dataqbs

[AINews] Riemann Hypothesis proved...twice independently

· Source: Latent Space

In recent hours, two researchers announced independent tests of the Riemann Hypothesis, the most renowned of the Millennium Prize Problems in mathematics. Levent Alpöge, affiliated with Anthropic, presented a proof generated by the company’s Claude model, accompanied by a 1.3 GB Lean file equivalent to roughly 250,000 pages of text. Analyst Elliot Glazer noted that the formalization corresponds to the Generalized Riemann Hypothesis (GRH) and that the likelihood of structural errors in the code is below 1 %. This suggests that, if the final compilation completes without adding extra axioms, the proof could be valid even though no human has reviewed it in its entirety.

At the same time, Alessandro Masullo released a 940‑page manuscript he wrote entirely over the past twelve years, describing a “work in progress” proof that also targets the GRH. Unlike the AI‑assisted effort, Masullo’s submission is fully readable, though experts warn it may contain technical flaws.

Both projects continue the recent wave of AI‑assisted mathematics, following the contentious Navier‑Stokes claims and the quasi‑Riemann limits. The community still needs several years to formally validate these results.

The significance lies in the potential first solution of a Millennium Prize Problem achieved through artificial intelligence, which would reshape how mathematical research is conducted and open new avenues for human‑machine collaboration in highly complex domains.

Read the original article on Latent Space

This summary is an informational synthesis produced by dataqbs.com. All rights to the original content belong to its author and the cited media outlet. We act solely as curators of technology news and claim no authorship.

Read this in Español · Deutsch · Français · Italiano