><h2>Magnets could be</h2><p>Many high-performance motors use neodymium permanent magnets, and <strong>China dominates their supply chain</strong>.</p><p>That becomes a problem if America, Europe, Japan and South Korea start building millions of robots at home while still relying on China for a basic ingredient inside the motors.</p><p>It also explains why Australia&#8217;s <strong>Lynas Rare Earths</strong> belongs in this story despite never making a robot.</p><h2>Ten thousand robots is one thing. Ten million is another.</h2><p>You can build 10,000 expensive robots with specialist suppliers and plenty of manual production. At ten million, you need something completely different.</p><p>Parts have to become cheaper and more standardised. Suppliers need much bigger factories. Production yields matter. And no manufacturer spending billions on a robot factory will want its entire production line dependent on one small supplier in Bavaria.</p><p>That&#8217;s when the obscure stuff starts to matter.</p><p>A shortage of planetary roller screws could do it. So could precision reducers, miniature motors, encoders, force sensors or permanent magnets.</p><p>It might be something even more boring that I&#8217;ve never heard of.</p><p><strong>That&#8217;s what I&#8217;m looking for.</strong></p><p></p><h3>Stop looking at the robot</h3><p>We&#8217;ve seen this with AI already. ChatGPT got everyone&#8217;s attention, then Nvidia made the fortune supplying the compute. That created demand for data centres, memory and networking, which needed transformers, copper, gas turbines and a frankly ridiculous amount of electricity.</p><p>Robotics is going to put the same sort of pressure on manufacturing.</p><p>We can build extraordinary robots today. Building ten million of them every year is a completely different problem. That means hundreds of millions of motors, bearings, gears, screws, encoders and sensors, all made accurately enough, cheaply enough and quickly enough.</p><p><strong>Something won&#8217;t keep up.</strong></p><p>It might be roller screws, precision reducers, magnets, motors or tactile sensors. It could just as easily be some obscure component made by a 70-year-old engineering company that nobody outside its industry has ever heard of.</p><p>Everyone is trying to work out <strong>who will win the race to build the robot.</strong></p><p><strong>I&#8217;m more interested in who already makes the thing they all discover they can&#8217;t get enough of.</strong><span><br><br></span></p><div><hr></div><h2>Go take a and look, For the &#10084;&#65039; of startups</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p><br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[A New Financial Cycle Has Begun]]></title><description><![CDATA[US Debt, currencies, central banks, China, commodities and AI are not separate stories. I connect them all and explain what is changing, why it matters and what comes next.]]></description><link>https://insights.fusion-42.com/p/a-new-financial-cycle-has-begun</link><guid isPermaLink="false">https://insights.fusion-42.com/p/a-new-financial-cycle-has-begun</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sun, 23 Aug 2026 21:27:42 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!dduo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<h3 style="text-align: justify;"><em><span>Why the race for AI, power and strategic materials is colliding with the global debt problem</span></em></h3><blockquote><p style="text-align: justify;"><em><span>The AI boom is not happening in the cloud. It is being built with debt, electricity, copper, chips and rare-earth magnets. That physical arms race is arriving just as governments need to refinance record debts&#8212;and China controls many of the industrial bottlenecks.</span></em></p></blockquote><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!dduo!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!dduo!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 424w, https://substackcdn.com/image/fetch/$s_!dduo!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 848w, https://substackcdn.com/image/fetch/$s_!dduo!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 1272w, https://substackcdn.com/image/fetch/$s_!dduo!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!dduo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png" width="1456" height="819" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/dc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:819,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:77613,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!dduo!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 424w, https://substackcdn.com/image/fetch/$s_!dduo!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 848w, https://substackcdn.com/image/fetch/$s_!dduo!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 1272w, https://substackcdn.com/image/fetch/$s_!dduo!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fdc7a2883-d75f-4c26-b857-9dcbdfdb1656_1600x900.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3 style="text-align: justify;"><span><br><br><br></span><strong><span>THE SIMPLE VERSION </span></strong></h3><p style="text-align: justify;"><span>Governments across the developed world need to borrow more just as inflation is stopping central banks from making money cheap. AI is adding another vast demand for capital, electricity, copper and strategic minerals. Japan is being squeezed by its currency, Europe by issuance, and America by debt and refinancing. China has weak domestic demand but controls critical processing and magnet supply chains. Oil is the accelerant. Gold and Bitcoin are the market&#8217;s vote against the policy mix.</span></p><h3 style="text-align: justify;"><strong><span>THE OVERLOOKED MACRO SHOCK </span></strong></h3><p style="text-align: justify;"><span>AI can eventually make goods and services cheaper. Building it is inflationary first. Technology companies are competing with governments for capital, with households for electricity and with industry for copper, transformers and strategic minerals. China&#8217;s control of refining and magnets turns that commodity squeeze into geopolitical leverage.</span></p><h3 style="text-align: justify;"><strong><span>This week&#8217;s market message</span></strong></h3><p style="text-align: justify;"><span>The clearest signal was not any single price. It was the combination: long-term government yields remained near multi-decade highs, the dollar fell, oil rose and monetary alternatives rallied. Normally, higher US yields support the dollar. When yields rise and the dollar falls together, investors are asking for more return because they see greater fiscal and inflation risk&#8212;not because American growth is suddenly stronger.<br></span></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!NIDW!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!NIDW!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!NIDW!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png" width="1080" height="1350" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1350,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:137989,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!NIDW!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!NIDW!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F15236b37-f3c8-464a-a7a0-c964929eddd4_1080x1350.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h3 style="text-align: justify;"><strong><span>The AI&#8211;China&#8211;commodity chain</span></strong></h3><p style="text-align: justify;"><span>This is the layer that connects the technology boom to bonds, inflation and geopolitics. AI infrastructure creates a new buyer for capital and physical resources while China controls much of the processing needed to turn those resources into usable components.</span></p><p style="text-align: justify;"></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!q3fr!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!q3fr!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 424w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 848w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!q3fr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png" width="1080" height="1080" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1080,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:101978,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!q3fr!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 424w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 848w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!q3fr!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F54005603-b7e6-478f-86f5-d6698de17b99_1080x1080.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p style="text-align: justify;"></p><h3 style="text-align: justify;"><strong><span>THE STRATEGIC CONTRADICTION </span></strong></h3><blockquote><p style="text-align: justify;"><span>America leads much of the AI stack, but China controls many of the physical inputs required to scale it. The competition is no longer just about models and chips. It is about capital, grids, copper, processing, magnets and power.</span></p></blockquote><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">1. The bond market is repricing the developed world</span></mark></strong></h2><p style="text-align: justify;"><span>For most of the past fifteen years, governments could borrow heavily because central banks held rates near zero and bought bonds. That regime has ended. Inflation is higher, central banks are no longer absorbing the same amount of debt, and governments are issuing more to fund deficits, defence, energy security, ageing populations and infrastructure. At the same time, technology companies are borrowing vast sums to finance AI infrastructure.</span></p><p style="text-align: justify;"><span>The result is simple supply and demand. More bonds are being sold, while the largest dependable buyer&#8212;the central bank&#8212;is stepping back. Private investors will still buy, but only at a price. That price is a higher yield.</span></p><h3 style="text-align: justify;"><strong><span>WHY IT MATTERS </span></strong></h3><p style="text-align: justify;"><span>A higher government yield becomes the base rate for the rest of the economy. Mortgages, corporate debt, project finance and equity valuations all become more expensive.</span></p><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">2. America is the centre of the refinancing problem</span></mark></strong></h2><p style="text-align: justify;"><span>US gross federal debt has passed $40 trillion. The more important point is not the headline number by itself; it is the speed at which old debt must be refinanced and new deficits funded. Treasury has deliberately issued heavily at the short end, using bills rather than locking in today&#8217;s long-term rates.</span></p><p style="text-align: justify;"><span>That swaps duration risk for refinancing risk. Short bills barely move in price when long yields jump, so the balance sheet looks less sensitive today. But those bills mature quickly. If rates stay high, the interest bill resets within weeks or months rather than years.</span></p><p style="text-align: justify;"><span>Using the estimate that roughly $6.45 trillion must be refinanced before year-end, every additional basis point paid across that amount adds about $645 million to annual interest expense. That is an approximation&#8212;the debt will mature at different points and different yields&#8212;but it shows the scale. A move of 50 basis points is roughly $32 billion a year.</span></p><h3 style="text-align: justify;"><strong><span>Treasury showed its pain threshold</span></strong></h3><p style="text-align: justify;"><span>When the 30-year yield reached 5.34%, Treasury announced that selected long-end buybacks would double from $2 billion to at least $4 billion per operation between September and November. The 30-year yield initially fell to about 5.19%, then returned to roughly 5.27% by Friday.</span></p><p style="text-align: justify;"><span>The buyback is not QE. Treasury is exchanging one form of government debt for another, not creating money through the Fed. It can improve liquidity and reduce long-bond supply, but it does not reduce the deficit. If the purchase is financed with more bills, it shortens the debt profile and increases future refinancing exposure.</span></p><p style="text-align: justify;"><span>The operation is also small beside a $32.2 trillion Treasury market. Its importance was the signal: Washington revealed that a 30-year yield around 5.30% is politically and financially uncomfortable. Markets test visible pain thresholds.</span></p><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">3. Japan is the pressure point</span></mark></strong></h2><p style="text-align: justify;"><span>Japan spent decades with near-zero rates. Its banks, insurers and investors therefore placed capital abroad, while global funds borrowed cheap yen to buy higher-yielding assets elsewhere. That is the yen carry trade.</span></p><p style="text-align: justify;"><span>Now Japanese inflation is rising, oil is expensive and the yen remains weak. The Bank of Japan is under pressure to raise rates. Its 10-year yield ended Friday around 2.88%, the 20-year near 3.76% and the 30-year around 4.06%. Those are enormous changes for a system built around almost-free money.</span></p><p style="text-align: justify;"><span>Higher Japanese yields make domestic bonds more attractive. They also make borrowing yen more expensive. Japanese capital has more reason to return home, while leveraged investors have more reason to close carry trades. Closing the trade means selling the foreign asset and buying back yen. That can hit US Treasuries, technology shares and other risk assets at the same time.</span></p><h3 style="text-align: justify;"><strong><span>Why the Fed&#8217;s FIMA facility matters</span></strong></h3><p style="text-align: justify;"><span>If Japan needs dollars to defend the yen, it could sell part of its large Treasury portfolio. FIMA offers a bridge: Japan can temporarily pledge Treasuries to the Fed and receive dollars rather than dumping the bonds into the market. It can then sell those dollars and buy yen.</span></p><p style="text-align: justify;"><span>That protects the Treasury market from forced selling in the short term. It is a secured, reversible loan&#8212;not automatic QE. The concern starts if seven-day funding is repeatedly rolled, expanded or treated as permanent. At that point, temporary liquidity support becomes ongoing official balance-sheet support.</span></p><p style="text-align: justify;"><strong><span>JAPAN&#8217;S DILEMMA </span></strong></p><blockquote><p style="text-align: justify;"><span>Raise rates and risk unwinding the carry trade. Hold rates down and risk further yen weakness, imported inflation and more intervention.</span></p></blockquote><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">4. Europe has the same disease with different symptoms</span></mark></strong></h2><p style="text-align: justify;"><span>Europe is not being singled out by the market; it is suffering from the same global repricing. Germany&#8217;s 30-year yield reached about 3.79%, its highest since 2011. French 10-year yields moved above 4.1%, while long French yields approached 5%. Italy and the UK face the same refinancing pressure.</span></p><p style="text-align: justify;"><span>Germany is loosening its old fiscal rules to fund defence and infrastructure. Euro-area governments are borrowing more for defence, healthcare and ageing populations. Record German issuance is expected, while the European Central Bank is allowing bonds on its balance sheet to mature rather than replacing them. More supply is therefore being pushed onto private investors.</span></p><p style="text-align: justify;"><span>The ECB faces an awkward split. It must control inflation across the whole bloc, but higher yields hurt highly indebted countries much more than Germany. If it supports the weakest sovereign markets too aggressively, it risks appearing to finance governments. If it does too little, spreads widen and the euro area fragments.</span></p><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">5. China is moving in the opposite direction</span></mark></strong></h2><p style="text-align: justify;"><span>China is the major exception. Its five-year government yield is around 1.39% and its 10-year yield near 1.69%, a 13-month low. While developed-market yields are rising because governments need more money and investors fear inflation, Chinese yields are falling because households save heavily, private credit demand is weak and the economy is struggling to generate enough productive domestic borrowing.</span></p><p style="text-align: justify;"><span>That is useful for Beijing: it can finance stimulus and infrastructure more cheaply than the West. It also gives China room to use fiscal policy while other governments are being disciplined by bond markets. But crashing yields are not proof of effortless strength. They also signal weak consumption, poor loan demand, pressure on banks and lingering property problems.</span></p><p style="text-align: justify;"><span>China could buy higher-yielding US, European or Japanese debt, but the decision is not purely about return. It must also consider currency risk, sanctions and the danger of holding reserves inside a system Washington can weaponise. That is why China is likely to diversify selectively: some foreign bonds, more gold, more renminbi trade settlement and more financial infrastructure outside the dollar system.</span></p><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">6. How the AI&#8211;China commodity shock works</span></mark></strong></h2><p style="text-align: justify;"><span>AI is usually discussed as software, productivity and company valuations. The immediate build-out is far more physical. Data centres require land, grid connections, transformers, cooling systems, gas turbines, advanced chips and large amounts of copper. High-performance motors, power systems, semiconductors, aerospace and defence also depend on rare earths and other strategic minerals.