Here’s your Monday dose of The AI Brief.
Your weekly updates of game-changing AI breakthroughs, hands-on startup case studies, velocity-boosting tools and no-nonsense strategic reminders, all designed to give founders the insights and tactics they need to stay ahead in an AI-driven market.
📈 Trending Now
The week’s unmissable AI headlines.
What’s buzzing in the AI world, the key facts behind it and why every founder should sit up and take notice.
💡 Innovator Spotlight
Meet the change-makers.
A down-to-earth snapshot of a founder using AI in a brilliantly unexpected way—and one practical lesson you can pinch straight away.
🛠️ Tool of the Week
Your speed-boost in a nutshell.
An AI tool or integration that’ll shave days off your dev cycle—plus how to get started in minutes and our top insider tip.
📌 Note to Self
Words above my desk.
A hard-won reminder or question to pose yourself—keep these front and centre to dodge the same old traps.
Weekly AI insights in 5 minutes flat
📈 Trending Now
Major Headlines
OpenAI secures $200 m U.S. defence contract
Microsoft & Premier League ink five-year AI partnership
Apple mulls ChatGPT/Claude for next-gen Siri
Top Trends
Agentic AI – multi-step reasoning with end-to-end task execution
AI in Sport & Entertainment – real-time analytics & personalised fan experiences
Gov & Defence AI – OpenAI’s Pentagon award kick-starts public-sector deals
Generative Reasoning Models – DeepSeek R1, Claude Opus 4 bring stepwise problem-solving
🛰 On the Radar
DeepSeek R1 – China’s cost-effective challenger heats up model competition
London Tech Week ’25 – £1 billion AI pledge underscores UK infrastructure push
Healthcare AI leap – Microsoft’s diagnostic system outperforms expert physicians
🔭 What to Watch
A clear shift from “features” to outcomes and “code” to adaptive integration. Keep an eye on AI agents that embed in messy workflows—and on partnership strategies that marry foundation models with domain expertise.
Services-as-Software -SAS
I wrote about SAS in November of last year, after having written about the death of the traditional SaaS model in Jan 24. (see articles below). Now, a fresh wave of AI-native founders is building end-to-end intelligent agents that don’t just automate tasks but deliver complete business outcomes, from SRE teams to accounting and legal support. In an era when every company can licence the same base models—GPT-4, Claude, Gemini—the true moat lies in implementation depth, embedding forward-deployed engineers to capture tribal knowledge and turn edge-case rules into reusable assets. This shift unlocks a$4.6 trillion opportunity , dwarfing traditional SaaS, and demands a relentless focus on the “last mile” of messy, human workflows.
Who
A new cohort of AI-first teams is emerging, building Services-as-Software companies—complete intelligent agents that replace entire functions such as site reliability engineering, accounting and paralegal work.
What’s bold
These startups aren’t merely shipping features; they’re delivering proven outcomes. In today’s world, where foundation models are universally accessible, the real competitive edge comes from how you integrate, customise and operate AI within complex, real-world processes. Leading teams embed forward-deployed engineers alongside users to codify tribal knowledge, turning every bespoke deployment into a product advantage for the next customer
Why it’s unexpected
The old SaaS playbook—feature velocity, proprietary code, slick UX—is on its way out. Instead, the moat now lies in implementation depth and the ability to prove business impact before contracts are signed, blurring the line between pre-sales and post-sales.
The scale of the opportunity
£1.8 trillion spent on salaries across sales, marketing, software engineering, security and HR each year
£1.8 trillion more spent on outsourced services, per Gartner
In the US alone, services roles account for roughly £3.9 trillion in wages across AI-exposed functions—highlighting an enormous under-automated market.
By comparison, global SaaS revenue hovers around £226 billion annually, with outsourced digital work at £1.5 trillion.
Practical tip
Focus on the “last mile”—the messy workflows where generic models falter. Design your implementation team as a strategic asset, not an afterthought, and build reusable scaffolding so each customer deployment compounds your competitive advantage.
Why founders should care
The next generation of unicorns will win not by cranking out features but by mastering deep, adaptive integration—turning AI promises into provable, repeatable business results. With analysts projecting that AI-driven services could boost global productivity by £5.5 trillion by 2030, the real prize lies in outcome-focused agents, not toolkits.
🚀 Service as Software (SaS)
🔄 The Evolution of SaaS into a Fully Automated Service as Software (SaS) Model
🚀 AI Launchpad • 4‐Week Live Programme
Go from GenAI fundamentals to full-scale automations—live masterclasses, hands-on builds and reusable blueprints. All paid registrants receive our ChatGPT Starter Playbook.
Who: Executives & founders
When: Four Wednesdays, 19:00–21:00 GMT+4 (16 Jul–6 Aug ’25)
Why join?
Automate key workflows & save hundreds of hours
Build-first demos with zero waffle
💡Innovator Spotlight
Pika: The AI Video Wizard Making Hollywood Jealous
Who
Demi Guo and Chenlin Meng—both Stanford AI drop-outs—launched Pika to turn every browser into a mini film studio.
