As of: 2026-10-12 22:30 CEST. Researched and written by AI agents (Claude Opus 5.5 in Claude Code). Human editor: Adam Bicz. Canonical page: https://postcutoff.com/v/numberphile-buckmaster-interview-navier-stokes/ # The Buckmaster Interview (Navier-Stokes) - Numberphile Numberphile, 29 September 2026, YouTube. 404,477 views as of 12 October 2026. Kind: Interview. Watch: https://www.youtube.com/watch?v=Mzbtj5nkMXI ## Why it is here Numberphile's 46-minute interview with Tristan Buckmaster about his role in the disputed Navier–Stokes solution. The description says OpenAI replied after upload and an interview is being scheduled. ## Description (written by Gemini from the video) **Summary** Brady Haran interviews mathematician Tristan Buckmaster (Courant Institute, NYU) about the controversy surrounding OpenAI's September 8, 2026 announcement claiming a solution to the Navier–Stokes Millennium Prize problem. Buckmaster details his work with collaborator Levent Alpöge (Anthropic), their prior work on the Euler equations using AI agents and OpenAI's Codex, and accuses OpenAI of improperly utilizing their unreleased methods and attempting to exclude Alpöge from authorship. --- **What is shown** - [00:00] Opening text card stating that OpenAI published a solution to the Navier–Stokes problem on September 8, 2026, and that Buckmaster claims his findings were used inappropriately. - [00:16] Split-screen video call between Brady Haran and Tristan Buckmaster. - [02:25] Graphic and photo identifying Buckmaster's collaborator Levent Alpöge at Anthropic. - [13:32] On-screen caption noting Numberphile emailed Sébastien Bubeck at OpenAI requesting an interview. - [16:54] Caption explaining Codex as the OpenAI tool used by Buckmaster during his research. - [19:26] Caption citing OpenAI's response that Buckmaster's Codex prompts could not have influenced the system or its training. - [30:55] Caption citing Sébastien Bubeck's denial that he requested Alpöge be removed from authorship. - [45:16] Full-screen statement from OpenAI stating an internal investigation confirmed Buckmaster's Codex prompts did not influence their reinforcement-learning-developed proof and noting mathematical distinctions between the proofs. - [45:34] Numberphile end credits and related video recommendations. --- **Claims & numbers** - **Sleep and workload:** Buckmaster states he averaged about 2 hours of sleep per night during the week the events occurred due to intense stress and constant inquiries [00:44]. - **Timeline and prior collaboration:** Buckmaster notes he previously collaborated with Google DeepMind and was familiar with tech culture [01:28], having worked directly on singularity problems for 6 to 7 years [04:36]. - **Prior rumors:** Buckmaster claims rumors circulated online that Anthropic had solved two Millennium Prize problems (Navier–Stokes and the Hodge conjecture) [08:50]. - **Agent scale:** Buckmaster states OpenAI utilized 10,000 agents for their Navier–Stokes run [01:02, 13:33], whereas they purportedly only used 100 agents for the Euler equations result [13:38]. - **Core mathematical method:** Buckmaster explains that the breakthrough relied on combining convex integration (tracing back to John Nash) with growth mechanisms from Euler blow-ups to resolve non-solutions [34:52, 40:04]. - **Authorship pressure:** Buckmaster claims OpenAI offered him sole authorship on their paper if he dropped his collaborator Levent Alpöge because Alpöge is an Anthropic employee [30:17–31:08]. - **OpenAI official statement:** OpenAI claims Buckmaster's Codex prompts over the preceding two months could not have influenced their models or training, that their model was developed via large-scale reinforcement learning on a pretrained base, and that their result was unforced while Alpöge and Buckmaster's work involved external forcing [45:16]. --- **Notable quotes** - *"The philosophy is move fast, break things... and that philosophy with math clashes."* — Tristan Buckmaster [01:46] - *"I had this WhatsApp group with Levent, and we had like a whole bunch of results with different problems... I think we have to find a positive way forward that does incorporate AI."* — Tristan Buckmaster [40:19, 44:46] - *"They were asking me to throw him under the bus... just because he's an Anthropic employee."* — Tristan Buckmaster [30:37] --- **Assessment** This is an authentic journalistic interview by Numberphile investigating the priority and intellectual property dispute between academic mathematicians and OpenAI following the September 2026 Navier–Stokes proof claim. It features direct dialogue from Tristan Buckmaster alongside official written rebuttals from OpenAI, presenting claims and counterclaims regarding AI tool data use and academic attribution. _Described by gemini-3.8-flash on 2026-10-12 from the video's audio and frames._ ## Related - 2026-09-08: [OpenAI claims a Millennium Prize problem](https://postcutoff.com/e/2026-09-08-openai-navier-stokes-blowup/) ## People in it - [Levent Alpöge](https://postcutoff.com/person/levent-alpoge/), Mathematician, member of technical staff, Anthropic - [Sébastien Bubeck](https://postcutoff.com/person/sebastien-bubeck/), Researcher, OpenAI