In a statement released Tuesday, OpenAI said its proprietary artificial‑intelligence system has produced a solution to the Navier‑Stokes problem, one of the seven Millennium Problems that have challenged mathematicians for more than two centuries. The company described the effort as a collaborative run of up to 10,000 AI agents working for roughly 88 hours, using tools that could read cached internet data and execute code.
How the AI tackled the fluid‑dynamics puzzle
According to OpenAI, the internal model was not made public but generated a “Lean” proof that demonstrates a solution to the long‑standing question of whether the Navier‑Stokes equations can develop singularities under certain conditions. The firm emphasized that the achievement represents a snapshot of ongoing progress in AI research rather than a final, unassailable answer.
Academic community reacts
Mathematicians quickly weighed in. Professor Tristan Buckmaster of New York University, who has been investigating the same problem with Anthropic researcher Levent Alpöge, said he has not seen OpenAI’s proof and cannot confirm how the AI arrived at its result. He stressed that he is not accusing anyone of wrongdoing and noted that OpenAI reached out to offer a joint release of findings.
Renowned UCLA mathematician Terence Tao expressed concern that reliance on AI could diminish the creative process of mathematics, likening it to watching only the beginning and ending of a film. Tao warned that an “indiscriminate” use of AI might turn the discipline into a mere production quota game, offering limited benefit to the field.
OpenAI’s response to the criticism
OpenAI acknowledged the possibility that de‑identified data from users of its products could have contributed to the model’s improvement, but it maintained that no specific user data was accessed during the Navier‑Stokes effort. The company also said it began work on the problem on September 1 after hearing a rumor about another Millennium Prize solution, and it offered Buckmaster and Alpöge visibility into the prompts used and the full proof.
Importantly, OpenAI clarified that it does not intend to claim the official $1 million Millennium Prize for this result, underscoring a focus on advancing AI capabilities rather than pursuing the monetary award.
What this means for the tech and research landscape
The announcement highlights the growing power of AI to assist in highly complex scientific endeavors. If the solution holds up under peer review, it could accelerate research in fluid dynamics, with downstream effects on engineering, climate modeling, and aerospace design. At the same time, the episode underscores the need for transparent collaboration between AI developers and the academic community to ensure that breakthroughs are both credible and ethically grounded.
For businesses and investors, the development signals that AI is moving beyond language tasks into domains traditionally reserved for human experts. Companies that can harness such technology responsibly may gain a competitive edge in sectors that rely on advanced mathematical modeling.
Next steps
The mathematical community will now subject OpenAI’s proof to rigorous scrutiny. If validated, the result could be recorded as the second official solution to a Millennium Problem, joining the 2003 proof of the Poincaré Conjecture. Until then, the story remains a vivid illustration of how AI is reshaping the frontiers of knowledge.
Original reporting: KTVZ (Central Oregon) — read the source article.
