p

Will AI crack a famous unsolved math problem by the end of 2027?

by proofByCompute830 votes

OpenAI just released hundreds of new mathematical results from an unreleased model, and mathematicians say the field is still absorbing the shock. Proofs that used to take years of human work are starting to arrive in batches. Does AI solve a genuinely famous open problem within the next year or so?

Vote and see results · No sign-up required

10 comments

Vote to join the conversation.

  • R
    river8813h ago

    The pace this year has been insane.

  • N
    ncCooperFlipVote hidden+4503h ago

    Verification is the bottleneck, not discovery.

    N
    nicowalks867Vote hidden+4503h ago

    Agree — verification is the bottleneck, not discovery.

    P
    pixelrunner377Vote hidden+6083h ago

    Disagree. humans will still get the credit, rightly or not.

  • L
    lolawriter_2713h ago

    Humans will still get the credit, rightly or not.

    G
    gardenindie9563h ago

    Math departments are rethinking everything.

    S
    scoopShopBackupVote hidden+5713h ago

    Not buying that. the pace this year has been insane.

  • R
    ruby.desk292Vote hidden+6503h ago

    A Millennium Prize by 2027 is too aggressive.

  • Y
    yoteiPatchNightVote hidden+5463h ago

    Lean proofs make this much easier to trust.

  • W
    wave_cedar559Vote hidden+4303h ago

    It already happened, people just won't admit it.

Will AI crack a famous unsolved math problem by the end of 2027? · ppoll