OpenAI announced that it had solved the Navier-Stokes problem, one of seven Millennium Prize Problems designated by the Clay Mathematics Institute as among the most significant unsolved challenges in mathematics. The company utilized an internal system more powerful than its GPT-6 Astra model, deploying approximately 10,000 autonomous AI agents that worked simultaneously on the problem and reached a solution within 88 hours. Verification of the solution took the GPT-6 Astra model approximately 17 hours.
The announcement quickly became subject to dispute when mathematician Tristan Buckmaster of New York University raised concerns about the timing of OpenAI’s effort. Buckmaster indicated that OpenAI had accelerated its work after learning that he and a researcher at Anthropic were preparing to announce their own breakthrough. He expressed additional concerns because the collaborative work had been stored in OpenAI’s Codex model, a system designed for computer programming that could potentially have exposed the research to OpenAI personnel. Buckmaster stated he could not determine what OpenAI’s model accomplished or whether the researchers’ data had been accessed, while clarifying he was not making direct accusations.
At a press briefing, OpenAI researcher Sebastien Bubeck denied that the company had utilized the pair’s work or accessed materials shared with OpenAI servers. However, OpenAI’s official statement acknowledged it could not exclude the possibility that data from the researchers’ use of OpenAI products “helped improve our models.” The company stated it had been motivated to pursue the problem after hearing rumors that two Millennium Prize Problems had been solved.
The Navier-Stokes problem concerns whether equations describing fluid movement fail under certain conditions. OpenAI’s solution suggests they do fail, with the equations occasionally becoming unbounded with fluid speeds reaching impossibly infinite values. Winners of any Millennium Prize Problem are eligible for a $1 million award from the Clay Mathematics Institute, though OpenAI indicated it would not seek the prize.
The breakthrough demonstrates the expanding role of advanced artificial intelligence systems in mathematics, following similar claims by Google DeepMind and OpenAI’s own announcement in May regarding another mathematical problem proposed 80 years earlier. The announcement also provides positive publicity for OpenAI during a period of heightened scrutiny regarding AI safety practices.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI