World · Urbans Post
Mathematician Tristan Buckmaster believes OpenAI used his work to rush ahead and beat him to solving a legendary math problem with a $1 million bounty. But that’s not been enough for him to stop using the company’s models—and he’s not the only mathematician that feels that way. “Even if you don't agree with any of this, you're kind of stuck.
With AI being so useful, it's hard to completely prevent oneself from using it,” Buckmaster tells WIRED. “These companies have a monopoly, and there is not much choice,” he adds. In the week and half since the New York University professor accused OpenAI of copying his approach, Buckmaster has been using the company’s coding agent Codex to tidy up his research papers.
When he has time to do math (which he says is rare, since the fallout thrust him into the spotlight), the tool has been helping him understand the logical steps OpenAI’s agents might have taken to get from his earlier workings to the final proof. Buckmaster had used Codex as well as Anthropic’s competing Claude to work on what’s known as the Navier-Stokes existence and smoothness problem, alongside Anthropic researcher Levent Alpöge. OpenAI deployed tens of thousands of agents to reach the solution, but only after it learned the equation was close to being solved, Buckmaster says.
When Buckmaster went public with his claims, it ignited a firestorm about artificial intelligence and whether it would make human mathematicians obsolete . It also led OpenAI to do an investigation and amend its announcement about solving Navier-Stokes to say it “confirmed that Buckmaster’s Codex prompts over the two months preceding this announcement and paper on September 8, 2026, could not have influenced the system in any way, including through training.” The company pointed WIRED to its announcement in an email. Showing that AI was pushing the boundaries of mathematics was “more important than the result,” Buckmaster says.
But churning out solutions to long-standing math problems without fully crediting the human work undergirding them—especially ahead of major IPOs—is irresponsible and “childish,” he says. Other mathematicians have raised similar concerns. Only a handful of people on the planet understand the techniques in geometric group theory that German mathematician Andreas Thom has dedicated the last two decades to developing.
So when OpenAI said in August that its Astra model had used them to prove a long-standing problem he had been working on, “I was amazed,” says Thom. “And of course I was wondering, how did they learn about it?” So he says he asked OpenAI researchers Mark Sellke and Sébastien Bubeck. In an August email, he pointed out that the firm’s assertion that “no progress” had been made on the problem in the last decade overlooked a 2019 paper of his, as well as other mathematicians’ work.
The company amended its press release. He and a colleague had been using ChatGPT to assist their work on the problem in the months running up to the result, but when he asked if their interactions had been fed into training data, Thom says Sellke replied: “That did not happen.” “I set it aside,” Thom recalls. “I'm not so much interested in these political things; I want to work on mathematics.” While Thom has seen OpenAI’s statement that Buckmaster’s prompts couldn't have influenced the system, he says he doesn’t trust this and concedes he will probably never know whether his work actually fed the result.
“AI really kills this entire idea that you could trace back who contributed what,” he says. “That is probably over.” It’s a big change from how science is typically done, with academics subjecting their findings to peer review and building on each others’ work with credit. Beyond upending attribution, the fact that humans still don’t fully understand each step OpenAI’s agents took to arrive at the Navier-Stokes result also poses existential questions for mathematicians, who see their field slipping from their understanding.
"If I want to make a contribution to mathematics, how do I do that as just a human nowadays when these trillion-dollar companies are in on the game?" says Cornell mathematician Alex Townsend, the coauthor of a forthcoming book on the field’s evolution. Since August, Townsend has seen many of his colleagues start to ask what they need to know about the technology and how they can set up subscriptions to access higher-powered models. "I feel both excited and nervous simultaneously," Townsend says.
"Excited because I can achieve things that I couldn't achieve without it, and nervous because I'm questioning: 'OK, what's my purpose here?'" Thom accepts that his area of study is changing, and has continued to use ChatGPT—with updated privacy settings to stop his work being used to train the data—to speed up writing papers because it is “extremely efficient.” (OpenAI’s offerings through universities and at the enterprise level default to not training models on users’ data.) “If a human had actively done that, then I would be very, very angry,” he says of someone using his work without attribution. But if information was pulled into the model through a back door, by an algorithm which nobody fully understands, “I could probably live with that,” he says. Some mathematicians are less forgiving.
Twenty-five Fields medalists wrote in an open letter that AI companies and mathematicians are “severely misaligned.” More than 4,000 people have signed the Leiden Declaration, which has a series of recommendations for how mathematicians, funders, and politicians can ensure that AI doesn’t swallow the field. Buckmaster fears the possibility of using AI to “clean up some of your grammar, and suddenly your years of work [could be] gobbled up in user data and sold to another mathematician or grad student. That's what I think most of the mathematicians tend to be worrying about, and I think it's a real issue.” With no oversight on the horizon and AI further engraining itself in the field, some mathematicians want to find a way to at least tap the brakes.
That includes more than 2,000 people with ties to Caltech who called on organizers of an AI math hackathon at the school to suspend the event. The hackathon was sponsored by Anthropic and OpenAI, though the latter has since dropped out. But any attempts to slow things down may be for naught, especially in the long term.
“I don't think this is really sustainable because of the efficiency gain” that AI offers, says Thom. Any mathematicians—especially early-career researchers—risk being “isolated” if they don’t use AI to accelerate their work, he adds. He and Buckmaster both believe the community needs to start thinking about what the technology means for younger mathematicians and how to smooth the transition.
The next generation of mathematicians is already looking for guidance: Students have also been asking about what their future could look like now that AI is becoming so capable at solving math problems, Townsend says. Buckmaster is calling for a detente in AI-driven mathematics while mathematicians and AI laboratories set some ground rules on how to release results, including getting their references right. He himself is going to clean up the papers he published prematurely last week to beat OpenAI’s announcement, one of which he described as “AI slop.” ”I have a responsibility to clean up the papers that I did post that weren't completed, and I think I have a responsibility to explain to mathematicians what we did,” he says.
He’s also open to discussing this with OpenAI. “I don't want to just engage in fights” he says. As for whether he would ever actually work on a problem with the company, “we have to be careful with that,” he says with a wry smile.
Source: WIRED
