OpenAI announced that a new internal model, whose training began on August 28, has reportedly solved the Navier–Stokes Millennium Prize problem and more than 100 other long‑standing open problems across many areas of mathematics. The pace of progress surprised mathematicians at OpenAI and prompted internal discussions about how best to inform and prepare the broader mathematical community.
The company emphasized that mathematics is a foundational science and that new mathematical discoveries can enable a wide range of applications. Because of this, the responsible development and deployment of math‑related AI capabilities is important beyond the discipline itself.
Community concerns and an open letter
Some mathematicians have expressed concern about the effects of using AI systems to solve open problems. In a recent open letter titled “A Severe Misalignment of AI in Mathematics,” they highlighted potential negative externalities when open problems are treated as benchmarks for new AI systems. Those criticisms underlined the need for AI developers to engage thoughtfully with the mathematics community.
Role and independence of the advisory group
In response, OpenAI is working with mathematicians who have formed an independent advisory group on mathematics and artificial intelligence. The group is intended to serve as a bridge between OpenAI, the mathematical community, and the general public, giving mathematicians a voice in decisions about how to proceed.
The advisory group will provide guidance on reviewing and communicating emerging results: it will help OpenAI assess the significance of findings, advise on coordinated dissemination, and advise on academic and professional standards for mathematical research. It will also offer input on how OpenAI’s tools can support mathematical research and education, with the aim of putting capable tools in mathematicians’ hands so they can pursue the questions they deem important.
The group will operate independently of OpenAI. Members will be free to offer unsolicited advice, comment on OpenAI’s impact on mathematics, and publish their recommendations. They will not be paid by OpenAI, and the group may change its membership at its discretion. Crucially, the advisory group will not be responsible for advising OpenAI on pacing its internal mathematical progress.
Initial membership and next steps
The advisory group is hosted at the Institute for Advanced Study. The initial members listed are:
- Timothy Gowers (Collège de France, Cambridge)
- Martin Hairer (EPFL, Imperial College London)
- Nikhil Srivastava (Berkeley, Simons Institute)
OpenAI characterizes this collaboration as an initial step. It acknowledges there are difficult questions ahead about how AI can support mathematical understanding and how the benefits of these capabilities can be distributed more broadly. The company states it wants mathematicians to be central in shaping those answers.
