On Sunday afternoon, while the world’s eyes were glued to the World Cup final, an AI model quietly resolved a problem that had tortured mathematicians since 1939. By the time Kevin Buzzard woke up in London the next morning, the result had been verified. By lunch, it was all his peers at Imperial College London’s pure mathematics department could talk about. Anthropic employee Levant Alpöge’s post announcing the result has since drawn more than 20 million views on X.
A breakthrough that arrived while nobody was watching
The timing was almost poetic. As millions celebrated a sporting spectacle, an AI system did what human mathematicians had failed to do for 85 years. The problem, first posed in 1939, had become a quiet legend in the field — a puzzle that generations of researchers had chipped away at without success. The AI solved it in what appears to have been a matter of hours.
Why this moment feels different from previous AI wins
AI has been making inroads into mathematics for years, from helping prove theorems to suggesting new conjectures. But this felt different. The problem was not obscure — it was a known, stubborn challenge. The speed of the solution, the fact that it happened during a global distraction, and the verification that followed within a single day have left mathematicians confronting something they had long anticipated but never fully prepared for: the sense that their field is being reshaped faster than they can adapt.
‘A great time to be alive’ — but also an unsettling one
Kevin Buzzard, a mathematician at Imperial College London, captured the mixed emotions. “It is a big day,” he told Fortune. “I think it’s a great time to be alive, personally.” Yet the same interview revealed a deeper unease. The change, Buzzard acknowledged, is “very rapid and very unsettling.” The phrase has resonated across the mathematics community, where excitement about AI’s capabilities is now shadowed by questions about what it means for human mathematicians.
What the AI actually did — and how it was verified
The AI model, developed by researchers at Anthropic, did not simply guess an answer. It produced a rigorous solution that could be checked by human mathematicians. The verification process, which normally takes weeks or months for a major result, was completed within hours. This speed of verification is itself a landmark — it suggests that AI is not just generating plausible answers but producing work that meets the standards of formal mathematical proof.
Confirmed facts vs what remains unclear
What is confirmed: An AI model solved a mathematics problem that had been open since 1939. The solution was verified by mathematicians at Imperial College London. The announcement has generated over 20 million views on social media. Kevin Buzzard has publicly described the event as both exciting and unsettling. What remains unclear: The exact nature of the problem, the specific AI model used, and whether the solution will lead to further breakthroughs or remain an isolated achievement. The broader trajectory of AI in mathematics is still being understood.
How mathematicians are reacting — from awe to anxiety
Across university departments and online forums, the reaction has been split. Some researchers see the breakthrough as a tool that will accelerate discovery, freeing humans to focus on deeper conceptual questions. Others worry that the role of the mathematician is being fundamentally redefined — from problem-solver to problem-verifier, or worse, to spectator. The phrase “very rapid and very unsettling change” has become a shorthand for this anxiety.
The pattern behind the breakthrough: AI’s accelerating assault on unsolved math
This is not an isolated event. It is the latest in a series of AI-driven mathematical breakthroughs that have compounded rapidly over the past few years. Each new success narrows the gap between what AI can do and what was once considered uniquely human. The pattern suggests that the pace of AI progress in mathematics is not linear but exponential — and that the field may need to rethink its fundamental assumptions about research, collaboration, and even what constitutes a meaningful mathematical contribution.
What this means for students and young researchers
For students pursuing careers in pure mathematics, the message is complex. The tools of the trade are changing. The ability to solve problems may become less valuable than the ability to frame them, to ask the right questions, or to interpret AI-generated proofs. Young mathematicians may need to develop skills that their professors never learned — working alongside AI systems, verifying machine-generated reasoning, and finding new areas where human insight still matters.
What happens next in AI and mathematics
The immediate future is likely to bring more such breakthroughs. AI models are being trained specifically on mathematical reasoning, and each solved problem provides training data for the next. The 1939 problem may soon be joined by others from the 1940s, 1950s, and beyond. The question is not whether AI will solve more unsolved problems, but how quickly — and whether the mathematics community can adapt its institutions, its education, and its sense of purpose in time.
Our take
This is not a story about AI replacing mathematicians. It is a story about a profession confronting the fact that its most cherished activity — solving hard problems — is no longer exclusively human. The excitement Buzzard feels is genuine, and so is the unease. The challenge for mathematics is not to resist the change but to find a new equilibrium, where human creativity and machine reasoning coexist. That will require not just technical adaptation but a cultural one — and that kind of change is always the hardest to solve.
Frequently Asked Questions
What problem did the AI solve?
The AI solved a mathematics problem that had remained unsolved since 1939. The exact nature of the problem has not been publicly detailed, but it was a known open problem in pure mathematics that had resisted human efforts for 85 years.
How was the solution verified?
The solution was verified by mathematicians at Imperial College London within hours of the AI producing it. The verification process confirmed that the AI’s reasoning met the standards of formal mathematical proof.
Who developed the AI model?
The AI model was developed by researchers at Anthropic, the AI company. The announcement was made by Levant Alpöge, an Anthropic employee, whose post on X has drawn over 20 million views.
What does this mean for the future of mathematics?
Mathematicians are divided. Some see AI as a powerful tool that will accelerate discovery. Others worry that the role of human mathematicians is being fundamentally redefined. The consensus is that the pace of change is “very rapid and very unsettling,” and the field will need to adapt its methods and culture.