</span></p><p style="text-align: justify;"><span>The largest US technology companies are expected to spend about $725 billion on AI-related capital expenditure in 2026. Alibaba has just proposed raising roughly $10.2 billion for chips, infrastructure and models. Nebius raised $5 billion through convertible bonds for data centres and AI capacity. This money does not come from nowhere. AI companies are competing with governments, utilities and industry for the same capital. That adds corporate issuance to an already crowded bond market.</span></p><h3 style="text-align: justify;"><strong><span>Power is the first bottleneck</span></strong></h3><p style="text-align: justify;"><span>The International Energy Agency expects global data-centre electricity use to more than double to around 945 terawatt-hours by 2030&#8212;slightly more than Japan consumes today. Data centres are expected to account for nearly half of US electricity-demand growth this decade. That means more grids, generation, storage, gas and nuclear investment. It also means higher power prices where infrastructure cannot keep up.</span></p><p style="text-align: justify;"><span>AI can be highly deflationary once it raises productivity. Building the infrastructure is inflationary first. It absorbs capital, skilled labour, energy equipment and commodities years before the full productivity benefit arrives.</span></p><h3 style="text-align: justify;"><strong><span>Copper is the clearest broad commodity signal</span></strong></h3><p style="text-align: justify;"><span>Copper traded around a record $6.73 per pound this week and is up more than 18% this year. AI data centres, grid reinforcement, renewable generation and electrification are all increasing demand while new mines take years to permit and build. The IEA still projects a copper supply gap of roughly 25% by 2035 based on announced projects. China also controls about half of global copper smelting capacity, so the vulnerability is not only in mining; it sits in processing.</span></p><h3 style="text-align: justify;"><strong><span>Rare earths are a geopolitical price, not one market price</span></strong></h3><p style="text-align: justify;"><span>Rare earths do not trade on one transparent global exchange like copper. There are different oxides, metals, alloys and magnets, often sold through bilateral contracts. China accounted for about 60% of magnet rare-earth mining in 2024, 91% of refining and 94% of sintered permanent-magnet production. The real leverage sits in separation, refining and magnet manufacturing&#8212;not simply in owning ore.</span></p><p style="text-align: justify;"><span>That has created two markets. China&#8217;s domestic rare-earth price index fell to 260.8 in August from roughly 310 in March, and its NdPr oxide benchmark dropped about 12% from July. Yet controlled heavy rare earths remained dramatically dearer outside China. Indicative August transactions put terbium oxide near $990&#8211;$999 per kilogram inside China against roughly $3,625&#8211;$4,500 outside; dysprosium showed the same divide. These are opaque bilateral prices, but the message is clear: export licences and political access can matter more than the domestic commodity quote.</span></p><p style="text-align: justify;"><span>China increased some yttrium and permanent-magnet shipments to the US ahead of trade talks while continuing to restrict dysprosium and terbium exports to Japan. That links the AI and mineral story directly to the yen and carry trade. China can apply industrial pressure to Japan without selling a single Treasury.</span></p><h3 style="text-align: justify;"><strong><span>THE STRATEGIC EDGE </span></strong></h3><blockquote><p style="text-align: justify;"><span>America leads much of the AI stack. China controls many of the physical inputs needed to scale it. The next phase of the contest is therefore about grids, copper, processing, magnets and power&#8212;not just models and chips.</span></p></blockquote><p style="text-align: justify;"><span>Western governments will respond with subsidies, price floors, strategic stockpiles and domestic processing. Those policies improve resilience, but facilities outside China can cost 20% to more than 150% more to build and roughly 50% more to operate. Supply-chain security therefore adds another layer of spending, deficits and inflation to the global debt problem.</span></p><h3 style="text-align: justify;"><strong><span>The coercion trap</span></strong></h3><p style="text-align: justify;"><span>The proposed US MATCH Act shows how technology controls feed back into this system. If enacted, it would give allied supplier countries such as the Netherlands and Japan 150 days to align their semiconductor-equipment controls with Washington. If they did not, the United States could extend its jurisdiction over foreign machines containing American technology and restrict both sales and servicing in China. The bill is not yet law, but it advanced from the House Foreign Affairs Committee by 36 votes to eight.</span></p><p style="text-align: justify;"><span>The short-term security logic is clear. American export controls achieve little if Dutch or Japanese suppliers can provide the same essential equipment. But forcing an ally to restrict its own national champion carries a strategic cost. China represented 29.1% of ASML&#8217;s total net sales in 2025, and the Dutch government has formally objected to the proposed law&#8217;s extraterritorial reach.</span></p><p style="text-align: justify;"><span>Each restriction gives China another reason to accelerate domestic lithography, retaliate through rare earths and redesign supply chains around non-American technology. It also gives Europe and Japan an incentive to remove US components from their own products so that Washington can no longer control where they are sold. The policy can slow China today while weakening the network of dependence and trust on which American power rests tomorrow.</span></p><h3 style="text-align: justify;"><strong><span>AMERICA IS SPENDING ITS POWER </span></strong></h3><blockquote><p style="text-align: justify;"><span>America is using its dominance to slow China today. But every time it coerces an ally, it gives the world another reason to build a system that America cannot control tomorrow.</span></p></blockquote><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">7. Oil is the accelerant</span></mark></strong></h2><p style="text-align: justify;"><span>Brent ended the week close to $95 as the Iran conflict and uncertainty around the Strait of Hormuz kept supply risk elevated. About a fifth of the world&#8217;s oil and liquefied natural gas normally passes through the strait. Higher energy prices feed directly into transport, food, manufacturing and household bills.</span></p><p style="text-align: justify;"><span>This matters because governments need lower yields at the exact moment oil is pushing inflation higher. Central banks cannot cut aggressively into a fresh energy shock without risking their credibility and currencies. The US threat of tougher sanctions on Iran may increase economic pressure on Tehran, but it also raises the oil and inflation risk facing the US Treasury and the Fed. One arm of policy is working against the other.</span></p><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">8. Every major central bank is trapped differently</span></mark></strong></h2><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!1D--!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!1D--!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 424w, https://substackcdn.com/image/fetch/$s_!1D--!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 848w, https://substackcdn.com/image/fetch/$s_!1D--!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!1D--!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!1D--!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png" width="1080" height="1080" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/e9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1080,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:85599,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!1D--!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 424w, https://substackcdn.com/image/fetch/$s_!1D--!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 848w, https://substackcdn.com/image/fetch/$s_!1D--!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 1272w, https://substackcdn.com/image/fetch/$s_!1D--!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe9aebc6e-dcf2-44e6-830c-3fe5d5b91a31_1080x1080.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><blockquote><p style="text-align: justify;"><span>This is why the world feels unstable. There is no single interest-rate policy that solves the problem. Each central bank can remove pressure from one part of the system only by moving it somewhere else.</span></p></blockquote><h2 style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">9. This is a dollar unwind&#8212;not yet &#8216;Sell America&#8217;</span></mark></strong></h2><p style="text-align: justify;"><span>A true Sell America event would mean the dollar, Treasuries and US equities falling together in disorderly fashion, with poor auctions and damaged market liquidity. We are not there. US shares rose on Friday, buyers still appeared at auctions and Treasuries continue to trade normally.</span></p><p style="text-align: justify;"><span>But the trade is spreading. The dollar fell nearly 1% over the week while long yields remained high. Gold rose more than 5%, silver finished near $69.62 and Bitcoin had its strongest week in roughly two and a half years. Investors are not abandoning America. They are reducing the amount of unhedged confidence they place in the dollar and nominal government debt.</span></p><h3 style="text-align: justify;"><strong><span>What reserve managers are actually doing</span></strong></h3><p style="text-align: justify;"><span>Central-bank balance sheets show a gradual diversification rather than a sudden flight from the dollar. The dollar&#8217;s share of disclosed global foreign-exchange reserves has fallen from about 71% in 1999 to 57.13% in the first quarter of 2026. But it rose from 56.42% in the previous quarter and has been broadly stable near current levels since 2022. The direction is clear; the pace is slow.</span></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!P10R!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!P10R!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!P10R!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!P10R!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!P10R!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!P10R!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png" width="1080" height="1350" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1350,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:112522,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!P10R!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!P10R!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!P10R!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!P10R!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd43759a3-ce9e-443c-9e85-f0818d00d79d_1080x1350.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p style="text-align: justify;"><span>The ownership mix matters as much as the total. Foreign private investors still buy American assets, including record amounts of US equities, while foreign official institutions have become less reliable Treasury buyers. The United States is therefore becoming more dependent on domestic investors, private foreign capital and leveraged market structures rather than reserve managers whose demand was traditionally less price-sensitive.</span></p><h3 style="text-align: justify;"><strong><span>DE-DOLLARISATION HAPPENS AT THE MARGIN </span></strong></h3><blockquote><p style="text-align: justify;"><span>The dollar does not need to be replaced for US borrowing costs to rise. Reserve managers only need to direct less of each new dollar into Treasuries and more into gold, other currencies or domestic assets. With Treasury issuance expanding, a weaker marginal buyer matters long before the dollar loses reserve status.</span></p></blockquote><h3 style="text-align: justify;"><strong><span>Why gold is the cleanest signal</span></strong></h3><p style="text-align: justify;"><span>Gold has no issuer, no refinancing calendar and no promise from a government. It benefits when real confidence in currencies and sovereign debt falls. Bitcoin expresses a similar idea with much greater volatility and technology risk. Silver also carries industrial demand, so it is a less pure monetary signal.</span></p><p style="text-align: justify;"><span>Equities can still rise in nominal terms. Companies with pricing power and real cash flow can pass on inflation. The danger is that persistently high real yields eventually reduce valuations and make financing harder&#8212;especially for businesses dependent on cheap capital.</span></p><p style="text-align: justify;"><strong><span>10. The feedback loop</span></strong></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!kpMB!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!kpMB!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!kpMB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png" width="1080" height="1350" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/ea998d15-7618-4787-8dd9-87c416954687_1080x1350.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1350,&quot;width&quot;:1080,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:71435,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/212442124?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!kpMB!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 424w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 848w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 1272w, https://substackcdn.com/image/fetch/$s_!kpMB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fea998d15-7618-4787-8dd9-87c416954687_1080x1350.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p style="text-align: center;"></p><p style="text-align: justify;"><span>That is the debt spiral people are worried about. It does not mean default tomorrow. Countries that borrow in their own currency have powerful tools. They can tax, cut spending, extend maturities, change regulation, provide liquidity and allow inflation to reduce the real value of old fixed-rate debt.</span></p><p style="text-align: justify;"><span>The problem is that these tools have costs. Tax rises weaken growth. Spending cuts are politically difficult. Short funding increases refinancing risk. Currency weakness raises inflation. Central-bank purchases damage credibility if they become permanent. The spiral begins when the market sees every rescue adding to the next problem.</span></p><h3 style="text-align: justify;"><strong><span>What this means for ordinary people</span></strong></h3><p style="text-align: justify;"><strong><span>BORROWERS </span></strong><span>Mortgage and loan rates can stay high even if central banks begin cutting overnight rates.</span></p><p style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">SAVERS </span></mark></strong><span>Cash earns more interest, but energy and food inflation can still reduce purchasing power.</span></p><p style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">PENSIONS </span></mark></strong><span>Higher yields help new bond buyers but damage the price of long-dated bonds already held.</span></p><p style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">BUSINESSES </span></mark></strong><span>Debt-heavy firms and companies without pricing power face the greatest squeeze.</span></p><p style="text-align: justify;"><strong><mark data-color="#f1de71" style="background-color: rgb(241, 222, 113); color: rgb(0, 0, 0);"><span data-color="#14101b" style="color: rgb(20, 16, 27);">INVESTORS </span></mark></strong><span>Diversification matters more when bonds and equities can fall together. Hard assets help, but they are not guaranteed one-way trades.</span></p><h3 style="text-align: justify;"><strong><span>What would make this worse</span></strong></h3><p style="text-align: justify;"><span>The dangerous combination is oil above $100, the US 30-year yield holding above 5.30%, dollar&#8211;yen breaking decisively through 160, weak Treasury auctions, wider French and Italian spreads, and official liquidity facilities being enlarged or repeatedly rolled. If the dollar falls while all of that happens, the market is adding a fiscal credibility premium.</span></p><h3 style="text-align: justify;"><strong><span>What would calm it down</span></strong></h3><p style="text-align: justify;"><span>A durable fall in oil and inflation would give central banks room to cut. Credible fiscal restraint would reduce future bond supply. Strong auctions without official support would show private demand is intact. A stable yen after a measured Bank of Japan move would reduce carry-trade risk. Better Chinese domestic demand would make its low yields look like successful easing rather than a warning about weak growth.</span></p><h3 style="text-align: justify;"><strong><span>The next tests</span></strong></h3><p style="text-align: justify;"><span>Next week brings 2-, 5- and 7-year US Treasury auctions, fresh US growth and inflation data, Nvidia&#8217;s results and the Federal Reserve&#8217;s Jackson Hole gathering. Japan&#8217;s September policy meeting is the next major pressure point for the yen and the carry trade. Europe&#8217;s auctions and sovereign spreads will show whether the stress remains global or begins concentrating in weaker countries.</span></p><p style="text-align: justify;"><span>Watch the relationships, not just the levels. Falling yields with a firmer dollar would be relief. Falling yields with a collapsing dollar would look like financial repression. Rising yields with a stronger dollar would be conventional tightening. Rising yields with a weaker dollar is the warning signal.