YC is a barometer for the startup world and this year I’ve been following closely their lists of “Requests for Startups” or RFS. You can find the previous lists from YC and other sources in part 1, part 2 and part 3.
Now, in the Fall 2025 edition, these are some of the key areas where YC sees immense potential:
1. Retraining the Workforce for the AI Economy
As the physical infrastructure for AI — think data centers and semiconductor fabs — is built out, we face a significant shortage of skilled tradespeople like electricians, HVAC technicians, and welders. The government’s new AI Action Plan is creating a “worker-first” agenda, directing departments to fund rapid retraining programs for these essential physical labor jobs.
This opens a huge opportunity for startups to build a new kind of vocational school for the AI economy. Imagine using AI to create personalized training programs that get people job-ready in months, not years. While teaching hands-on skills like welding can’t be done via a keyboard, multimodal AI offers solutions: a voice AI could coach individuals, or AR/VR simulations with AI tutors using vision models could provide practice and feedback. The business model is clear: employers will pay to hire these well-trained workers, and unlike traditional coding bootcamps, AI can scale the quality of instruction infinitely. This is a chance to build a huge business that empowers everyone to participate in the new AI economy.
2. Generative Video: The New Digital Primitive
Video generation models are rapidly advancing, producing photorealistic, sound-on clips that are often indistinguishable from reality at a marginal cost approaching zero. YC believes that video will become a new basic building block for software.
This fundamental shift unlocks a wealth of new ideas:
- Media and Entertainment: Imagine creating new seasons of canceled TV series, personalized kids’ cartoons starring your family, or an AI-native successor to TikTok where every video is unique to a single viewer.
- Shopping: See yourself wearing clothes online or have your furniture auto-staged in apartment listings.
- Gaming and Simulation: Developing video games without traditional game engines or creating APIs that return infinite robotic training data.
- Personal Connections: Even having video calls with loved ones long after they’re gone.
YC is looking for founders who treat generative video as a new computing primitive, not just an output, building novel apps, tooling, and infrastructure for a world of limitless, low-latency video.
3. Lean, Mean, AI-Powered Machines: The Rise of $100B Startups
New AI tools are making it possible for small, high-agency teams – even solo founders – to build multi-billion dollar companies with significantly less capital, potentially with as little as $500k in funding. This echoes the impact of cloud computing, which removed the need for massive infrastructure spending.
The best startups of the future will optimize for revenue per employee. These smaller, efficient teams have inherent advantages over bloated incumbents, avoiding the pitfalls of politics, excessive meetings, and lack of focus. They can instead concentrate on winning through superior speed and execution. YC is eager to fund these founders to help them build the first 10-person, $100 billion company.
4. Taming Complexity: Infrastructure for Multi-Agent AI
The evolution of AI agents is leading to distributed workflows where single runs fan out into many sub-agent calls. These multi-agent systems are proving useful for complex tasks, from long-running workflows to “agentic mapreduce jobs” that apply human-level judgment to filter and search vast amounts of data in parallel.
However, building these systems is challenging. They involve traditional distributed systems problems like ensuring high throughput and reliability while controlling costs, alongside new, higher-level issues such as writing effective prompts, handling untrusted context, and monitoring and debugging agents. YC is looking for founders who have experienced these pains in production and want to build the tools that make these multi-agent systems easier to build and maintain, making operating “fleets of agents as routine and reliable as deploying a web service”.
5. Reinventing Enterprise Software with AI at its Core
Just as cloud computing gave rise to giants like Salesforce and ServiceNow by enabling 10x better products and catching incumbents off guard, AI presents the same once-in-a-generation opportunity for today’s founders.
Tomorrow’s enterprise software won’t just be systems of record; they will have AI embedded deeply and thoughtfully throughout, helping employees perform their work faster and more accurately. Think of it as a “Cursor for sales, HR, and accounting”. Incumbents will struggle to rebuild their existing products around this new technology, providing startups with the crucial time needed to gain a winning edge.
6. Disrupting Government Consulting with LLMs
The U.S. government spends over $100 billion a year on consulting, an area ripe for disruption due to its inefficiency and reliance on often-poor custom software. With political pressure to cut wasteful spending and the significant advancements in Large Language Models (LLMs), there’s a strong belief that this sector is poised for change.
LLMs are already capable of performing many tasks traditionally handled by consulting firms. YC has already funded companies helping businesses navigate government approvals like FedRAMP, and others using LLMs to help the government cut regulation and ensure legal compliance for new policies. There’s a vast amount of work currently done by major government consulting firms like Deloitte and Accenture that startups can now tackle by building LLM-powered software.
As I mentioned, this is a barometer, not a compass, but the message is clear. YC (and the broader market) are looking to fund fast growing, lean, AI first companies that disrupt the old way things are done. But come to think of it, wasn’t it always the case?
At Remagine Ventures, we write first cheques for Israeli founders building the future with AI. It’s never too early to get in touch and you don’t need a warm intro.
I'm a former general partner at google ventures, head of Google for Entrepreneurs in Europe and founding head of Campus London, Google's first physical hub for startups.
I'm also the founder of Techbikers, a non-profit bringing together the startup ecosystem on cycling challenges in support of Room to Read. Since inception in 2012 we've built 11 schools and 50 libraries in the developing world.
- Weekly Firgun Newsletter – March 6 2026 - March 6, 2026
- Weekly Firgun Newsletter – February 27, 2026 - February 27, 2026
- The Rise of the Personal AI Agent - February 26, 2026
