OpenAI CEO Altman: Mathematicians Have Moved from "We're Doomed" to Accepting AI
OpenAI CEO Altman said that mathematicians' attitude toward AI has shifted from "we're doomed" to acceptance, and compared this process to mourning. In September this year, OpenAI announced that its model had solved the Navier-Stokes equations, and subsequently announced that it had solved more than 700 problems; Terence Tao said that the "romantic era" of mathematics has ended, and that top scholars are shifting toward verifying AI's work.
Body · AI translation
Altman compared this process to mourning: the first day is disbelief, "we're done for, this field is over, let's have the graduate students change majors"; by the third day, it might be anger, "OpenAI doing this is terrible"; and by the fifth day, many will say, "I'll have a different, but hopefully better, career path as a mathematician."
Altman's optimistic assessment stands in sharp contrast to the "existential shift" described by some in the mathematics community. In September of this year, OpenAI suddenly burst into the field of mathematics, announcing that its model had solved the Navier-Stokes equations—a famous problem that has stumped the field's top figures since the 19th century. Two scholars, one from New York University and one from Anthropic, accused OpenAI of using data from their chat logs to help crack the problem. Other mathematicians complained that even if the AI model may indeed have solved the problem, it did not show the derivation process, leaving mathematicians with the less pleasant work of verification.
The controversy did not die down. A group of undergraduates at Caltech, worried that AI is rapidly upending the field of mathematics and hoping not to be marginalized, launched a "math marathon" that encourages participants to use AI to solve long-unsolved problems. More than 700 scholars signed an open letter condemning the competition as promoting AI "slop" (referring to low-quality AI content). OpenAI subsequently withdrew its sponsorship. The organizers then adjusted the rules, and the competition is scheduled to be held in November.
Terence Tao has expressed similar concerns. This mathematical prodigy at the University of California, Los Angeles, earned his doctorate at age 20. He said that the "romantic era" of mathematics is over, and that top scholars are adapting, as he is, to a new process: verifying AI's work rather than solving problems alone. After OpenAI announced this month that it had solved more than 700 additional problems, Tao said: "Something is being lost."
Meanwhile, Altman stands on the other side of the debate, maintaining the optimism of a disruptor. Speaking of AI prompting companies to restructure their workforces, he said: "I don't think there will be a jobs apocalypse." But in mathematics, Altman believes the impact is coming faster than in other disciplines and industries. "This is the fastest I think any of us have seen a field go through this cycle."