Diffusion LM Best-Paper Author Dropped Out of Stanford PhD to Join OpenAI
aaron_lou · x · 2026-10-11
At COLM, SOTA TV interviewed Aaron Lou, who won the ICML Best Paper on diffusion LMs and then dropped out of his Stanford PhD to join OpenAI right at the o1 moment. They discussed autoregressive + speculative decoding vs. diffusion, and why he doesn't yet buy the fully self-improvement story. Full podcast coming soon.
More from Companies & People
- Ex-Apple engineer recalls demoing Siri automation QA tool to Craig Federighi — jdluk87 · 2026-10-12
- Waymo CEO Dmitri Dolgov: problem convergence tells you if shipping is one year or ten away — a16z · 2026-10-12
- Ex-NASA engineer hired by SpaceX at 62: more design iterations in 6 months than years at NASA — SumitGup · 2026-10-12
- Under 1% of the world pays for AI tools — software is not dead — sahilypatel · 2026-10-12
- LinkedIn lets users hide comments flagged as 'AI slop' — lilyraynyc · 2026-10-12
- 113 AI PM job listings: 69% demand AI experience, only 14 want production-level work — aakashgupta · 2026-10-12