OpenAI Claims AI Solved Navier-Stokes, Sparking Credit

OpenAI announced today that its artificial intelligence has produced a solution to the Navier-Stokes equations, a 200-year-old mathematical problem describing fluid motion. The discovery, one of the unsolved Clay Millennium Prize problems worth $1 million, was made after dedicating over 1,000 AI agents to the task for more than 50 hours.

Sebastien Bubeck, a mathematician and AI researcher at OpenAI, said the company began training a new model with advanced math capabilities on August 28. After hearing rumors that rival AI firm Anthropic was making progress on Navier-Stokes, OpenAI allocated more resources. The computational effort cost "in the millions of dollars," according to Mark Chen, OpenAI's head of research.

The AI-Driven Discovery

Bubeck described initial disbelief at the result. "I thought there must be a mistake somewhere," he said. The solution was finalized and formalized using the Lean programming language on a Sunday morning. OpenAI emphasized this proof required vastly more computing power than its previous mathematical work.

The announcement was swiftly contested. On Monday, mathematician Tristan Buckmaster of NYU and Levent Alpöge, a researcher at Anthropic, posted documents claiming key advances on a related area of the problem. They used several AI models, including Claude and Codex.

Allegations of Rushed Credit

Buckmaster alleges OpenAI became aware of his and Alpöge's work last week and then rushed to solve the problem. In a statement, he claims he asked OpenAI leaders if the company had accessed their Codex logs. He says he was told the model "didn't look up user data" but received no answer on training data questions.

Buckmaster further claims OpenAI presented several proposals. One proposal, he says, involved him publishing a paper announcing the Navier-Stokes problem was solved by an internal OpenAI model, but excluding Alpöge's name. Buckmaster, Alpöge, and Anthropic did not respond to WIRED's request for comment.

OpenAI's Denial and Recognition

During a press briefing, Bubeck and other executives denied inspecting the rival researchers' work. "We, whether it’s the researchers or the agents, did not see any of their work until it was released publicly last night," Bubeck stated.

He offered congratulations. "I want to be extremely clear that we recognize the priority of Levent Alpöge and Tristan Buckmaster’s work on unforced Euler, and we have nothing but congratulations to them on this monumental achievement," Bubeck said. "To be clear, we did not use their prompt or proof to prompt our models or direct our agents."

OpenAI stated it wants to recognize the prior work. However, Ven Chandrasekaran, a mathematician at OpenAI, stressed their AI-generated solution is fundamentally different in nature from the one produced by Buckmaster and Alpöge.

The dispute highlights how competition in AI-driven research could lead to more conflicts over attribution. As AI takes on more complex proof-finding work, such controversies may become more frequent.