OpenAI says 10,000 AI agents produced a proposed solution to a $1 million Navier‑Stokes problem
What happened
OpenAI says an internal experimental system that used about 10,000 AI agents produced a proposed solution to the Navier‑Stokes existence and smoothness problem, one of the seven Millennium Prize Problems that carries a $1 million prize.
Key numbers
- ≈10,000 AI agents collaborated.
- 2.7 million messages exchanged.
- ≈130 billion tokens generated.
- 88 hours of work before the solution was drafted.
- Additional 17 hours spent formalizing the proof.
How the AI system worked
The system, described by OpenAI as more capable than its publicly available GPT‑6 Astra model, let agents pass messages and generate text. After the initial phase, the model named Astra spent another 17 hours checking and formalizing the proof.
The proposed proof suggests a fluid vortex could become infinitely fast in finite time while keeping total energy finite, a scenario that would break the smoothness condition of the Navier‑Stokes equations.
Mathematicians’ concerns
Some mathematicians, including NYU’s Tristan Buckmaster, have questioned whether the AI’s result was truly independent. They note that de‑identified data from two researchers’ related work may have been part of the training data, though OpenAI says the model did not access those specific papers before they were published.
Why it matters
If the proof holds up, it would resolve a problem that has been open for 90 years and could improve understanding of fluid dynamics used in aircraft design, weather forecasting, and medical research.
Beyond the mathematics, the ability of thousands of AI agents to tackle a long‑standing problem shows a new way to approach complex scientific challenges.
Next steps
The solution must undergo rigorous peer review before it can be accepted as a valid proof. The mathematics community is expected to scrutinize the claim in the coming months.
