Will Solving a legendary math problem with AI be a candidate for a Fields Medal?

6 points | by DivingForGold 3 hours ago

4 comments

  • DivingForGold 3 hours ago
    Should anyone solving a legendary math problem using a substantial amount of AI, be a candidate for a Fields medal ?
  • atmosx 2 hours ago
    It's going to be one of the hard questions the community will have to answer. It's a brand new world out there.
  • paulpauper 2 hours ago
    no. the evidence suggests AI contributed to the entirety of the proof. I say give it to Anthropic/Claude. It also easily meets the age requirement.
  • turtleyacht 2 hours ago
    The solver gets the medal, not their advisors, mentors, colleagues, teaching assistants, manuscript reviewers, correspondents, conference organizers, and so on.

    There's a whole infrastructure behind the work, but execution is what's recognized.

    Even if the medal is not awarded, the solution is now out there. It would be hard to prove independent development the longer time goes. Hence the hurried rush of concomitant drafts.

    Suppose computer-aided proofs are now a requirement. Math may not survive its DevOps moment unscathed. Everything now needs an IT component.

    (Or, where do we see a computer science department without a computer?)

    • DivingForGold 2 hours ago
      I posed this question to several PhD candidates after my Masters swim class at the university, I got back 50% opposed, 50% allowed to be a candidate. Apparently it could be a hot topic. Would be interesting to see how it scores on Kalshi prediction market.