An AI-assisted proof of the Riemann hypothesis faces its hardest test
An AI-assisted proof of the Riemann hypothesis faces its hardest test
A claim that AI has helped solve one of mathematics’ most famous puzzles is colliding with a familiar question: does a computer-checked proof amount to a breakthrough if mathematicians have yet to verify and understand it?
Anthropic mathematician Levent Alpöge says Claude helped him develop a proof of the Riemann hypothesis, a problem about the distribution of prime numbers that has resisted solution since 1859. He released the argument as Lean code, which can check formal proofs. But the claim has not been independently confirmed, and the Clay Mathematics Institute still listed the hypothesis as unsolved after his announcement.
Alpöge’s account is strikingly enthusiastic: “No, Claude flew, and danced, and sparkled near that finish line.” The Associated Press report presents a more measured picture of what the code establishes. Rutgers mathematician Alex Kontorovich said Claude had followed the proof code, but stressed, “Does that mean I understand it? Not yet.” He added that he still had to do the work of examining it.
That distinction is central. Formal verification can show that a proof follows specified rules; it does not by itself mean experts have grasped the mathematics or accepted the result. Martin Bridson, president of the Clay institute, called the announcement potentially transformative, while warning that “the journey from today’s announcement to a place of real understanding will be long and difficult.”
The timing also invites scrutiny. Alpöge’s announcement came just after Anthropic’s stock-market debut, and he acknowledged the timing was “horrible.” The liberal account situates the claim amid concerns about costly AI competition and its environmental footprint, quoting a mathematician who called such corporate displays “immature playground boasting writ large.” The AP, meanwhile, highlights the scale of the work—millions of lines of Lean code—and the potential significance if it holds.
For now, neither the excitement nor the skepticism settles the mathematics. The decisive test is whether independent specialists can verify the argument and make sense of it.
Write a comment