OpenAI claims proof for Navier-Stokes equation, mathematicians disagree on approach

OpenAI presented a 166-page potential solution to the Navier-Stokes equation in early September, claiming it follows the approach of Spanish mathematicians Diego Córdoba and Luis Martínez-Zoroa. However, several expert groups, including mathematician Gonzalo Cao-Labora of EPFL, state the similarities are only surface-level and the approach is fundamentally different. The proof, written in the formal language Lean, has been under review by the Clay Mathematics Institute. Disputes arose when NYU mathematician Tristan Buckmaster accused OpenAI of building on his and Levent Alpöge's prior work, which OpenAI denied. Furthermore, OpenAI researcher Noam Brown revealed that the project deployed roughly 10,000 model instances for 88 hours using 130 billion tokens, but admits less than 10 percent of the success is attributable to the multi-agent setup, stating the base model itself is what matters.