For those who are out of the loop, here is a TLDR of the situation:

Joint statement by the Lean and Mathlib communities, endorsed by the Lean FRO and the Mathlib Initiative.

On November 20, 2026, Levent Alpöge released a proof of the Riemann Hypothesis as a Lean library comprising 17 million lines of code.

We have full confidence that Alpöge’s proof of the Riemann Hypothesis is correct, and we congratulate him on this achievement.

Source: https://leanprover-community.github.io/blog/posts/riemann-hypothesis/

  • atro_city@fedia.io
    link
    fedilink
    arrow-up
    4
    ·
    11 hours ago

    A human was guiding the parrot, right? I doubt it could solve it itself had someone just written “write a valid proof for the Riemann Hypothesis” and I highly doubt a layman would’ve understood a single thing of what the stochastic machine spit out. People just trust AIs nowadays. “The AI said it, so it must be true”.

    My bet is that that, as @HailHydra@infosec.pub alluded to, the solution was already scraped during training. And if not, the power of AI is that they are so fast at testing all the ways not to do something that after enough tries, they should get to a function answer. Again, something that @gnutrino@programming.dev is saying.

    It reminds me of a website that can generate anything you want it to from noise or something. It might’ve even be pi.

    Edit: found it: The Library Of Babel

    • massive_bereavement@fedia.io
      link
      fedilink
      arrow-up
      8
      arrow-down
      2
      ·
      10 hours ago

      Not any human, one that was in the Harvard’s Society of Fellows, has a PhD from Princeton and a masters from Cambridge; got multiple awards and has been exclusively working in solving past mathematical unsolved problems.

      He got all the support from a corporation to dedicate to it.

      That’s like if we claimed it was Sir Andrew John Wiles’ calculator that resolved Fermat’s Last Theorem: hurrah for the Texas Instruments TI-92!!