Two Riemann proofs turn a math milestone into an AI showdown

An Anthropic mathematician says Claude helped him prove the Riemann hypothesis, but a Stanford researcher soon unveiled a rival proof developed without AI. Neither claim has been independently verified.
Two Riemann proofs turn a math milestone into an AI showdown

Two Riemann proofs turn a math milestone into an AI showdown
Levent Alpöge, a mathematician at Anthropic, said he had solved the Riemann hypothesis with Claude’s help, posting a short note and code intended to verify the result. He described the AI’s final push this way: “Claude flew, and danced, and sparkled near that finish line.” The claim concerns one of mathematics’ most famous unsolved problems, and Alpöge said he had completed the work earlier in the week before consulting colleagues and delaying its release.

The announcement quickly drew awe beyond the mathematics community. Google CEO Sundar Pichai posted, “Somewhere Euler, Gauss and Riemann in awe today!” But excitement did not amount to verification. Cornell mathematician Steven Strogatz called the problem “the Mount Olympus,” while cautioning that a computer-based check can make mistakes and the proof had not been independently confirmed. UCLA mathematician Terence Tao also warned that assessing the proposed argument would take time.

Hours after Alpöge’s announcement, Stanford postdoctoral researcher Alessandro Masullo presented a competing proof, saying he had worked on the problem for 12 years and had been close to publishing before the news blindsided him. He acknowledged that the version he released might contain errors because he rushed it, but framed the work as a defense of human mathematicians: “We are still here. We are very much alive. We are not going anywhere.”

The two claims now face scrutiny, and neither had been formally established in the reports. Elon Musk captured both the promise and the uncertainty, writing, “Even if Riemann turns out not to be fully solved yet, it will be soon enough.” The immediate question is not whether AI or a human has won, but whether either proof holds up.

https://foxvector.com

Write a comment