</span></p><h3 style="text-align: justify;"><strong><span>The bottom line</span></strong></h3><p style="text-align: justify;"><strong><span>The world is leaving the cheap-money era with far more debt than it entered with. America is fighting refinancing risk. Japan is fighting the consequences of ending zero rates. Europe is fighting issuance and fragmentation. China is fighting weak domestic demand while controlling strategic processing and magnets. AI is creating a new call on capital, power and copper. Reserve managers are adding gold and reducing their marginal dependence on Treasuries. Oil is making every decision harder. The market has found the contradiction: governments need lower yields while inflation, debt supply and physical scarcity demand higher ones.</span></strong></p><blockquote><p style="text-align: justify;"><span>This is not yet a global debt crisis. It is the stage before one, when policy still works&#8212;but each intervention lasts for less time, costs more and moves the pressure somewhere else. It&#8217;s not looking pretty that&#8217;s for sure.</span></p></blockquote><p style="text-align: justify;"></p><p style="text-align: justify;"></p><p style="text-align: justify;"></p><p style="text-align: justify;"><strong><span>Sources and further reading</span></strong></p><p style="text-align: justify;"><a href="https://www.reuters.com/world/china/global-markets-global-markets-2026-08-21/"><span>Global markets: yields and oil remain high</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/world/us-treasury-double-sizes-some-debt-buyback-operations-least-4-billion-2026-08-19/"><span>US Treasury doubles selected long-end buybacks</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/business/bonds-bounce-us-buybacks-relief-may-be-brief-2026-08-20/"><span>Why the bond-market relief may be brief</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/business/record-german-debt-sales-deepen-strains-europes-battered-bond-market-2026-08-21/"><span>European debt issuance and rising yields</span></a></p><p style="text-align: justify;"><a href="https://www.brecorder.com/news/40435931"><span>Japan&#8217;s bond yields and inflation pressure</span></a></p><p style="text-align: justify;"><a href="https://www.ft.com/content/c3223b7c-6683-42df-9e60-9dbd2dbdfe07"><span>China&#8217;s falling yields and weak credit demand</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/world/china/china-seen-holding-loan-rates-steady-august-despite-economic-weakness-2026-08-19/"><span>China holds benchmark lending rates</span></a></p><p style="text-align: justify;"><a href="https://www.iea.org/reports/energy-and-ai/executive-summary%C2%A0"><span>IEA: energy and AI</span></a></p><p style="text-align: justify;"><a href="https://www.iea.org/reports/rare-earth-elements/executive-summary"><span>IEA: rare-earth concentration and economic exposure</span></a></p><p style="text-align: justify;"><a href="https://www.iea.org/reports/global-critical-minerals-outlook-2026/executive-summary"><span>IEA: 2026 critical-minerals outlook</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/business/retail-consumer/alibaba-proposes-hong-kong-share-placement-worth-10-billion-2026-08-23/"><span>Alibaba&#8217;s proposed $10.2bn AI capital raise</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/technology/nebius-plans-45-billion-convertible-debt-sale-fund-data-centers-ai-platform-2026-08-19/"><span>Nebius raises $5bn for AI data centres</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/world/asia-pacific/china-exports-us-rise-rare-earth-critical-aerospace-sector-2026-08-20/"><span>China&#8217;s strategic rare-earth shipments to the US and Japan</span></a></p><p style="text-align: justify;"><a href="https://www.mainrichinternational.com/resources/articles/rare-earth-market-update-august-2026-ndpr-pricing-mofcom-western-supply"><span>August rare-earth pricing and regional premiums</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/legal/government/china-criticizes-us-chip-equipment-bill-run-up-beijing-talks-2026-05-13/"><span>Reuters: MATCH Act and allied semiconductor controls</span></a></p><p style="text-align: justify;"><a href="https://www.sec.gov/Archives/edgar/data/937966/000162828026011378/asml-20251231.htm"><span>ASML 2025 annual report and China exposure</span></a></p><p style="text-align: justify;"><a href="https://data.imf.org/en/news/imf%20data%20brief%20july%201"><span>IMF: currency composition of reserves, Q1 2026</span></a></p><p style="text-align: justify;"><a href="https://www.federalreserve.gov/econres/notes/feds-notes/the-international-role-of-the-u-s-dollar-2025-edition-20250718.html"><span>Federal Reserve: the international role of the dollar</span></a></p><p style="text-align: justify;"><a href="https://www.gold.org/goldhub/research/gold-demand-trends/gold-demand-trends-q2-2026/central-banks"><span>World Gold Council: central-bank buying in Q2 2026</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/world/china/foreign-holdings-us-treasuries-fall-june-led-by-japan-uk-china-data-shows-2026-08-17/"><span>Reuters: Japan and China reduce Treasury holdings</span></a></p><p style="text-align: justify;"><a href="https://www.federalreserve.gov/monetarypolicy/fima-repo-facility.htm"><span>Federal Reserve explanation of FIMA repo</span></a></p><p style="text-align: justify;"><a href="https://www.reuters.com/business/gold-steadies-heads-third-straight-weekly-gain-2026-08-21/"><span>Gold and precious-metals close</span></a></p><p style="text-align: justify;"><a href="https://fiscaldata.treasury.gov/datasets/debt-to-the-penny/"><span>US Treasury Debt to the Penny</span></a></p><p style="text-align: justify;"><em><span>This report is commentary for general information, not personal investment advice. Market levels are approximate closing figures for Friday, 21 August 2026, and can change rapidly.</span></em></p><p style="text-align: right;"></p><p><span><br><br><br></span></p><h2>Go take a and look</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Intelligence Is Abundant. Judgement Is Not.]]></title><description><![CDATA[Arthur's starting judgement wasn't Googled. It was earned the expensive way.]]></description><link>https://insights.fusion-42.com/p/intelligence-is-abundant-judgement</link><guid isPermaLink="false">https://insights.fusion-42.com/p/intelligence-is-abundant-judgement</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Tue, 14 Jul 2026 14:31:18 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!KJsM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!KJsM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!KJsM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 424w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 848w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1272w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" width="800" height="361" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:361,&quot;width&quot;:800,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:674287,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/206570350?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!KJsM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 424w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 848w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1272w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><span>For most of history, intelligence was the bottleneck. That constraint has gone. Frontier models will draft your strategy, your financial model and your investor update in seconds, and rather well. </span></p><p><span>Yet founders are not obviously better off, because intelligence was never the whole job. Intelligence produces options. Judgement decides what matters when every option looks plausible. The models flooded the first market and left the second untouched. Intelligence became abundant. Judgement stayed scarce. Most AI tooling is still solving the old problem.</span></p><h2><span>You have already met the problem</span></h2><p><span>You have done this. Pasted your deck into a chatbot, had a genuinely useful conversation, then come back a week later and started from zero. It did not know your company then and it does not know it now.</span></p><p><span>That is not a memory bug waiting for a patch. A chat transcript is a pile of language. Nothing in it is dated, nothing is versioned, nothing separates a passing remark from an established fact, and advice built on a pile of language is generic by construction. That is why the chatbot&#8217;s answer to your company sounds suspiciously like its answer to every company.</span></p><p><span>And it agrees with you. These systems are tuned to be pleasant because pleasant retains users. You are already the most optimistic person in every room. The last thing you need is a brilliant assistant whose deepest instinct is to tell you that you are right.</span></p><h2><span>Trust is an architecture problem</span></h2><p><span>The obvious question about any AI that claims to know your company: why won&#8217;t it make things up?</span></p><p><span>Fair question, because we watched it happen. An early version of Arthur kept a running summary of each company, written by a model. One day it stated, with complete confidence, that a company was incorporated in a jurisdiction it had never touched. No data said so. The model completed a plausible pattern, because that is what models do. We wrote rules against it. The rules failed.</span></p><p><span>So we stopped asking the model to remember anything important. Arthur keeps the facts of your company separately from the AI, in a record where every entry is dated and traceable to something real: a document you uploaded, a thing you said, a result that happened. The consequence is simple. Arthur cannot invent your company&#8217;s history because it never owns that history in the first place. And when new information contradicts the record, the conflict is put in front of you to settle, not quietly papered over.</span></p><p><span>The parts that judge you are deterministic. Same inputs, same answer, and you can inspect the reasoning. A chatbot cannot tell you why it said something. Arthur can show you.</span></p><h2><span>Strong opinions, loosely held</span></h2><p><span>Strong opinions, loosely held, create healthy discourse and are the fastest way to truth. A co-founder who agrees with everything is dead weight.</span></p><p><span>So when Arthur says your raise is not ready, that is a strong opinion: a verdict with the workings attached. Attack it. Win the argument and the record updates, along with every judgement that depends on it. Lose the argument and you&#8217;ve learned something before an investor teaches it to you at far greater cost.</span></p><p><span>And because Arthur holds a real picture of your company, it does not wait to be asked. It finds the gap between what you are and what a fundable company at your stage looks like, and brings you the next decision with the reasoning attached. Ask a chatbot nothing and you get nothing.</span></p><h2><span>Where the standard comes from</span></h2><p><strong><span>Ready by whose definition? </span></strong><span>This is where judgement gets concrete. Judgement is knowing which missing customer reference matters and which does not. When $300k is enough to raise and when $3m is too much. That fixing your pricing is worth more than your next feature. Which investor objection is fatal and which is simply the opening move of a negotiation.</span></p><p><span>Frontier models reason surprisingly well, and that is not the gap. What they lack is a standard grounded in outcomes. Most written startup advice comes from people who never operated or wrote a cheque, and the calls above were never written down at all. They live in results: the rounds that closed, the deals that died, the companies that ran out of road six months after everyone agreed they were fine.</span></p><p><span>That judgement was earned at Fusion42, over years on both sides of the table: mentoring founders, assessing businesses, investing capital and seeing which decisions survived contact with reality. And because markets move, that judgement keeps evolving. Your company is measured against this quarter&#8217;s market, not a model&#8217;s memory of an old one.</span></p><h2><span>Judgement as infrastructure</span></h2><p><span>Arthur starts with Fusion42&#8217;s judgement. But the bigger opportunity is not preserving one firm&#8217;s experience. It is creating a system where judgement itself becomes an asset that can be captured, tested and improved.</span></p><p><strong><span>The future is not one AI with one opinion.</span></strong><span> Every investor, accelerator and operator holds their own views about what makes a company investable. Over time, those views become theses: explicit beliefs about what matters, what does not, and why. Today those theses mostly live in heads: applied inconsistently, invisible to scrutiny, and gone when the person leaves.</span></p><p><strong><span>The next step is opening the judgement layer: </span></strong><span>letting investors encode their own theses into the same transparent framework everything else runs on, adjusting the weightings, adding the criteria that matter to them, making their assumptions visible. Not by training the AI, which would reopen every trust problem above, but by writing judgement down where it can be read. And it reads both ways: the same system that lets investors encode how they judge lets founders see exactly how they are being judged.</span></p><p><strong><span>And written down means testable.</span></strong><span> The record says: here is your thesis, here is where it worked, and here is where reality disagreed. Theses become decisions. Decisions become outcomes. Outcomes refine judgement. That loop is why the system cannot quietly become a mirror for anyone&#8217;s bias. Bad judgement gets exposed. Good judgement compounds.</span></p><p><strong><span>Intelligence scales by replication. </span></strong><span>Judgement scales by encoding. The scarce asset was never one firm&#8217;s opinion. It is the ability to capture, structure and compound human judgement. Arthur is the mechanism.</span></p><h2><span>Your company is the asset</span></h2><p><span>Everything Arthur records about your company accumulates into an asset you can inspect, move and own, not dissolved into a model&#8217;s weights. Six months in, that record holds what no chatbot session can: what your company believed, when it changed its mind, and why. New tools forget. Arthur compounds.</span></p><p><strong><span>And it is yours. In institutional deployments it is completely ringfenced: </span></strong><span>nothing flows back, nothing trains a global model, nothing averages you into everyone else. The prevailing AI business model runs the other way, one central brain quietly enriched by absorbing what makes each customer distinct. If judgement is the scarce asset, that concentrates it in exactly the wrong place.</span></p><h2><span>The point</span></h2><p><span>The question is no longer whether AI can think. It is whether it knows enough about your company to be critical and objective for right reasons.<br><br><br></span></p><h2>Go and look</h2><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout - who just got funded</a></strong><br>Newly funded startups, pre-seed to Series B, as the rounds happen.</p></li></ul><ul><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise - who&#8217;s deploying</a></strong><br>New funds with fresh capital and cheques to write.</p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire - what changed, what matters</a></strong><br>The intelligence feed that filters the noise, with the &#8220;so what&#8221; attached.</p></li></ul><p></p><h2></h2><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[We built this for Arthur. It felt selfish to keep it.]]></title><description><![CDATA[Scout , Raise and Wire by Fusion 42]]></description><link>https://insights.fusion-42.com/p/we-built-this-for-arthur-it-felt</link><guid isPermaLink="false">https://insights.fusion-42.com/p/we-built-this-for-arthur-it-felt</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sun, 12 Jul 2026 14:30:40 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!KJsM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!KJsM!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!KJsM!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 424w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 848w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1272w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif" width="800" height="361" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:361,&quot;width&quot;:800,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:674287,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/206570350?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!KJsM!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 424w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 848w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1272w, https://substackcdn.com/image/fetch/$s_!KJsM!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff2ac8c4b-5c89-4a18-87bd-e45979e8200f_800x361.gif 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p>For the past year we&#8217;ve been building Arthur - a startup operating system that works  alongside founders as your co-founder, shaping your company on live intelligence rather than gut feel and stale advice. Part of that intelligence is pure market signal: who just got funded, who&#8217;s deploying fresh capital, what changed overnight - localised to your geography, personalised to your news type and sector. </p><p>We built that signal layer for Arthur. Then we realised it was too useful to keep private.</p><p><strong>So we&#8217;re releasing it. Three products, free to every founder.</strong></p><h2><a href="https://scout.fusion-42.com/scout-frontend">Scout - know who just got funded</a></h2><p>Newly funded startups, pre-seed to Series B, as the rounds happen.</p><p>Every round in your space is intelligence. A competitor raising &#163;8m tells you what investors are paying for. A fund backing three companies adjacent to yours tells you exactly who to email next.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Dnbj!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Dnbj!