Key facts
- OpenAI told The New York Times it has made "substantial progress" on a second Millennium Prize Problem.
- The company stated it is not seeking the $1 million prize money.
- Online speculation suggests the problem may be the Hodge Conjecture.
- The Clay Mathematics Institute's review of OpenAI's Navier-Stokes proof is pending.
- Mathematician Tristan Buckmaster claims he and a colleague finished a related proof before OpenAI published its Navier-Stokes solution.
OpenAI has announced significant progress on a second Millennium Prize Problem, a prestigious mathematical challenge offering a $1 million reward. This follows the company's September 8 claim of solving the Navier-Stokes equations, another of the seven problems established by the Clay Mathematics Institute in 2000.
While OpenAI has not officially named the second problem, speculation on social media, particularly among mathematicians and enthusiasts, points towards the Hodge Conjecture. This 76-year-old puzzle concerns counting the "holes" in complex geometric shapes using algebraic shortcuts.
However, OpenAI's recent claims are accompanied by ongoing scrutiny. The Clay Mathematics Institute's review of the Navier-Stokes proof is still pending. Furthermore, a dispute has arisen with mathematician Tristan Buckmaster, who alleges that OpenAI's Sébastien Bubeck raced to publish the Navier-Stokes solution after Buckmaster and a colleague had already completed a related proof.
OpenAI has stated its motivation is not the prize money but to use these complex problems as a benchmark for the rapid improvement of its AI models. The company's unreleased internal model, potentially the rumored Aeon version designed for long-term tasks, is reportedly behind this latest mathematical endeavor. The broader impact of AI in mathematics was highlighted by Anthropic's recent announcement that its Claude model formalized Fermat's Last Theorem in 11 days, prompting warnings from mathematician Terence Tao about the accelerated consumption of challenging mathematical problems.