OpenAI Claims Math Breakthrough, Faces Controversy

OpenAI announced a solution to the Navier-Stokes problem, sparking both excitement and accusations of impropriety.

OpenAI claims to have solved the Navier-Stokes problem, a major mathematical challenge that has remained unsolved for approximately 90 years. The company announced this discovery in a blog post on Tuesday, stating that an internal AI model, more powerful than GPT-6 Astra, achieved the solution using 10,000 concurrent agents. The Navier-Stokes problem is one of the Millennium Prize Problems, each associated with a $1 million reward, though OpenAI states it does not plan to take the prize.

This announcement has been met with controversy. New York University mathematics professor Tristan Buckmaster, along with Anthropic researcher Levent Alpöge, had been working on a related problem. Buckmaster claims he contacted OpenAI after learning of their progress and found OpenAI's proof for the Navier-Stokes equation utilized a route he and Alpöge had explored using OpenAI’s Codex and Anthropic’s Claude. Buckmaster raised concerns about potential access to their Codex data, a suspicion OpenAI has attempted to address by stating no specific user data was accessed, but acknowledging that de-identified data from product usage might have contributed.

OpenAI stated that their proofs differ significantly from Buckmaster and Alpöge's work, which they claim to have seen only after public release. However, Buckmaster countered on Mastodon, alleging OpenAI admitted to using training data from a period after his team's result. This situation highlights a debate around the origins of AI-driven discoveries and the ethical considerations surrounding data access and intellectual property in advanced AI research.

Find the right AI tool for your business

Chat with Insta and get matched to the right tool in seconds.