Y Combinator JUST announced what startups they want to fund next in 2025.

They want 10-person companies worth $100 billion (and more).

AI-Powered Vocational Training The Problem: AI infrastructure needs skilled tradespeople (electricians, welders, HVAC techs), but training takes years The Opportunity: Use multimodal AI, AR/VR to train workers in months, not years Revenue Model: Employers pay for well-trained workers; AI teachers scale infinitely
Video Generation as Computing Primitive The Reality: Google's Veo 3 creates photorealistic 8-second clips for dollars; costs approaching $0 Applications: Personalized content (TikTok for one viewer, canceled TV show revivals) Commerce (try-on experiences, auto-staged apartments) Gaming (no game engines needed, infinite robotic training data) Focus: Build infrastructure for limitless, low-latency video
The 10-Person, $100B Company The Shift: Like cloud computing 15 years ago, AI eliminates need for large teams Key Metric: Revenue per employee becomes critical optimization target Advantages: No politics, fewer meetings, faster execution vs. bloated incumbents Goal:$ 100B company with just $500K in YC funding
Multi-Agent System Infrastructure Evolution: AI agents moving from single-threaded to distributed workflows with thousands of sub-agents Challenges: Traditional distributed systems problems plus new issues (prompt engineering, monitoring, untrusted context) Opportunity: Make operating agent fleets as routine as deploying web services
AI-Native Enterprise Software Historical Pattern: Salesforce & ServiceNow (both $10B+ revenue,$ 200B+ market cap) rode cloud wave 25 years ago to beat incumbents Today's Opportunity: AI presents same once-in-generation chance Vision: "Cursor for sales, HR, and accounting" - AI embedded throughout, not just system of record Advantage: Incumbents will struggle to rebuild around AI
Replacing Government Consulting with LLMs Market Size: US government spends $100B+ annually on consulting (Deloitte, Accenture) Drivers: Political pressure to cut waste, LLMs can already do consulting work Examples: FedRAMP approval automation, regulation compliance checking Opportunity: Build LLM software to replace traditional consulting functions

What YC Wants: Deep domain expertise in specific verticals, Full automation, not just human assistance, Clear B2B revenue paths, Infrastructure/tools that help AI scale, Treating AI as primitive building block

Guest post by Greg

Author

Ai Base Network (ABN), ABN ASIA was founded by people with deep roots in academia, with work experience in the US, Holland, Hungary, Japan, South Korea, Singapore, and Vietnam. ABN Asia is where academia and technology meet opportunity. With our cutting-edge solutions and competent software development services, we're helping businesses level up and take on the global scene. Our commitment: Faster. Better. More reliable. In most cases: Cheaper as well.

Feel free to reach out to us whenever you require IT services, digital consulting, off-the-shelf software solutions, or if you'd like to send us requests for proposals (RFPs). You can contact us at [email protected]. We're ready to assist you with all your technology needs.