ACCESS The 'AI Coachella' prof wants to take back the data centers
19 snips
May 7, 2026 
Anjney Midha, venture capitalist, Stanford AI lecturer, and founder of AMP focused on compute infrastructure. He discusses cooperation across AI labs and how compute should work like electricity. He talks about his viral Stanford class culture, one-person frontier labs, and AMP’s plan for transparent, shared data-center capacity and ‘nutrition label’ style disclosure.
AI Snips
Chapters
Transcript
Episode notes
How Stanford Classes Shaped A VC Mindset
- Midha recounts taking Peter Thiel's CS183B class and later Sam Altman's YC-sponsored course, showing Stanford's lineage of startup instruction.
- He used those experiences to contextualize YC's institutional role and why students should respect its history.
Cooperation Could Multiply AI Progress
- Zero-sum thinking dominates frontier AI but cooperation across labs could produce far larger capability jumps.
- Midha imagines co-training world-class models together and borrowing biotech collaboration patterns to accelerate breakthroughs like cancer research.
A VC Who Cooperates With Competing Firms
- Midha describes investing in and working with many firms and founders, explaining his cooperative VC approach across funds and boards.
- He cites examples like leading seed rounds, being on seven boards, and incubating Periodic Labs with mixed investors.