OpenAI has published a solution to the Navier-Stokes existence and smoothness problem, one of the seven Millennium Prize Problems. The breakthrough was achieved using an internal next-generation model described as significantly more capable than GPT-6 Astra, coordinated through a swarm of approximately 10,000 AI agents. The system reached its result in about 88 hours, exchanging 2.7 million messages and consuming roughly 130 billion output tokens. The compute cost for this operation alone is estimated to be in the millions of dollars.
The solution identifies a finite-time singularity in three-dimensional incompressible Navier-Stokes dynamics, suggesting that smooth fluid motion can break down under specific conditions. To ensure mathematical rigor, the proof was formalized in Lean, a process completed by GPT-6 Astra over an additional 17 hours.
Controversy and Authorship Disputes
Despite the technical milestone, the announcement is overshadowed by allegations of intellectual theft. NYU mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge claim that OpenAI pressured them to drop Alpöge from the authorship of their own AI-assisted work on the problem. Buckmaster alleges that an OpenAI researcher asked, "Why would you ruin your career?" when he refused to comply. This follows broader concerns previously reported by AlexTech.ai regarding the use of private research to train frontier models.
OpenAI has denied these accusations, stating that its team did not see the work of Buckmaster and Alpöge until it was released publicly. The company maintains that its solution is fundamentally different from the researchers' approach.
Academic and Practical Skepticism
Not all experts are convinced by the magnitude of the discovery. Chris Combs, a professor of aerospace engineering, has cautioned that the result may be more of a "mathematical curiosity" than a practical breakthrough. Combs argues that the proof involves niche scenarios and specific assumptions, noting that it does not provide a general closed-form solution for Navier-Stokes and therefore changes nothing in how these equations are used in real-world engineering practice.
The event highlights a growing tension in the scientific community: as progress on complex mathematical problems increasingly requires massive compute resources available only to a few AI labs, the traditional norms of academic collaboration are being challenged.
No comments yet. Be the first!