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 424w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 848w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 1272w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Dnbj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png" width="1456" height="1288" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/cc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1288,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:2161990,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/206570350?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Dnbj!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 424w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 848w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 1272w, https://substackcdn.com/image/fetch/$s_!Dnbj!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fcc7ebef0-e0c3-46ee-8aca-0e930d2b50dc_4732x4185.png 1456w" sizes="100vw"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2><a href="https://raise.fusion-42.com/raise-frontend">Raise - know who&#8217;s deploying</a></h2><p>New fund announcements: fresh capital, mandate, cheque size, focus.</p><p>A fund that closed eighteen months ago is a very different conversation to one that closed last month. New funds need to write cheques. Raise tells you who&#8217;s sitting on fresh powder - so you stop cold-emailing funds that finished deploying in 2024.</p><blockquote><p><em><strong>And watch for the next update: we're adding the investors behind every new raise in Scout.</strong></em></p></blockquote><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!ZFYn!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!ZFYn!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 424w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 848w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 1272w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!ZFYn!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png" width="1456" height="1425" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/f8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1425,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:2354410,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/206570350?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!ZFYn!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 424w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 848w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 1272w, https://substackcdn.com/image/fetch/$s_!ZFYn!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8551124-f8e0-426d-b923-25fa79a78fbf_4381x4288.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2><a href="https://wire.fusion-42.com/wire-frontend">Wire - know what changed, know what matters</a></h2><p>The global intelligence feed, filtered hard, with the &#8220;so what&#8221; attached.</p><p>Every story carries The Wire Takeaway - not &#8220;Samsung raises foundry prices&#8221; but &#8220;your wafer costs just went up 10-15%, call your suppliers this week. Filter by News type, sector, region to get fully personalised and localised news. Weekly reports roll it up so you can skim Monday morning and get back to building.</p><p><strong>Reading everything is a full-time job. You already have one.</strong></p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!pEyB!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!pEyB!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 424w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 848w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 1272w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!pEyB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png" width="1456" height="1464" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1464,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:579566,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/206570350?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!pEyB!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 424w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 848w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 1272w, https://substackcdn.com/image/fetch/$s_!pEyB!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F765642c9-c561-4e95-940f-f5581e06849c_2296x2308.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2>Where this goes</h2><p>These feeds are one part of what Arthur runs on. Combined with what Arthur knows about your company - your stage, your traction, your gaps - raw signal becomes something sharper: which investors to approach this month, which shift to act on, which assumption someone else&#8217;s round just invalidated.</p><p>You&#8217;ll hear much more about Arthur soon. For now, this layer is yours.</p><h2>Go and look</h2><p>No paywall to browse. </p><ul><li><p><strong><a href="https://scout.fusion-42.com/scout-frontend">Scout</a></strong><a href="https://scout.fusion-42.com/scout-frontend"> - who just got funded</a></p></li><li><p><strong><a href="https://raise.fusion-42.com/raise-frontend">Raise</a></strong><a href="https://raise.fusion-42.com/raise-frontend"> - who&#8217;s deploying</a></p></li><li><p><strong><a href="https://wire.fusion-42.com/wire-frontend">Wire</a></strong><a href="https://wire.fusion-42.com/wire-frontend"> - what changed, what matters</a></p></li></ul><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[The Raise Report]]></title><description><![CDATA[New funds with capital to deploy June 2026]]></description><link>https://insights.fusion-42.com/p/the-raise-report-cf4</link><guid isPermaLink="false">https://insights.fusion-42.com/p/the-raise-report-cf4</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Tue, 30 Jun 2026 08:29:56 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Qr-U!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Qr-U!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Qr-U!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 424w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 848w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 1272w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Qr-U!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png" width="392" height="277.03846153846155" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1029,&quot;width&quot;:1456,&quot;resizeWidth&quot;:392,&quot;bytes&quot;:2766147,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/170330465?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Qr-U!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 424w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 848w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 1272w, https://substackcdn.com/image/fetch/$s_!Qr-U!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F4d4634a4-3568-4ca2-b4c1-65993ab1991f_3368x2380.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><strong><span>This report provides an overview of new venture capital funds, CVCs and accelerators that have announced new capital to invest in early-stage startups during June 2026.  </span></strong></p><p></p><h2>Monthly Report</h2><p>June 2026 saw a robust activity in early-stage funding, with a total of <strong>91 new funds </strong>identified raised <strong>$14.1B.</strong><br>Artificial Intelligence (AI) infrastructure continues to dominate investment interest, followed by generative AI and climate technology. The United States remains the leading hub for these investments, though the United Kingdom and several Asian and European countries also show significant activity. Several large funds, including Sky Factory and Menlo Ventures XVII, have made substantial commitments, indicating continued confidence in the early-stage ecosystem.</p><h2 style="text-align: center;"><span><br>Full details of all funds and people live at </span><strong><a href="https://raise.fusion-42.com"><span>RAISE.FUSION-42</span></a></strong></h2><div class="captioned-image-container"><figure><div class="image-link image2" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!znAt!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 424w, https://substackcdn.com/image/fetch/$s_!znAt!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 848w, https://substackcdn.com/image/fetch/$s_!znAt!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 1272w, https://substackcdn.com/image/fetch/$s_!znAt!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!znAt!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png" width="1200" height="146.77002583979328" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:142,&quot;width&quot;:1161,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:24345,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/204238573?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!znAt!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 424w, https://substackcdn.com/image/fetch/$s_!znAt!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 848w, https://substackcdn.com/image/fetch/$s_!znAt!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 1272w, https://substackcdn.com/image/fetch/$s_!znAt!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F43458fa2-0b58-43cd-a9a8-ca0ec816112d_1161x142.png 1456w" sizes="100vw"></picture><div></div></div></div></figure></div><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!bBZ-!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!bBZ-!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 424w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 848w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 1272w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!bBZ-!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png" width="638" height="779.5343406593406" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;normal&quot;,&quot;height&quot;:1779,&quot;width&quot;:1456,&quot;resizeWidth&quot;:638,&quot;bytes&quot;:455090,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/204238573?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!bBZ-!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 424w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 848w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 1272w, https://substackcdn.com/image/fetch/$s_!bBZ-!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F88ae10da-f108-4efb-8c86-b0841e9b407a_1457x1780.png 1456w" sizes="100vw"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><div><hr></div><h3><span>SECTORS</span></h3><p>AI-related sectors, particularly AI Infrastructure and Generative AI, are at the forefront of early-stage investments, reflecting the ongoing technological advancements and market demand in artificial intelligence. Climate Tech and Clean Energy also feature prominently, highlighting a growing focus on sustainable solutions. Traditional sectors like Agtech, Enterprise Software, and Digital Health continue to attract significant capital, indicating sustained innovation across diverse industries.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!x6Nh!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!x6Nh!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 424w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 848w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 1272w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!x6Nh!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png" width="572" height="526.4285714285714" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/d8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1340,&quot;width&quot;:1456,&quot;resizeWidth&quot;:572,&quot;bytes&quot;:118388,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/204238573?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!x6Nh!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 424w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 848w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 1272w, https://substackcdn.com/image/fetch/$s_!x6Nh!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8ba2c41-5bf1-49cf-b435-b14838be760c_1600x1473.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3>GEO</h3><p>The United States overwhelmingly leads in the number of new early-stage funds, reinforcing its position as a global innovation hub. The United Kingdom is a strong second, with notable activity in supporting emerging fund managers and specific tech sectors. Countries like China, Japan, and Israel are also significant players, particularly in advanced technology and AI-focused investments.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!WwaG!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!WwaG!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 424w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 848w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 1272w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!WwaG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png" width="594" height="593.184065934066" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1454,&quot;width&quot;:1456,&quot;resizeWidth&quot;:594,&quot;bytes&quot;:111939,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/204238573?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!WwaG!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 424w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 848w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 1272w, https://substackcdn.com/image/fetch/$s_!WwaG!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5cda1cf5-a0f0-45b6-8a7d-7b05e15e0f49_1600x1598.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3><span>Notable Large Fund Launches</span></h3><p>Several funds with substantial capital commitments were launched in June 2026. Notably, Sky Factory by Dreame Technology and Menlo Ventures XVII stand out with multi-billion dollar funds, indicating a strong appetite for significant investments in promising early-stage ventures. These large funds are poised to make a considerable impact on the growth and development of startups in their respective focus areas.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!--J0!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!--J0!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 424w, https://substackcdn.com/image/fetch/$s_!--J0!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 848w, https://substackcdn.com/image/fetch/$s_!--J0!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 1272w, https://substackcdn.com/image/fetch/$s_!--J0!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!--J0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png" width="1456" height="1226" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1226,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:175973,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/204238573?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!--J0!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 424w, https://substackcdn.com/image/fetch/$s_!--J0!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 848w, https://substackcdn.com/image/fetch/$s_!--J0!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 1272w, https://substackcdn.com/image/fetch/$s_!--J0!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F3ba94ba9-c4d2-406e-93c7-7811534093ff_1600x1347.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2></h2><p><span>June 2026 demonstrates a vibrant and dynamic early-stage investment landscape, heavily influenced by advancements in AI and a growing commitment to climate technology. While the United States continues to be a dominant force, global participation is robust, with significant contributions from the UK, China, Japan, and other nations. The launch of numerous large funds underscores a healthy capital environment for innovative startups, particularly those addressing critical technological and environmental challenges.</span></p><div><hr></div><div class="pullquote"><p><em>If you have not joined the Fusion42 Community on Telegram &#8212;<br>it is probably time to do so.</em></p><p><em>For the &#10084;&#65039; of Startups</em></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://www.fusion-42.com/&quot;,&quot;text&quot;:&quot;Join us here&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://www.fusion-42.com/"><span>Join us here</span></a></p></div><p></p><p>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek<br><br><br><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><div><hr></div>]]></content:encoded></item><item><title><![CDATA[How did SpaceX get to be a $2.5tn company?]]></title><description><![CDATA[That&#8217;s the right question. Not &#8220;is it worth it&#8221; - how did it get there.]]></description><link>https://insights.fusion-42.com/p/how-did-spacex-get-to-be-a-25tn-company</link><guid isPermaLink="false">https://insights.fusion-42.com/p/how-did-spacex-get-to-be-a-25tn-company</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Fri, 19 Jun 2026 06:55:51 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!maip!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<p>That&#8217;s the right question. Not &#8220;is it worth it&#8221;, <em>how did it get there.</em> Because it&#8217;s not a valuation. It&#8217;s a story with a massive price tag, and the story is the best one ever told. </p><p>Read the S-1. They claim a $28.5tn market, then admit a few pages later it &#8220;may not be accurate.&#8221; And how&#8217;d they get to $28.5tn? Forty-odd consultant reports they didn&#8217;t pay for, run through &#8220;management&#8217;s interpretation,&#8221; topped with three separate disclaimers telling you not to rely on it. <strong>Big-number masturbation</strong> to blow our little minds, with no consequences for any of it.</p><h3><strong>Take the story away and what&#8217;s actually there?</strong> </h3><p>Starlink. A real business, $11.4bn revenue, fat margins. Everything else is just burning: $4.9bn lost last year, another $4.3bn in Q1, a $41bn hole and counting. Starlink alone doesn&#8217;t get you near $2.5tn.</p><blockquote><p><em><strong>So where&#8217;s the rest coming from? The bit written by Asimov, of course. The company tells you itself, describing the Moon bases and the asteroid mining: &#8220;technologies that do not exist.&#8221; Their words, in the filing.</strong></em></p></blockquote><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!maip!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!maip!