AI Summary
5 min readDid OpenAI Steal the Navier-Stokes Solution?
NYU mathematician Tristan Buckmaster put out a statement accusing OpenAI of only starting to prompt its model on the Navier-Stokes problem—one of seven Millennium Prize problems carrying a $1 million reward—after word of his own work reached the company. Buckmaster had spent most of the year on the problem with Lev Alantpaj, a mathematician who works at Anthropic, and had been running draft reports through OpenAI's Codex the whole time. When they asked whether the model had been trained on their sessions, they never got a clear answer. Buckmaster also claims OpenAI twice pushed him to drop Alantpaj from the paper because he works for their competitor. He then threatened to go public and was asked, "Why would you ruin your career?" OpenAI's Sebastian Boubeck called all of this "false and inflammatory" and has since apologized for that line, though OpenAI's own write-up confirms its effort only began on September 1st after hearing the rumor the problem was being worked on.
The Doomerism Debate That Won't Die
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of This Week in Startups
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **AI & Research: The Navier-Stokes Accusation** - NYU mathematician Tristan Buckmaster accuses OpenAI of using his private research to crack the Navier-Stokes problem, sparking a debate on trust.
- 2 (02:02) **Introducing the VC Roundtable** - The hosts and guests are introduced for the weekly investor discussion.
- 3 (03:14) **AI Doomerism: Are We All Going to Die?** - The panel reacts to Anthropic researchers quitting and warning of a 10% chance of AI wiping out humanity by 2030.
- 4 (11:23) **The PR Disaster & Public Sentiment** - The panel argues that AI companies have scared the public so badly that a backlash is now inevitable.
- 5 (17:43) **The 10x Employee: AI Adoption in the Workplace** - Jason shares how he forced his firm to adopt AI, leading to a massive productivity gap between users and non-users.
- 6 (23:52) **Autonomous Vehicles: The Long, Slow Rollout** - The panel discusses why self-driving cars and drone delivery have taken longer than expected.
- 7 (24:55) **Zipline: A Missed Investment & A Hard Lesson** - The investors share their personal stories of missing out on the drone delivery company Zipline.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
This Week In Startups is made possible by:
Northwest Registered Agent - https://www.northwestregisteredagent.com/twistdomain
NetSuite - https://www.netsuite.ai/twist
Vanta - http://www.vanta.com/twist
Check out a full transcript and summary of today's show, created by Plaud: https://web.plaud.ai/s/pub_c2b45bfe-abad-43ce-b7ef-41118f55f738::Vo2MH6PuXBzTCaxKRDjHH0zu8FbzgDpQL5OnfBfMzLP22CQRxciqE74Mq4YDD-LwYXGZCazgIeglPuwCToday's show:
An Anthropic researcher quit this week, saying both leading labs are gambling with everyone's lives. His coworker chimed in and predicts there is a greater than 10% chance that AI will end humanity IN THE NEXT 10 YEARS. Jason and Lon take it to Yohei Nakajima of Untapped Capital, Ben Lerer of Lerer Hippeau, and Rebecca Lynn of Canvas Prime, who cannot decide whether it is the greatest PR strategy in history or some kind of mass psychosis.
Learn why Jason says founders should never hand a frontier lab their proprietary data, whether or not they're offered free credits. PLUS, who really solved Navier-Stokes, and what does the OpenAI–Buckmaster fight mean for anyone doing real work inside ChatGPT or Claude?
Guests:
Yohei Nakajima on X: https://x.com/yoheinakajima
Untapped Capital: https://untapped.vc/
Ben Lerer on X: https://x.com/BenjLerer
Lerer Hippeau: https://www.lererhippeau.com/
Rebecca Lynn on X: https://x.com/VCRebecca
Canvas Prime: https://www.canvas.vc/
Relevant Links:
Jacob Coxon quits Anthropic - "gambling with our lives" https://techcrunch.com/2026/09/09/gambling-with-our-lives-anthropic-researcher-quits-warns-against-self-improving-ai/
Evan Hubinger, Anthropic's Alignment Science lead, on X: >10% odds within a decade — https://x.com/EvanHub/status/2097497037956891126
Future of Life Institute — the 2015 Puerto Rico AI safety conference Jason asks Lon to pull up → https://futureoflife.org/
Meta debuts Muse (Axios) —<
More from this podcast
This Week in Startups →