OpenAI claims to have solved a 90-year-old mathematics problem using artificial intelligence, according to a report published by BBC Business. The company stated that it used a new AI model and thousands of AI bots to solve the Navier–Stokes existence and smoothness problem in just 88 hours. The problem, part of the Navier-Stokes equations, concerns how fluids move and has lacked a proof for nearly a century.

The solution, which OpenAI called a 'milestone,' resolved two out of the four statements in the proof required for the Millennium Prize, a $1 million reward for solving certain mathematical conundrums. However, the result has not yet been verified independently or accepted by the Clay Mathematics Institute, which oversees the prize. OpenAI emphasized that it does not intend to claim the prize for this result.

The company admitted that it had 'heard rumors' about two Millennium Prize problems being resolved by September 1 and decided to deploy its AI bots to tackle the remaining challenges. By September 5, the AI had reportedly found a solution to the Navier–Stokes problem. The process involved nearly 3 million messages and 130 billion output tokens, costing an estimated $10 million, based on OpenAI's own pricing for its most advanced models.

Tristan Buckmaster, a mathematics professor at New York University, raised concerns about the timeline of OpenAI's work. He stated that he and Levent Alpöge, a mathematician at OpenAI rival Anthropic, had been working on the problem using OpenAI's Codex tool. Buckmaster claimed that on September 3, he discovered that 'information about our progress had been passed to OpenAI,' suggesting that the company may have accessed data from their work. He argued that OpenAI did not begin working on the Navier–Stokes equations until 'after information about our work had reached OpenAI.'

OpenAI responded by congratulating Buckmaster and Alpöge on their 'concurrent work,' calling it 'remarkable.' The company stated that it had not seen any of their work until it was released publicly and that no user data was accessed in its Navier–Stokes research. However, it acknowledged that 'de-identified data derived from their usage of our products' could have contributed to its models, though it emphasized that its proofs and results differ significantly from theirs.

Despite the controversy, OpenAI has not provided the full proof of its solution, and the Clay Mathematics Institute has not yet reviewed or accepted the result. The company's claim highlights the growing role of AI in mathematical research but also raises questions about data sharing, intellectual property, and the verification process for such groundbreaking claims. The broader implications of AI's ability to solve complex problems remain to be seen, as the scientific community continues to evaluate the validity and significance of OpenAI's achievement.