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 424w, https://substackcdn.com/image/fetch/$s_!maip!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 848w, https://substackcdn.com/image/fetch/$s_!maip!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 1272w, https://substackcdn.com/image/fetch/$s_!maip!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!maip!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png" width="535" height="943" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:943,&quot;width&quot;:535,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:525522,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/202682484?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!maip!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 424w, https://substackcdn.com/image/fetch/$s_!maip!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 848w, https://substackcdn.com/image/fetch/$s_!maip!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 1272w, https://substackcdn.com/image/fetch/$s_!maip!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F8b32799c-9e44-41bd-8586-c29f251d7954_535x943.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><p></p><h3><strong>Now watch the engineering, because this bit already happened.</strong></h3><p>Buy Twitter at $44bn. Fold it into xAI. Fold xAI into SpaceX. Every step a common-control merger, every step a fresh mark, the whole lot recast in the accounts like it was always one company.</p><p>Revenue barely moved. Losses went vertical,  $6bn gone in the AI segment alone. And the mark went up the entire way.</p><p>That&#8217;s not value creation. That&#8217;s value declaration. He didn&#8217;t build the business, he re-papered it three times and the market clapped each time. Value asserted through structure, not earned through cash flow.</p><h3><strong>And here&#8217;s the cleverest bit, the same trick aimed at himself.</strong></h3><p>Everyone reads his pay packages as the ultimate alignment. Tie my comp to the share price, I only win if you win. Brilliant optics. But look at what it actually does. The targets are market-cap milestones, not cash milestones. So he&#8217;s not paid to make money,  he&#8217;s paid to move the price. And the price moves on the story. He&#8217;s literally incentivised to grow the narrative, because the narrative is the thing he&#8217;s best at. The KPI isn&#8217;t a check on him. It&#8217;s a reward for doing exactly what he was already doing.</p><p>And the size does its own work. A pay number so absurd it stops reading as money and starts reading as conviction. <em>Look what he&#8217;s tied himself to, he must believe.</em> Same as the $28.5tn, too big to evaluate, so you stop evaluating and start feeling. <strong>The bigness is the persuasion.</strong></p><p>Then check the votes. Class B, ten votes a share, electing a majority of the board. &#8220;Controlled company&#8221; status, which lets him skip the independent compensation committee, the one body that&#8217;s meant to sit between a CEO and his own pay. So he sets the package, through a board he elects, with votes that outweigh what he actually owns. He sells you &#8220;fully aligned&#8221; while building the exact structure that means nobody can ever hold him to it.</p><h3><strong>It&#8217;s the same play at every level.</strong></h3><blockquote><p>On the market, a TAM you can&#8217;t check. On the businesses, value declared through mergers, not made. On himself, pay that rewards storytelling and a vote structure that removes the referee. Three layers, one trick: pick a number too big to test, and let our own greed and awe do the validating for him.</p></blockquote><p><strong>And it&#8217;s not hidden. It&#8217;s all in the filings. We just don&#8217;t read them.</strong></p><h3><strong>So what comes next?</strong></h3><p>Watch for the reverse-merge into Tesla. Tesla&#8217;s already in the S&amp;P. SpaceX, losing money, can&#8217;t get in on its own numbers. Fold it in and every index tracker on Earth becomes a forced buyer overnight. Then <strong>pile the sci-fi </strong>back on top, Optimus on the Moon, robots mining asteroids, factories in orbit, and the merged thing carries a $10tn cap. A dream too far out to ever be proven wrong in time to matter.</p><p><strong>The story is the funding. </strong>The dream makes capital cheap, cheap money builds the one thing that works, and that makes the dream look real. The belief pays for itself. Works right up until the real business stops outrunning what the dream burns. Right now the fire&#8217;s winning.</p><p>He&#8217;s not going to build a meaningful off-world economy in his lifetime. But he&#8217;s laid the foundations for two things: a real space economy, and proof that we&#8217;re suckers, driven by greed.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!DGv1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!DGv1!,w_424,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 424w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_848,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 848w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_1272,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 1272w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_1456,c_limit,f_webp,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!DGv1!,w_1456,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif" width="516" height="472" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;normal&quot;,&quot;height&quot;:472,&quot;width&quot;:516,&quot;resizeWidth&quot;:516,&quot;bytes&quot;:8312597,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/gif&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/202682484?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!DGv1!,w_424,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 424w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_848,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 848w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_1272,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 1272w, https://substackcdn.com/image/fetch/$s_!DGv1!,w_1456,c_limit,f_auto,q_auto:good,fl_lossy/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F442e8397-5fc0-40f0-a0f8-8cc38630fb52_516x472.gif 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3><em>Respect to the man. He&#8217;s a genius, not at rockets, at us. At reading the system and the psychology that moves the herd. He didn&#8217;t really hack physics. He hacked us and our belief system.</em></h3><p></p><p></p><p></p><p></p><p><br>For the &#10084;&#65039; of startups<br><br>&#9996;&#127996; &amp; &#128153;<br>Derek</p><div><hr></div><p><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div><p></p><div><hr></div>]]></content:encoded></item><item><title><![CDATA[The AI labs won’t follow you down. Go build the hard thing.]]></title><description><![CDATA[The models are great. Infrastructure built. Capital committed. What's missing is what makes it move.]]></description><link>https://insights.fusion-42.com/p/the-ai-labs-wont-follow-you-down</link><guid isPermaLink="false">https://insights.fusion-42.com/p/the-ai-labs-wont-follow-you-down</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Mon, 15 Jun 2026 04:37:19 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!jbPc!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<blockquote><p>Let me start with the receipts, because they matter for what comes next.  </p></blockquote><p>In January 2024 I wrote that vertical SaaS was about to get smashed, that AI would collapse the cost of replicating specialised software, compress those famous 80&#8211;90% gross margins back toward the 45% the rest of the economy lives on, and force a shift from per-seat subscriptions to outcome pricing. People thought that was spicy. Two years on, the financial press has a name for what happened: the &#8220;SaaSpocalypse.&#8221; Software lost its premium for the first time in two decades, the SaaS Capital Index fell from around 7x ARR at the start of 2025 to roughly 3.8x by March 2026, the software ETF dropped about 30% from its September 2025 peak, and somewhere on the order of $2 trillion in market cap came off the sector. The seat-based model, Salesforce, Workday, Atlassian, is the explicit thing analysts are now downgrading, because once an agent does the work of ten seats, the link between headcount and revenue doesn&#8217;t just slow, it reverses.</p><p>If anything, it arrived more completely than I described. I said companies would stop buying specialised software and start building it in-house, because the cost of building was about to collapse. I pictured the low-code of the day, dragging blocks on a canvas. What actually showed up was more total than that: natural-language app-builders like Lovable, Bolt and v0, where a non-technical founder describes the tool they need and watches it get built and deployed, no engineers, no seat licences, no canvas. The wrapper wave softened the ground first: thousands of thin AI products that died fast but proved, publicly, how cheap software had become to produce. Then the build-it-yourself tools removed the last excuse to pay $100 a seat for something you could now just make. Same deflationary force I named, arriving harder than I&#8217;d dared predict. The destruction came, and it came early.</p><p>By November 2024 I was writing about the other half: if the old thing is dying, here&#8217;s the new thing. Software stops being a tool you operate and becomes a service that delivers the outcome, Service as Software, $4.6 trillion of work that used to need a human, done by software. Outcome pricing, agents that complete tasks rather than assist with them. I wasn&#8217;t the first to put a name to it,  the idea was already moving through the VC world by late 2024, but I was on it early, while it was still a niche thesis rather than the wallpaper it&#8217;s become. It&#8217;s the consensus take of 2026 now, which is a polite way of saying it stopped being controversial.</p><p>So I&#8217;m not here to walk any of it back. Both halves landed. The destruction and the replacement both arrived roughly on schedule. I&#8217;m writing now because the part that actually matters has come into focus, the part neither of those pieces had: where the defensible thing is. Everyone can see the category. Almost nobody&#8217;s saying the right thing about the moat.</p><p><strong>Here&#8217;s the thing though. &#8220;Go vertical&#8221; gets said a lot, </strong> but almost nobody&#8217;s actually doing it in a tight enough space to matter. They say it, then they flinch, because they&#8217;re convinced the labs will crush anything they build. That fear is the whole opportunity. The gap isn&#8217;t that nobody knows where the moat is. It&#8217;s that everybody knows, and almost nobody&#8217;s brave enough, or frankly equipped enough, to go and take it. So let me spell out exactly why the labs can&#8217;t follow you down. Once you see the mechanism, the fear looks misplaced.</p><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!jbPc!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!jbPc!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 424w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 848w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 1272w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!jbPc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png" width="1456" height="861" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:861,&quot;width&quot;:1456,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:668334,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/202069900?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!jbPc!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 424w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 848w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 1272w, https://substackcdn.com/image/fetch/$s_!jbPc!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F721b66c9-0064-457b-8988-89f2b54c143c_2062x1220.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2>The asymmetry nobody&#8217;s pricing in</h2><p>OpenAI and Anthropic sell open loops. They have to. A model that serves a billion people can&#8217;t close a loop, because closing a loop means owning the outcome, the permissions, the exceptions, the liability, the write-back into someone&#8217;s real system, the moment your software becomes the system of record. &#8220;Be a safe, general assistant to everyone&#8221; and &#8220;form a specific opinion about your specific business and act on it&#8221; are opposite postures. The first is their entire franchise. The second is the only thing a real business actually wants. So they hand you intelligence and let you do the last mile.</p><p>And the last mile isn&#8217;t a mile. It&#8217;s the whole job. The distance between an AI that can tell you how to build a financial model and an AI that builds yours, watches your edits, catches the contradiction three rows down, recomputes everything downstream, and then holds a view you have to argue with, that distance is the entire business. Everything before the loop closes is a demo. Only the closed loop is a product.</p><h2>Open loop vs closed loop</h2><p>An open loop ends with advice. A closed loop ends with the work done,  and then goes round again.</p><div class="callout-block" data-callout="true"><p>Open loop: &#8220;Here&#8217;s how you&#8217;d structure your cap table.&#8221; You now go and do it.</p><p>Closed loop: the cap table exists, it&#8217;s right, it updated when your round terms changed, and the system flagged that your new option pool breaks an assumption in the model you signed off last week.</p></div><p>The labs live on the left. Every business problem worth money lives on the right. The chatbot that explains the thing is commoditised the day a better model ships. The system that owns the thing end to end, with the right permissions, handling the exceptions, carrying the liability, holding state,  gets more valuable that same day.</p><p>But here&#8217;s where most people, including me in 2024, get the word &#8220;outcome&#8221; wrong. We sold Service as Software as outcomes, deliver the result, not the tool. True, but incomplete, because &#8220;outcome&#8221; smuggles in a finish line. Deal closed. Model built. Task done. Loop ends. That&#8217;s not how a business works. A business doesn&#8217;t shut down the day it closes a deal, it books the deal, learns from how the deal went, adjusts, and goes again. And again. The outcome isn&#8217;t the end of the loop. It&#8217;s one turn of it.</p><p>That&#8217;s what closed actually means. Not &#8220;finished&#8221;, completes the circuit and comes back round. An organic loop, the way a living business is organic: every turn feeds the next, it compounds, it gets better the more it runs, and it never stops. A one-off outcome is just an open loop with a tidier ending. The thing worth building is the loop that learns from its own outcome and uses it to make the next one sharper.</p><p>And notice this is precisely what a single brilliant answer can never be. The labs are the apex of the one-off: ask, receive, done, forget. Stateless by design. The most powerful one-shot outcome machine ever built, and structurally incapable of going round again on your specific business, because going round again means remembering, and remembering everyone&#8217;s business is not a product you can ship to a billion people.</p><p>That&#8217;s the part I want to nail, because it inverts the usual moat argument. Most moats erode as models improve. A thin wrapper gets more disposable with every release. A closed-loop business gets stronger, because the expensive part the loop, the state, the proprietary corpus, the texture of a workflow you understand better than anyone alive, is precisely the part a model upgrade doesn&#8217;t touch. A better engine makes an open loop cheaper and a closed loop deadlier. You want to be on the side of that trade where the giants&#8217; progress is your tailwind, not your obituary.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!b4WT!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!b4WT!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 424w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 848w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 1272w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!b4WT!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png" width="1200" height="642.1875" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:548,&quot;width&quot;:1024,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:901967,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/202069900?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!b4WT!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 424w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 848w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 1272w, https://substackcdn.com/image/fetch/$s_!b4WT!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F777ebcc9-9c2b-449a-938b-3fdd0998cbc9_1024x548.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2>Why the labs can&#8217;t follow you  and why nobody&#8217;s exploiting it</h2><p>So if a closed loop is the whole game, why is almost nobody building one? Because it&#8217;s an engineering beast, and the reason it&#8217;s hard isn&#8217;t the number of parts, it&#8217;s that they stack, each one depending on the one above it.</p><p>At the top sits the goal, the outcome the business actually exists to move. Not a task. A direction. This is the part every harness diagram going around right now leaves out, and leaving it out is why those systems never close: a loop with no goal to measure against is just a hamster wheel. It acts, but it has no standard to judge the result by, so it can&#8217;t learn.</p><p>Below the goal sits intent, the goal broken down into the move that matters right now. What is the founder actually trying to do, as opposed to the words they typed? Get intent wrong and everything downstream executes the wrong thing perfectly.