What’s funky
Half a million creators now whip up millions of videos a week—whether it’s scuba divers on alien reefs or tigers break-dancing in Times Square—with no GPU farms in sight.
Why it’s unexpected
Until Pika, you needed expensive render farms or awkward enterprise tools; now it’s just “click → watch” in any modern browser.
Pinch this lesson
Strip away friction—hardware, jargon, loading bars—and you open your product to everyone, from indie filmmakers to your gran.
Practical tip
Treat that final “Generate” click as sacred. When users hit it, everything else must vanish, leaving only delight.
Why founders should care
The next AI unicorns won’t be named for model size but for the joy they deliver. By turning users into creators, you build engagement and defensibility far beyond the core algorithm.
AI Avatars & Virtual Influencers: Building Brands and Getting Paid in 2025
The big picture
AI avatars and virtual influencers now form a $4.6 billion market, growing at 26% a year—no flash-in-the-pan trend, but a full-blown industry.
Top earners
Lil Miquela is on track for $15–20 million in 2025, fronting Prada, Balenciaga and even metaverse concerts.
Shudu Gram pulls in $10–12 million, thanks to campaigns for Fendi and Balmain plus NFT art drops.
Imma, Tokyo’s pink-haired icon, nets $8–10 million with Dior, Puma and IKEA tie-ups.
Livestream shocker
On Baidu’s Youxuan, AI doubles of top host Luo Yonghao generated $7.65 million in six hours—outperforming his human-hosted streams.
Fashion & beauty highlight
Mango’s teen line was the first campaign ever created entirely with generative AI—no human models needed.
Why brands are all in
Creative control: zero scandals, total localisation.
Scalability: avatars never sleep or burn out.
Engagement: Gen Z can’t get enough of digital storytellers.
Key takeaway
Virtual influencers aren’t a gimmick—they’re a thriving, outcome-focused business. Blend human warmth with AI reliability, and you’re onto something powerful..
🛠️ Tools of the Week
How AI Avatar Content Automation Really Works
1. Content Database
Use a simple spreadsheet or workspace as your single source of truth:
Google Sheets (free, cloud-based) keeps scripts, captions, hashtags and video URLs in one place.
Notion lets you build a wiki-style repo with templates, comments and version history.
Airtable gives you a no-code database view, Kanban boards or galleries—ideal for richer metadata.
2. AI Video Generation
Automatically turn text into talking-head or character clips in the browser:
HeyGen: Browser-based avatars with multilingual voice cloning (API available).
Synthesia: 230+ avatars, 140+ languages—just send your script to their API and get back an MP4 .
D-ID: Photorealistic AI presenters straight from a single image—ideal for lifelike spokespeople.
Pika Labs: Instant text-to-video loops, no GPU farms—just paste your prompt and watch the magic.
3. Automated Social Posting
Orchestrate the whole flow—database → AI tool → social networks—without code:
n8n: Fair-code, self-hostable or cloud, with 400+ integrations and native AI nodes.
Make.com: Visual canvas, AI-powered agents and 2,000+ app connectors, all drag-and-drop.
Zapier: 8,000+ apps, prebuilt templates and AI orchestration—just link your Sheets to Instagram, TikTok, LinkedIn, YouTube, you name it.
Why this matters
• Hands-off scaling: Once it’s wired up, new scripts become new videos and posts overnight.
• Total consistency: One source of truth means no mixed-up hashtags, captions or forgotten links.
• End-user joy: Spend time on ideas, not editing or scheduling.
Give it a whirl—your next viral avatar video could be fully automatic.
Check out our content store at insights.fusion-42.com
Unlock the Full Fusion42 Content Store
Dive into our all-in-one knowledge hub for founders, framers and funders:Everything42 – Your weekly “Life, the Universe and Fusion42” deep dive on big-picture trends, live cohort metrics and candid build-in-public updates.
Startup-X – The AI Brief every Monday: Trending Now, Innovator Spotlight, Tool of the Week and a Note to Self, all in a rapid-fire five-section format.
Investor-X – The Raise Report each Wednesday: tracking who’s raising, who’s deploying—and why. Since launch, we’ve covered 700 + funds that raised over USD 240 billion in fresh capital. Money isn’t the issue; understanding why deployment lags is.
Masterclass Insights – On-demand replays of every Global Super Accelerator workshop, Raise Summit panel and candid Q&A—complete with slides, demos and real-time pivots.
eBooks for Founders – Bite-sized, no-fluff playbooks on everything from framing your problem and sizing your market to term-sheet traps and retention strategies—distilled from our own accelerator cohorts.
Ready to plug in to every guide, workshop and insider update?
Subscribe to Fusion42+ at insights.fusion-42.com and spend less time hunting answers and more time building what matters.
📌 Note to Self
Thank you for reading. If you liked it, share it with your friends, colleagues and everyone interested in the startup Investor ecosystem.
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.
Please let me know what you think of it, love a feedback loop 🙏🏼
🛑 Get a different job.
Subscribe below and follow me on LinkedIn or Twitter to never miss an update.
For the ❤️ of startups
✌🏼 & 💙
Derek