OpenAI closer to solving another millennium prize problem

On September 17, 2026, The Information reported that OpenAI is nearing a solution to the Hodge conjecture, one of the seven Millennium Prize problems. A source close to the project indicated that the team expects to announce the breakthrough soon. The Hodge conjecture, which deals with algebraic structures in geometric shapes, has not yet been solved, though OpenAI may delay public disclosure to avoid past PR issues. Earlier, in September 2026, OpenAI claimed its internal model had solved the Navier-Stokes existence and smoothness problem, another Millennium Prize challenge. The effort reportedly cost millions, using a model variant named 'Doug'. Researchers believe that mathematics is the next natural field for large models, due to its reliance on logical reasoning and verifiable results. Some predict that automation in mathematics could follow the recent trends in software engineering. Solving complex math problems also advances machine learning automation, a key step toward recursive self-improvement in AI. OpenAI’s success in tackling these problems could signal progress in tackling challenges in biology and chemistry.