</p><p>Then memory, which is not one thing but two. Short-term, what happened in this thread, this session, the last five minutes,  is cheap, volatile plumbing. Long-term, every decision, correction, contradiction and exception accumulated across months, is the opposite: durable, curated, and the single hardest thing to build. Fold them into one &#8220;memory&#8221; box, as everyone does, and you hide the fact that the long-term one is the entire moat.</p><p>And only then, at the bottom, action, actually doing the thing, with permissions, in the real world, handling the cases where reality doesn&#8217;t match the plan.</p><p>Goal, intent, the two memories, action, each one load-bearing for the next, and the whole stack measured against the goal at the top so the loop can close and come round sharper. That&#8217;s not a prompt. It&#8217;s a system, and a genuinely hard one. (How you actually wire it, and why the standard &#8220;model in the middle&#8221; diagram has the topology backwards, is its own piece. Next one.)</p><p>Which is exactly why the opportunity is sitting there unexploited. It&#8217;s not that people haven&#8217;t spotted the gap. It&#8217;s that building this is genuinely difficult, and the fear of the labs gives everyone a convenient excuse not to attempt the hard thing. So they build another open-loop wrapper, ship a single-answer chatbot, and wonder why the next model release eats them. The people who&#8217;ll win are the ones willing to do the masterpiece-level engineering the labs structurally won&#8217;t,  because doing it would mean abandoning the thing that makes them labs.</p><h2>The new blocker isn&#8217;t the model. It&#8217;s the knowledge.</h2><p>Here&#8217;s what changed underneath all of this, and almost nobody&#8217;s saying it out loud. The models got good, shockingly good, shockingly fast. The blocker to automating real work is no longer intelligence. It&#8217;s domain knowledge.</p><p>Every company runs on know-how that lives nowhere you can point to. Some of it is in people&#8217;s heads. The rest is scattered across old email, Slack threads, support tickets, a database nobody fully understands, and a process three people half-remember. The business functions because humans vaguely know where the knowledge is and how to apply it. That&#8217;s fine for humans. It&#8217;s useless for an agent. You cannot hand an AI &#8220;we kind of do it like this, ask Dave.&#8221; An agent needs the actual rules, how a refund gets handled, how a pricing exception gets decided, how an engineer triages an incident at 3am, structured, current, and executable.</p><p>So the missing layer isn&#8217;t a smarter model and it isn&#8217;t search-over-your-documents with a chat box on top. It&#8217;s a living map of how a company actually works, turned into something an AI can execute against safely and consistently. Pull the knowledge out of all those fragmented sources, structure it, keep it current, and you&#8217;ve built the thing that sits between raw company data and reliable automation. Every company in the world is going to need that layer. Most don&#8217;t have it and can&#8217;t build it.</p><p>And here&#8217;s the wedge, the bit that&#8217;s mine. Everyone chasing this is building it for companies that already exist: a decade of Slack to excavate, ten years of tickets to mine. The whole job is archaeology, dig the knowledge out of where it&#8217;s buried.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!-FzH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!-FzH!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 424w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 848w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 1272w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!-FzH!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png" width="1200" height="655.2197802197802" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:795,&quot;width&quot;:1456,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:604556,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/202069900?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!-FzH!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 424w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 848w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 1272w, https://substackcdn.com/image/fetch/$s_!-FzH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F91e6ad86-9a63-4f45-8436-fd192f0b33fe_2048x1118.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><br>A startup at pre-seed has nothing to dig. There&#8217;s no buried corpus. The company&#8217;s brain doesn&#8217;t exist yet, it&#8217;s being formed in real time, usually half-wrong, in the founder&#8217;s head. So you don&#8217;t excavate it. You build it from zero, as the company comes into existence. And because it&#8217;s being built rather than retrieved, it can do the thing archaeology never can: it doesn&#8217;t just record how the company works, it holds a view on whether the company is working. A map tells you where things are. This forms an opinion about a business the founder hasn&#8217;t fully figured out yet, and pushes back. That&#8217;s not memory. That&#8217;s judgment. Memory is for companies with a past. Judgment is for companies that only have a future.</p><p>Notice this is the same closed loop, one level down. A company brain that only answers questions is still an open loop, it tells you how your company works, you go act. Close it, and the brain forms the view, drives the action, watches the result, and corrects. The knowledge layer is the substrate. The closed loop is what you do with it. The corpus accreting underneath is what makes it impossible to copy.</p><h2>&#8220;Vertical&#8221; was the floor, not the destination</h2><p>So go vertical, but understand it&#8217;s the floor you start from, not the moat itself. Owning &#8220;legal&#8221; or &#8220;finance&#8221; isn&#8217;t narrow enough to be safe; those are still categories a lab could credibly wave at. The defensibility sits one or two levels below that: in the specific, annoying, regulated-adjacent workflow the labs will never bother to learn the texture of, where closing the loop requires knowing the exceptions that aren&#8217;t written down anywhere.</p><p>Not &#8220;accountancy.&#8221; The exact three-statement model a pre-seed founder actually needs, in their jurisdiction, with their messy assumptions, that recomputes when reality moves. Not &#8220;founder tools.&#8221; The continuous loop of observing a founder, forming a view of their company, and driving the next concrete action without waiting to be asked. That&#8217;s a closed loop by construction, and it&#8217;s not one a general assistant can ship, because forming an opinion about one company and acting on it is the opposite of being everyone&#8217;s safe co-pilot.</p><h2>The loop underneath the loop</h2><p>I buried this in the 2024 piece as a single section called &#8220;eating our own medicine.&#8221; It should have been the whole article. Your first 100 users teach the system things your 101st gets for free, automatically, in a way nobody outside can reconstruct. Every closed loop you run leaves a deposit, the correction a founder made, the exception that broke the template, the pattern that only shows up across hundreds of real companies. That corpus is the only asset on the board a competitor cannot buy, scrape, or prompt their way into. And it compounds.</p><p>The orchestration, master agents coordinating sub-agents, is essential. You can&#8217;t build a real vertical service without it, and it&#8217;s exactly how I&#8217;m building Arthur. What I got wrong in 2024 was treating it as the moat. It isn&#8217;t. It&#8217;s table stakes now, the labs and the frameworks ship the pattern for free. The defensibility was never the conductor. It&#8217;s the closed loop the conductor drives, and the corpus that loop is quietly building underneath.</p><h2>The three objections you&#8217;re already forming</h2><p>If you&#8217;ve followed this far you&#8217;ve got three rebuttals loaded. Good. Here they are, and here&#8217;s why none of them lands.</p><p><strong>&#8220;Context windows are huge now, just feed the whole company in.&#8221; </strong>A bigger window makes the model a better reader. It does nothing to make it a writer. You can paste your entire operating history into a million-token context and the model will hand you a brilliant answer, and then the turn ends and nothing in the world changed. It didn&#8217;t write the corrected number back, reconcile the three systems that disagree, or commit a change the next turn builds on. Reading context is not owning state. The window lets the model see your business; it never lets it run it. An open loop with an infinite window is still an open loop, it just reads more before it forgets. And every improvement in context size upgrades the cheap, readable half of my loop for free while leaving the expensive half, state, write-back, reconciliatio, exactly where it was.</p><p><strong>&#8220;Owning liability is ruinous, you&#8217;ll pay for every cleanup.&#8221; </strong>Only if &#8220;owning the outcome&#8221; means letting a probabilistic model make the consequential decision and then insuring the wreckage. It doesn&#8217;t. You earn the right to own the outcome not by being brave enough to absorb the losses, but by building the system so the losses can&#8217;t happen: the cap table isn&#8217;t generated by a model that might be off by a factor of ten, it&#8217;s computed by deterministic code that&#8217;s right every time, with the founder signing off the consequential write before it commits. The liability I&#8217;ll take on is precisely the liability the labs structurally can&#8217;t, because their architecture is the probabilistic thing that makes the number unguaranteeable in the first place. Owning liability isn&#8217;t the cost of closing the loop. It&#8217;s evidence you closed it properly. (How that architecture works is its own piece, the next one.)</p><p><strong>&#8220;Pre-seed founders have no money, selling to startups is a graveyard.&#8221; </strong>It is, for anyone whose startups are the revenue. Mine aren&#8217;t. The pre-seed founder is the forge, not the market, the cheapest, most varied training ground there is to perfect the loop, where a founder who fails still leaves their corrections in the corpus. The company dies; the learning doesn&#8217;t. And they&#8217;re not even the buyer: the accelerators and funds who back them pay, because they want a whole portfolio executing faster, founder as distribution surface, not wallet. The ones who survive don&#8217;t graduate off the product onto a &#8220;real&#8221; enterprise tool; they grow into the enterprise tier of it, because by then the system is their company&#8217;s operating history and ripping it out means ripping out their own memory. The graveyard kills tools you can leave. You can&#8217;t leave the thing that remembers how your company works.</p><h2>So, practically</h2><p>If you&#8217;re building in AI right now, the question isn&#8217;t &#8220;what&#8217;s my vertical.&#8221; Everyone has a vertical. The question is: where does my loop close, and who else could possibly close it?</p><p>If your answer ends in advice, you&#8217;ve built a demo, and a model release will end you. If it ends in the work done, owned, with the liability, the exceptions, the write-back, and a corpus accreting underneath that the next customer benefits from for free, then every time OpenAI and Anthropic get better, so do you.</p><p>And stop waiting for permission. The reason this opportunity is still open isn&#8217;t that it&#8217;s hidden, it&#8217;s that the engineering is hard and the fear of the labs gives everyone an excuse to flinch. That fear is misplaced. The labs aren&#8217;t going to come down into your closed loop; they can&#8217;t, without ceasing to be the labs. The only thing standing between you and the moat is whether you&#8217;re willing to build the hard thing.</p><p>Three pieces, two years, one direction of travel: the old model dies, the new one takes its shape, and the moat turns out to live in the closed loop and the corpus underneath it. I&#8217;ve called each step before it was obvious. I&#8217;m calling this one too. And this time I&#8217;m not just writing it, I&#8217;m building it.</p><p>Go find the ugliest, most painful, most specific loop you understand better than anyone. Then close it.</p><p>For the &#10084;&#65039; of startups.</p><p>&#9996;&#127996; &amp; &#128153; Derek</p><p></p><p></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://derekwatsonfusion42.substack.com/?utm_source=substack&amp;utm_medium=email&amp;utm_content=share&amp;action=share&quot;,&quot;text&quot;:&quot;Share Fusion42 by DeReK WaTSoN&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://derekwatsonfusion42.substack.com/?utm_source=substack&amp;utm_medium=email&amp;utm_content=share&amp;action=share"><span>Share Fusion42 by DeReK WaTSoN</span></a></p><div><hr></div><p><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Ready Player One (2018): Sci-Fi vs Reality]]></title><description><![CDATA[The future already happened. It was just fiction first.]]></description><link>https://insights.fusion-42.com/p/ready-player-one-2018-sci-fi-vs-reality</link><guid isPermaLink="false">https://insights.fusion-42.com/p/ready-player-one-2018-sci-fi-vs-reality</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sun, 29 Mar 2026 08:57:46 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!Bk7k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<blockquote><p><strong>Sci-Fi vs Reality</strong> <em>Did art imitate life, or did life imitate the inspiration?  </em></p><p>Every week I watch a sci-fi film and ask a simple question: where did this idea actually come from? Did fiction imagine the future first &#8212; or did reality quietly leak into the story before we noticed? </p><p>From there, I do the reality check. What already exists in today&#8217;s tech? What&#8217;s genuinely caught up with the film? And what still doesn&#8217;t &#8212; not because no one&#8217;s tried, but because something real is in the way. Physics. Economics. Regulation. Human behaviour.</p><p>Some ideas turn out to be pointless. Some are just early. Others are waiting on one or two very specific breakthroughs.</p><p>The chicken and the egg were never separate. They were always in conversation.</p></blockquote><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!Bk7k!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!Bk7k!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!Bk7k!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png" width="1200" height="654.5454545454545" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/bb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:768,&quot;width&quot;:1408,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:2007073,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192489846?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!Bk7k!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!Bk7k!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbb0286e5-2916-4118-b0a1-3b182293df9b_1408x768.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h1></h1><h3>Fiction gave us one OASIS. Reality gave us many, and none of them talk to each other. That&#8217;s the entire story of the metaverse, right there.</h3><h2><strong>What They Imagined</strong></h2><p>In 2018, <em>Ready Player One</em> showed us a bleak 2045 where humanity escapes into a single, seamless virtual universe called the OASIS. It was an infinite digital playground, economy, and social fabric rolled into one. With a simple VR visor and haptic suit, you could be anyone, go anywhere, and own anything. The OASIS solved for societal collapse by offering a better, more exciting digital alternative. It was one world, one login, one identity.</p><p>The writers saw it coming: a fully immersive digital layer for life. The key assumption was that it would all be connected.</p><p></p><div id="youtube2-cSp1dM2Vj48" class="youtube-wrap" data-attrs="{&quot;videoId&quot;:&quot;cSp1dM2Vj48&quot;,&quot;startTime&quot;:null,&quot;endTime&quot;:null}" data-component-name="Youtube2ToDOM"><div class="youtube-inner"><iframe src="https://www.youtube-nocookie.com/embed/cSp1dM2Vj48?rel=0&amp;autoplay=0&amp;showinfo=0&amp;enablejsapi=0" frameborder="0" loading="lazy" gesture="media" allow="autoplay; fullscreen" allowautoplay="true" allowfullscreen="true" width="728" height="409"></iframe></div></div><p></p><h2><strong>What We Actually Built</strong></h2><p>I used to think a unified metaverse was inevitable. Turns out, the economics are fucking brutal. Why would Roblox want you to take the digital goods you bought from them over to Meta&#8217;s world? There is zero incentive.</p><p>So instead of one OASIS, we built walled gardens. <em>Ready Player One</em> showed a unified world. Reality delivered platform-specific fortresses. Roblox built a colossal one, pulling in $3.6 billion in revenue in 2024 with nearly 83 million people logging in every day. Meta made a multi-hundred-billion-dollar bet on its own version, building the Quest hardware and Horizon Worlds.</p><p>Each platform is a powerhouse. Each is a silo. We have good-enough goggles, not a second skin. And the gap between these fortresses is where the dream of a true, open metaverse dies&#8212;and where the real opportunity for founders begins.</p><h2><strong>The Gap They Missed</strong></h2><p>The film&#8217;s biggest blind spot wasn&#8217;t the tech; it was the business model. It completely glossed over the raw, competitive drive for proprietary control. Building a single OASIS assumes a level of collaboration and open standards that capitalism just doesn&#8217;t favour. Each platform wants to be <em>the</em> platform, not a neighbourhood in someone else&#8217;s city.</p><p>Here&#8217;s the thing: the motivation is also different. We aren&#8217;t building the metaverse to escape a dystopian future; we&#8217;re building it to augment our slightly lonely, inefficient present. It&#8217;s about connection and utility, not desperation. The pattern I keep seeing is founders assuming people want pure escapism, when what they really want are better tools for reality. That&#8217;s a fundamental disconnect between the fiction and the market need.