OpenAI's willdepue defends agent race on Navier-Stokes: dual credit was goodwill, not foul play
BorisMPower · x · 2026-09-09
OpenAI's willdepue defended the company's agent race on the Navier-Stokes problem:
- His account: upon hearing 'Anthropic solved Navier-Stokes,' OpenAI threw a bunch of agents at it without knowing the other team's approach.
- The agents discovered the same result and more, and OpenAI still offered dual credit as a goodwill gesture — which he argues is totally above board.
The real crux, he says, is whether OpenAI had prior knowledge of Tristan Buckmaster's approach and seeded the model with it.
Related event: OpenAI Claims Navier-Stokes Breakthrough Amid Attribution Dispute(150 posts)→
More from Fun
- FreeBots indie AI bot world adds home decorating and owner claiming — Daniel_Farinax · 2026-09-09
- Trillions of tokens spilled analyzing a possibly wrong model of reality — MikePFrank · 2026-09-09
- In the ChatGPT Images 2.5 ad, a woman gets an AI-designed tattoo inked on her arm — yvgh233 · 2026-09-09
- User admits deliberately burning Claude's free tier to hurt Anthropic's investor numbers — justlikemedics · 2026-09-09
- 'Small towns of geniuses in a datacenter are here': the multi-agent cluster moment goes viral — austinc3301 · 2026-09-09
- Researcher jokes that Muse will 'totally not escape' its isolated VMs — suchenzang · 2026-09-09