Three mathematicians say OpenAI's Navier-Stokes claim dodges the problem rather than solving it
AnkaReuel · x · 2026-09-22
OpenAI recently claimed to have solved the Navier-Stokes problem, one of math's biggest open questions — a claim some AI safety researchers cited as evidence that AI is advancing beyond human capacity. Per Scientific American, three mathematicians have now shown that OpenAI's work "dodges the question rather than solving it."
The poster draws a broader lesson: never accept vendor claims until third parties validate them, and warns that over-trusting model outputs — a real phenomenon he calls AI psychosis — won't help us understand actual AI risks.
More from Companies & People
- Eric Schmidt rejects AI pause: it defies incentives and can't be verified — mark_k · 2026-09-22
- Codos launches 'virtual Chief AI Officer', claims 21% FTE capacity freed at fintech — rohanpaul_ai · 2026-09-22
- Cursor's Composer 1 powers a new AI-native coding interview: build a real project in 1 hour — steipete · 2026-09-22
- CS teachers struggle with AI; Pamela Fox says teach protocols and system design — lee_stott · 2026-09-22
- Inside a large Chinese teleoperation and robot data-collection startup — AIFlow_ML · 2026-09-22
- Steve Yegge floats 'Agentic TPMs' as the fastest path for coding agents into enterprises — Steve_Yegge · 2026-09-22