</p><h2><strong>The Players</strong></h2><p>The giants built the islands. The startups are building the bridges. They aren&#8217;t trying to create another OASIS; they&#8217;re creating the protocols and tools to connect the ones we already have. This isn&#8217;t a Silicon Valley story, either. The work is happening everywhere.</p><ul><li><p><strong>Decentraland (Buenos Aires, Argentina):</strong> One of the original Web3 worlds where users own the land via NFTs. It&#8217;s a live experiment in creating the kind of user-owned, interoperable world the film imagined, proving the concept beyond a single corporate entity.</p></li><li><p><strong>The Sandbox (Hong Kong):</strong> A major player in the user-generated content space, backed by $93M and big partnerships. Their entire model is based on interoperable NFT assets, directly tackling the walled garden problem by design.</p></li><li><p><strong>Somnium Space (Prague, Czech Republic):</strong> A VR-first metaverse that&#8217;s deeply integrated with blockchain, allowing for true asset ownership and cross-chain compatibility. They&#8217;re building for the immersive future the film promised, but with an open economic layer.</p></li><li><p><strong>Aethir (Singapore):</strong> You can&#8217;t have a seamless metaverse without massive rendering power. Aethir is building a decentralized GPU cloud to provide the infrastructure needed for high-fidelity, interoperable experiences, solving a huge technical bottleneck.</p></li><li><p><strong>Highstreet (Los Angeles, USA):</strong> A hybrid of online shopping and gaming, creating a marketplace for digital goods that can &#8220;world-hop.&#8221; They are explicitly building for asset interoperability, allowing brands to exist across fragmented virtual worlds.</p></li><li><p><strong>Wilder World (London, UK):</strong> Aiming for a photorealistic, AAA-quality metaverse on the blockchain. Their focus on universal NFT standards is a direct attempt to create the &#8216;take it with you&#8217; asset portability that was effortless in the OASIS.</p></li><li><p><strong>Beam (Ho Chi Minh City, Vietnam):</strong> Instead of building another destination, Beam is building the engine. It&#8217;s a modular metaverse framework with interoperability baked in, giving developers the tools to create connected worlds from the start.</p></li><li><p><strong>Voxels (Sydney, Australia):</strong> An early, browser-based virtual world that has already formed interoperability pacts with platforms like Decentraland. They are proof that collaboration, not just competition, is a viable path forward.</p></li><li><p><strong>Everyworld (Tel Aviv, Israel):</strong> Using AI to help users generate interoperable 3D assets and content. This tackles the scalability problem of content creation while ensuring the assets aren&#8217;t locked into a single platform from day one.</p></li><li><p><strong>Ixana (Cape Town, South Africa):</strong> Pure-play middleware. Ixana is building the technical plumbing&#8212;the protocols and asset bridges&#8212;that allows different metaverse platforms to communicate. They&#8217;re not building a world; they&#8217;re building the universal translator for them.</p></li></ul><p></p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!izc5!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!izc5!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!izc5!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!izc5!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!izc5!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!izc5!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png" width="1200" height="654.5454545454545" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:768,&quot;width&quot;:1408,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1437030,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192489846?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!izc5!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!izc5!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!izc5!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!izc5!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F07adcf99-dd6a-403b-8f90-755022e11754_1408x768.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h2><strong>What&#8217;s Left to Build</strong></h2><p>We have startups building bridges and protocols. But the gap they missed is still the most personal one: you.</p><p>In <em>Ready Player One</em>, your avatar, your inventory, and your reputation were singular and persistent. We don&#8217;t have that. Your identity is fractured across a dozen platforms. The startup that still needs to exist isn&#8217;t another world, it&#8217;s a universal passport. A truly portable, user-owned digital identity that plugs into any game or virtual space.</p><p>This isn&#8217;t just an avatar. It&#8217;s a wallet, a social graph, and a reputation score that you own and control. It&#8217;s the thing that proves you are you, wherever you go in the digital realm. It&#8217;s less about building another world and more about building the passport that lets you travel between them.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!GHu1!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!GHu1!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 424w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 848w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 1272w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!GHu1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png" width="768" height="1376" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1376,&quot;width&quot;:768,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1348988,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192489846?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!GHu1!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 424w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 848w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 1272w, https://substackcdn.com/image/fetch/$s_!GHu1!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5dec2785-1f1c-4ffe-8f7d-ac9a66390094_768x1376.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-tqxFUX" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-tqxFUX" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2><strong>The Timing Signal</strong></h2><p>A generation is growing up inside Roblox and Fortnite; for them, this isn&#8217;t a strange new world, it&#8217;s just... the world. Their expectation of digital presence and ownership is native. That&#8217;s a massive behavioural shift.</p><p>Technically, the pieces are falling into place. The launch of devices like Apple&#8217;s Vision Pro validates the market for spatial computing, pushing the hardware and developer tools forward. On-device AI is becoming powerful enough to handle real-time rendering and interaction. And open standards like OpenXR are slowly gaining ground, creating a common language for developers. The incentive to connect the islands is finally starting to outweigh the desire to own the ocean.</p><div><hr></div><p>For the &#10084;&#65039; of startups &#9996;&#127996; &amp; &#128153;</p><h2><strong>Go Deeper</strong></h2><ul><li><p><strong><a href="https://www.facebook.com/Stockstoearnpage/posts/mark-zuckerbergs-meta-stake-is-currently-worth-about-230-billionmeta-platforms-i/122151394412938282/">Meta&#8217;s Scale:</a></strong><a href="https://www.facebook.com/Stockstoearnpage/posts/mark-zuckerbergs-meta-stake-is-currently-worth-about-230-billionmeta-platforms-i/122151394412938282/"> </a></p></li><li><p><strong><a href="https://s27.q4cdn.com/984876518/files/doc_financials/2024/ar/Roblox-2025-Proxy_2024-AR-1-1.pdf">Roblox&#8217;s 2024 Performance:</a></strong><a href="https://s27.q4cdn.com/984876518/files/doc_financials/2024/ar/Roblox-2025-Proxy_2024-AR-1-1.pdf"> </a></p></li><li><p><strong><a href="https://dimensionmarketresearch.com/report/spatial-computing-market/">Spatial Computing Market Growth:</a></strong><a href="https://dimensionmarketresearch.com/report/spatial-computing-market/"> </a></p></li><li><p><strong><a href="https://www.mordorintelligence.com/industry-reports/wearable-technology-market">Wearable Tech Market Projections:</a></strong><a href="https://www.mordorintelligence.com/industry-reports/wearable-technology-market"> </a></p></li></ul><p></p><p></p><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://derekwatsonfusion42.substack.com/?utm_source=substack&amp;utm_medium=email&amp;utm_content=share&amp;action=share&quot;,&quot;text&quot;:&quot;Share Fusion42 by DeReK WaTSoN&quot;,&quot;action&quot;:null,&quot;class&quot;:&quot;button-wrapper&quot;}" data-component-name="ButtonCreateButton"><a class="button primary button-wrapper" href="https://derekwatsonfusion42.substack.com/?utm_source=substack&amp;utm_medium=email&amp;utm_content=share&amp;action=share"><span>Share Fusion42 by DeReK WaTSoN</span></a></p><div><hr></div><p><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you've got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I'm keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe below and follow me on&nbsp;<a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a>&nbsp;or&nbsp;<a href="https://twitter.com/derek__Watson">Twitter</a>&nbsp;to never miss an update.&nbsp;</em></p></li></ul><p class="button-wrapper" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe now&quot;,&quot;action&quot;:null,&quot;class&quot;:null}" data-component-name="ButtonCreateButton"><a class="button primary" href="https://insights.fusion-42.com/subscribe?"><span>Subscribe now</span></a></p><div><hr></div>]]></content:encoded></item><item><title><![CDATA[Domain Histories: Privacy & Identity]]></title><description><![CDATA[The best founders are students of history]]></description><link>https://insights.fusion-42.com/p/domain-histories-privacy-and-identity</link><guid isPermaLink="false">https://insights.fusion-42.com/p/domain-histories-privacy-and-identity</guid><dc:creator><![CDATA[DeReK WaTSoN]]></dc:creator><pubDate>Sat, 28 Mar 2026 16:10:10 GMT</pubDate><enclosure url="https://substackcdn.com/image/fetch/$s_!zSP8!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png" length="0" type="image/jpeg"/><content:encoded><![CDATA[<blockquote><p><em>This is <strong>Domain Histories</strong> &#8212; a weekly series mapping startup domains from first  principles. The best founders are students of history. Patrick Collison met Visa's founder before building Stripe. Apoorva Mehta studied Webvan's collapse before building Instacart. They used history to make specific decisions about what to build, what to avoid, and when to move. Each issue takes one domain and gives you the full  map: who tried before, why they failed, where the money sits, what's shipping now, and what a founder entering today should do differently because of it. Not a market overview. Not a research report. The history that changes what you build.  </em></p></blockquote><h3></h3><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!zSP8!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!zSP8!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 1456w" sizes="100vw"><img src="https://substackcdn.com/image/fetch/$s_!zSP8!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png" width="1200" height="654.5454545454545" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:768,&quot;width&quot;:1408,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1945679,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:false,&quot;topImage&quot;:true,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192425282?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!zSP8!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!zSP8!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7c958b46-219e-404b-ac89-3fbb76f8bf33_1408x768.png 1456w" sizes="100vw" fetchpriority="high"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><h2></h2><h2><strong>You have 200 passwords. The industry that created that problem wants to sell you the solution.</strong></h2><p>Here&#8217;s the thing. The chaos of digital identity &#8212; the endless logins, the data breaches, the feeling that you&#8217;ve been copied and pasted across a thousand databases you&#8217;ll never find &#8212; isn&#8217;t the result of some grand, malicious plan.</p><p>The original sin of digital identity was a lack of one.</p><p></p><h3><strong>The Original Problem</strong></h3><p>The internet wasn&#8217;t built for people. It was built for machines to talk to each other. Identity was an afterthought, a patch. When services needed to know who you were, they did the simplest thing possible: they asked you to create a username and a password.</p><p>Every. Single. Time.</p><p>Each new service built its own little castle with its own little gate. Your bank, your email, your social network, your pizza delivery app. Hundreds of castles, hundreds of keys. This created two massive problems that define the market to this day.</p><p>First, a usability nightmare. You, the user, were forced to become a human keychain, juggling hundreds of credentials. Or, more likely, you just used the same password everywhere, turning one leaky castle into a catastrophic failure of the entire kingdom.</p><p>Second, a massive, fragmented attack surface. Your identity wasn&#8217;t a single thing you owned; it was a thousand little pieces of you, scattered across servers you had no control over. A goldmine for hackers and a business model for the data brokers who figured out how to stitch it all back together and sell it.</p><h3><strong>The Numbers</strong></h3><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!MRTH!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!MRTH!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 424w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 848w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 1272w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!MRTH!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png" width="1200" height="630.9559939301972" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/fb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:693,&quot;width&quot;:1318,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1387225,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192425282?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe7d35546-f239-43b3-9990-a8479c60e5ad_1408x768.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!MRTH!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 424w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 848w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 1272w, https://substackcdn.com/image/fetch/$s_!MRTH!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ffb163dbd-c8d9-415f-93e3-96817fe316e9_1318x693.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3><strong>The Timeline</strong></h3><p>The history of identity is a story of bolting on solutions, with each new layer creating new problems. It unfolds in three acts.</p><p><strong>Era 1: The Gatekeepers (Pre-2016)</strong></p><p>Before 2016, identity was a free-for-all. Every company rolled their own login system. The dominant players weren&#8217;t tech companies; they were the old-guard credit bureaus &#8212; TransUnion, Experian, Equifax. They quietly built empires by aggregating offline data and selling it back to businesses as a proxy for trust. Their business was, and still is, data arbitrage. They profited from the fragmentation because they were the only ones with a semi-complete picture.</p><p>Startups in this era mostly tinkered with Single Sign-On (SSO), trying to create a master key for all your digital locks. But adoption was slow. The problem wasn&#8217;t seen as urgent enough to solve.</p><ul><li><p><strong>Why it ended:</strong> A European regulation lit a fire under the entire industry.</p></li><li><p><strong>What it means for you today:</strong> The DNA of this era persists. The incumbents still hold the most data, and the default behaviour for any new app is still to build its own siloed identity system.</p></li></ul><p><strong>Era 2: The Compliance Gold Rush (2016-2024)</strong></p><p>The General Data Protection Regulation (GDPR) changed the game. Suddenly, mishandling user data wasn&#8217;t just bad practice; it was a business-ending liability with massive fines.</p><p>Fear became the product.</p><p>This shift created companies like <strong>OneTrust</strong>. Founded in 2016, it was perfectly timed. It didn&#8217;t sell a better user experience. It sold certainty to corporate lawyers. It was a compliance engine, a suite of tools to manage consent, map data flows, and generate reports to keep regulators happy. It worked. OneTrust grew into a multi-billion dollar company by focusing entirely on the enterprise&#8217;s pain, not the user&#8217;s.</p><p>The playbook was simple: find a regulatory threat and sell the antidote. The low-hanging fruit of compliance has been picked. Building another cookie-banner tool is a death wish.</p><ul><li><p><strong>Why it ended:</strong> Compliance became table stakes, and a new technological wave &#8212; AI &#8212; introduced a level of complexity that basic compliance tools couldn&#8217;t handle.</p></li><li><p><strong>What it means for you today:</strong> Regulation is the most powerful catalyst in this market. If you&#8217;re not tracking global data and AI policy, you&#8217;re flying blind.</p></li></ul><p><strong>Era 3: The AI Reckoning &amp; Consolidation (2025-Now)</strong></p><p>I used to think AI would disrupt the old data gatekeepers. Turns out, it&#8217;s their new favourite weapon. Incumbents like TransUnion and Experian are aggressively rebranding as &#8220;AI-driven powerhouses,&#8221; using machine learning to &#8220;weaponize proprietary data.&#8221; Their revenue is growing, not shrinking. They&#8217;re using AI to strengthen their grip, offering more sophisticated fraud detection and credit scoring than ever before.</p><p>This forces the market to evolve. The problem is no longer just <em>privacy compliance</em>; it&#8217;s <em>AI governance</em>. How do you prove your AI models aren&#8217;t biased? How do you manage the data they&#8217;re trained on?