OpenAI and Alpöge-Buckmaster both leaned on Lean for Navier-Stokes claims
fortnow · x · 2026-09-10
Lance Fortnow notes both the OpenAI and Alpöge-Buckmaster Navier-Stokes announcements leaned heavily on Lean: OpenAI fully formulated its results in Lean, while Buckmaster's team has Lean-verified proofs for three public results but held back the blowup result pending verification. The post reviews Lean's history since Leonardo de Moura's 2013 Microsoft project, its backward-compatibility pains, and big formalization projects (Hales' Kepler proof, Scholze's liquid tensor experiment), asking whether Lean is becoming a new requirement for publishing.
Related event: OpenAI's Navier-Stokes Proof Verified in Lean Within 17 Hours(3 posts)→
More from AGI Musings
- Kelsey Piper: Labs plan to automate AI R&D with AI, shrinking human oversight within two years — round · 2026-09-11
- 100-agent experiment: when 9% of AI agents cheated, 24% blew the whistle on peers — jzl86 · 2026-09-11
- zetalyrae: extensional definitions of alignment only work retrospectively — zetalyrae · 2026-09-11
- "AI won't destroy humanity": Y2K analogy sparks AGI-doom debate — granawkins · 2026-09-11
- Economist Warns Grammarly Is Nudging Students to Outsource Thinking, Becoming Next Chegg — paulnovosad · 2026-09-11
- Scott Belsky: trust, progressive personalization and agent-to-agent effects define the 'first mile' — _AustinCalvert_ · 2026-09-11