</p><p>Simultaneously, the market is consolidating. Fast. The cybersecurity space saw a record $96 billion in M&amp;A in 2025. Standalone privacy and identity features are being absorbed into bigger platforms. <strong>Securiti AI</strong>, a data governance startup, was acquired by infrastructure giant Veeam for $1.73 billion in 2025. It became a feature.</p><ul><li><p><strong>What it means for you today:</strong> Competing with incumbents on data is a losing battle. Your advantage has to be in building integrated, trustworthy infrastructure that they can&#8217;t &#8212; because their business models and legacy tech are too siloed.</p></li></ul><p></p><p></p><p></p><h3><strong>The Graveyard</strong></h3><p>Failure in this space often looks like a successful acquisition. Standalone companies get absorbed into larger platforms, becoming a feature, not a business. The lesson isn&#8217;t that they failed, but that building a venture-scale, independent identity company is brutally hard.</p><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!9k6c!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!9k6c!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!9k6c!,w_2400,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png" width="1200" height="654.5454545454545" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:false,&quot;imageSize&quot;:&quot;large&quot;,&quot;height&quot;:768,&quot;width&quot;:1408,&quot;resizeWidth&quot;:1200,&quot;bytes&quot;:1035285,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192425282?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:&quot;center&quot;,&quot;offset&quot;:false}" class="sizing-large" alt="" srcset="https://substackcdn.com/image/fetch/$s_!9k6c!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 424w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 848w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 1272w, https://substackcdn.com/image/fetch/$s_!9k6c!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F2c955006-ca7b-44e8-b75f-69a7cd997be7_1408x768.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3><strong>Where The Money Sits</strong></h3><p>The money sits with the gatekeepers: TransUnion, Experian, Equifax. They hold the foundational datasets on billions of people. They profit from complexity, selling access, verification, and analytics back to every other business. Their moat is a labyrinth of proprietary data, regulatory capture, and enterprise sales teams that have been in place for decades.</p><p>They are not just data brokers anymore. They are AI companies. TransUnion&#8217;s revenue grew 13% in 2025. They are successfully layering AI on top of their core asset &#8212; data &#8212; to create new, high-margin products.</p><p>The tax everyone pays is the &#8220;trust tax.&#8221; Every time a business needs to verify a user, they pay one of these giants. Every time a new regulation appears, consultants and software vendors get paid to help navigate the complexity that the incumbents thrive on.</p><p>Their blind spot? They are slow, siloed, and their products are built for their own benefit, not the user&#8217;s. They are trying to bolt on privacy and modern identity, but their core business is data exploitation. That contradiction is the opportunity.</p><h3><strong>What&#8217;s Actually Hard</strong></h3><p>The tech is not the hardest part. Zero-knowledge proofs, decentralized identifiers, biometrics &#8212; the building blocks are there.</p><p>The hard part is the sales cycle.</p><p>You are not selling software. You are selling trust. You are selling a promise that your system is more secure, more compliant, and more reliable than the status quo. Your buyer isn&#8217;t a product manager looking for a cool feature. Your buyer is a Chief Information Security Officer or a General Counsel.</p><p>Their first question isn&#8217;t &#8220;How does it work?&#8221;. It&#8217;s &#8220;Who do we sue if this goes wrong?&#8221;.</p><p>You are asking enormous, risk-averse organisations to rip out a core piece of their infrastructure &#8212; how they know who their customers are &#8212; and replace it with yours. This involves navigating procurement hell, legal reviews, and security audits that can take 18 months or more. You don&#8217;t win with a better demo. You win with better liability terms, better compliance certifications, and a story about risk reduction that the board can understand.</p><h3><strong>What&#8217;s Shipping Now</strong></h3><p>The most interesting work today is happening at the infrastructure layer, with founders who have learned the lessons of the past. They&#8217;re not building another social network with a new login. They&#8217;re building the pipes.</p><ul><li><p><strong>Killing the Password:</strong> Startups are finally tackling the original sin. <strong>Beyond Identity</strong> is pushing passwordless authentication using device-bound cryptographic keys, making phishing nearly impossible. <strong>1Kosmos</strong> combines biometrics with a private blockchain to prove identity without storing centralised credentials.</p></li><li><p><strong>Building Reusable Identity:</strong> The dream of a single, portable identity is getting closer. <strong>ID.me</strong> has carved out a huge niche with government agencies, offering a reusable login for accessing public services. On the decentralized front, <strong>Civic</strong> and <strong>SelfKey</strong> are using blockchain to build user-controlled identity systems.</p></li><li><p><strong>Delivering Dev Tools:</strong> The smartest founders know that developers choose the next generation of infrastructure. <strong>uPort</strong>, now evolved into projects like <strong>Veramo</strong>, provides open-source tools for building with decentralized identity. In Web3, <strong>Privy</strong> is giving developers simple APIs to embed wallets and authentication into their apps, abstracting away the complexity.</p></li></ul><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!mTsV!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!mTsV!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 424w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 848w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 1272w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!mTsV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png" width="768" height="1376" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/b1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:1376,&quot;width&quot;:768,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:1289395,&quot;alt&quot;:null,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/192425282?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" srcset="https://substackcdn.com/image/fetch/$s_!mTsV!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 424w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 848w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 1272w, https://substackcdn.com/image/fetch/$s_!mTsV!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fb1bc1536-35fd-4f8e-aa22-915a9e12a07a_768x1376.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p></p><h3><strong>Where This Goes</strong></h3><p>The noise and fragmentation are signs of an underlying convergence. Privacy, identity, and AI governance are not three separate markets. They are becoming one single problem: how do you establish and manage trust in a digital world?</p><p>The next 36 months will be a race to build the integrated &#8220;trust infrastructure&#8221; that solves this.</p><p>The bet is that a startup, unburdened by legacy tech and siloed business units, can build this unified platform faster and better than the incumbents who are trying to bolt these pieces together. The winner won&#8217;t just offer passwordless login, or a consent dashboard, or an AI governance report. They will offer a single platform that manages the entire lifecycle of digital trust for an enterprise.</p><p>This is an infrastructure play. The exits will look less like consumer apps and more like the $96 billion in cybersecurity M&amp;A we saw in 2025 &#8212; strategic acquisitions by major cloud and security players.</p><h3><strong>Moves Worth Stealing</strong></h3><ol><li><p><strong>Ride the Regulatory Wave (The OneTrust Play):</strong> Don&#8217;t wait for demand to appear. Find the next GDPR-level event &#8212; a new AI liability law, a federal privacy bill &#8212; and build the tool that makes compliance automatic. Sell certainty to lawyers.</p></li><li><p><strong>Become Essential Infrastructure (The Securiti AI Play):</strong> Don&#8217;t try to be everything. Build one critical component of the trust stack so well that a larger platform like Veeam, Microsoft, or Salesforce has no choice but to acquire you. Be the definitive solution for one piece of the puzzle.</p></li><li><p><strong>Solve the Original Pain (The Beyond Identity Play):</strong> Go back to first principles. The password is a 60-year-old hack. Find the root of the problem that everyone else is just managing and build the thing that eliminates it entirely.</p></li></ol><h3><strong>The Why Now</strong></h3><p>Three forces are creating a brief window of opportunity.</p><ol><li><p><strong>AI Creates New Urgency:</strong> The explosion of AI makes data governance and identity verification a C-suite priority. The old tools aren&#8217;t equipped to manage AI models, training data, and the new attack vectors they create.</p></li><li><p><strong>Regulation is Expanding:</strong> The compliance wave that started with GDPR is now global and expanding to include AI. This creates new, mandatory budget items inside every large company.</p></li><li><p><strong>The Incumbents are Buying:</strong> The record-breaking M&amp;A in cybersecurity shows that the big players are acquiring innovation, not building it. This provides a clear path to exit for startups that can build something valuable.</p></li></ol><p>The convergence of these forces means that for the first time, there is both the budget and the urgency to finally fix the internet&#8217;s broken identity layer.</p><h3><strong>If You&#8217;re Starting Monday</strong></h3><ul><li><p><strong>The Wedge:</strong> Don&#8217;t build a generic identity platform. Pick a single, painful, high-stakes verification problem in one regulated industry. Example: proving the identity of participants in decentralised clinical trials for a pharma company.</p></li><li><p><strong>The Trap:</strong> Building a consumer-first &#8220;digital wallet&#8221; for identity. You will burn millions trying to solve a two-sided market problem and get crushed by platform players. Sell to the enterprise. They have the problem and the money.</p></li><li><p><strong>The Customer:</strong> The Chief Information Security Officer (CISO) or the Chief Compliance Officer. The person who gets fired when a data breach or an AI audit goes wrong.</p></li><li><p><strong>The First Hire:</strong> Your first hire should not be a growth hacker. Your first hire should be a lawyer who has spent years advising companies on data privacy and AI ethics. Your product is not just code; it&#8217;s a manifestation of legal and ethical rules.</p></li></ul><h3><strong>Go Deeper</strong></h3><ul><li><p>Momentum Cyber 2025 Report: Analysis of the record $96B in cybersecurity M&amp;A. (<strong><a href="https://momentumcyber.com/wp-content/uploads/2026/01/Cybersecurity-Almanac-Momentum-Cyber-2025-Year-End-Report-1-1.pdf">Link</a></strong>)</p></li><li><p>The Information on OneTrust: The inside story of the compliance gold rush and market consolidation. (<strong><a href="https://www.theinformation.com/articles/onetrust-privacy-startup-last-valued-at-4-5-billion-discusses-private-equity-sale">Link</a></strong>)</p></li><li><p>TransUnion 2025 Earnings: How incumbents are using AI to drive growth. (<strong><a href="https://newsroom.transunion.com/transunion-announces-strong-fourth-quarter-and-full-year-2025-result/">Link</a></strong>)</p></li><li><p>Rothschild &amp; Co on AI Investment: Data on the 2025 venture capital boom driven by AI. (<strong><a href="https://www.rothschildandco.com/siteassets/publications/rothschildandco/global_advisory/2025/growth_equity_update/05/en_ga_growth_equity_update_edition_38.pdf">Link</a></strong>)</p></li><li><p>Contrary Research on OneTrust: A breakdown of the platform serving 75% of the Fortune 100. (<strong><a href="https://research.contrary.com/company/onetrust">Link</a></strong>)</p></li></ul><p>For the &#10084;&#65039; of startups</p><p></p><div><hr></div><div class="captioned-image-container"><figure><a class="image-link image2 is-viewable-img" target="_blank" href="https://substackcdn.com/image/fetch/$s_!sxpE!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png" data-component-name="Image2ToDOM"><div class="image2-inset"><picture><source type="image/webp" srcset="https://substackcdn.com/image/fetch/$s_!sxpE!,w_424,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 424w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_848,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 848w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_1272,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 1272w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 1456w" sizes="auto, 100vw"><img src="https://substackcdn.com/image/fetch/$s_!sxpE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png" width="831" height="387" data-attrs="{&quot;src&quot;:&quot;https://substack-post-media.s3.amazonaws.com/public/images/bef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png&quot;,&quot;srcNoWatermark&quot;:null,&quot;fullscreen&quot;:null,&quot;imageSize&quot;:null,&quot;height&quot;:387,&quot;width&quot;:831,&quot;resizeWidth&quot;:null,&quot;bytes&quot;:40187,&quot;alt&quot;:&quot;&quot;,&quot;title&quot;:null,&quot;type&quot;:&quot;image/png&quot;,&quot;href&quot;:null,&quot;belowTheFold&quot;:true,&quot;topImage&quot;:false,&quot;internalRedirect&quot;:&quot;https://insights.fusion-42.com/i/181702527?img=https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png&quot;,&quot;isProcessing&quot;:false,&quot;align&quot;:null,&quot;offset&quot;:false}" class="sizing-normal" alt="" title="" srcset="https://substackcdn.com/image/fetch/$s_!sxpE!,w_424,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 424w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_848,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 848w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_1272,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 1272w, https://substackcdn.com/image/fetch/$s_!sxpE!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fbef954cf-4a50-4da8-bf5d-299326ff9e3f_831x387.png 1456w" sizes="auto, 100vw" loading="lazy"></picture><div class="image-link-expand"><div class="pencraft pc-display-flex pc-gap-8 pc-reset"><button class="pencraft pc-reset pencraft icon-container restack-image buttonBase-GK1x3M" tabindex="0" type="button"><svg aria-hidden="true" width="20" height="20" viewBox="0 0 20 20" fill="none" stroke-width="1.5" stroke="var(--color-fg-primary)" stroke-linecap="round" stroke-linejoin="round" xmlns="http://www.w3.org/2000/svg" class="icon-noB79L"><g><path d="M2.53001 7.81595C3.49179 4.73911 6.43281 2.5 9.91173 2.5C13.1684 2.5 15.9537 4.46214 17.0852 7.23684L17.6179 8.67647M17.6179 8.67647L18.5002 4.26471M17.6179 8.67647L13.6473 6.91176M17.4995 12.1841C16.5378 15.2609 13.5967 17.5 10.1178 17.5C6.86118 17.5 4.07589 15.5379 2.94432 12.7632L2.41165 11.3235M2.41165 11.3235L1.5293 15.7353M2.41165 11.3235L6.38224 13.0882"></path></g></svg></button><button class="pencraft pc-reset pencraft icon-container view-image buttonBase-GK1x3M" tabindex="0" type="button"><svg xmlns="http://www.w3.org/2000/svg" width="20" height="20" viewBox="0 0 24 24" fill="none" stroke="currentColor" stroke-width="2" stroke-linecap="round" stroke-linejoin="round" class="lucide lucide-maximize2 lucide-maximize-2 icon-noB79L"><polyline points="15 3 21 3 21 9"></polyline><polyline points="9 21 3 21 3 15"></polyline><line x1="21" x2="14" y1="3" y2="10"></line><line x1="3" x2="10" y1="21" y2="14"></line></svg></button></div></div></div></a></figure></div><p><em>Arthur is the AI native startup operating system I&#8217;m building in public &#8212; not hype, but a system that turns input into structured execution and tracks founder progress. If you want to follow or use it, it&#8217;s open for early access.</em><br><strong><a href="https://arthur.fusion-42.com/">&#8594; Access Arthur</a></strong></p><p></p><p><em>Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.<br><br>If you&#8217;ve got suggestions, an article, research, your tech stack, or a job listing you want featured, just let me know! I&#8217;m keen to include it in the upcoming edition.</em></p><p>Please let me know what you think of it, love a feedback loop &#128591;&#127996;</p><ul><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#10084;&#65039; Love it, keep them coming!</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128077; Great, always a good read.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128522; Enjoyable, but let&#8217;s see some variety.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128078; Okay, seen better.</a></p></li><li><p><a href="https://airtable.com/appUpqq3uT1iM4xUD/pagUI0CWQKcm3Y3Di/form">&#128721; Get a different job.</a><br><br><em>Subscribe and follow me on <a href="https://www.linkedin.com/in/derekwatsonf42/">LinkedIn</a> or <a href="https://twitter.com/derek__Watson">Twitter</a> to never miss an update.</em></p></li></ul><div class="subscription-widget-wrap-editor" data-attrs="{&quot;url&quot;:&quot;https://insights.fusion-42.com/subscribe?&quot;,&quot;text&quot;:&quot;Subscribe&quot;,&quot;language&quot;:&quot;en&quot;}" data-component-name="SubscribeWidgetToDOM"><div class="subscription-widget show-subscribe"><div class="preamble"><p class="cta-caption">Insights by Fusion42 is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.</p></div><form class="subscription-widget-subscribe"><input type="email" class="email-input" name="email" placeholder="Type your email&#8230;" tabindex="-1"><input type="submit" class="button primary" value="Subscribe"><div class="fake-input-wrapper"><div class="fake-input"></div><div class="fake-button"></div></div></form></div></div>]]></content:encoded></item></channel></rss>