• FishFace ( FishFace@piefed.social ) 
    link
    fedilink
    English
    arrow-up
    2
    ·
    2 hours ago

    This is extremely dumb. There are still people doing mathematics without AI, and doing mathematics with AI no more robs you of all your eurekas than collaborating on a problem with someone else robs you of all your eurekas; your collaborator (or the AI) will work some stuff out, and you will work other stuff out.

    • eah ( eah@programming.dev ) OP
      link
      fedilink
      English
      arrow-up
      1
      ·
      12 minutes ago

      Nah, it’s definitely robbing us of joyful things to do. There’s far less motivation to do math on your own if computers are so much faster. The next rebuttal will be “oh so you think computers took away joy by mechanizing arithmetic calculations” and to that I would say no because while probably there were interesting techniques used to do calculation by hand quickly which are no longer useful, there were far more, interesting problems to solve created by the invention computers, as well as better efficiency and new forms of art. In any case, computers are now threatening to generate literally everything else for which there exists a sufficient quantity of training examples in the form of bits, and I fear we won’t get anything new from that, just a higher quantity of what we can happily produce ourselves, at possibly lower quality, speaking about creative work more generally, be that music, video games, visual art, or writing.

      I don’t see very much point to racing ahead solving problems, a vast number of which don’t have practical applications in the real world. We could have left it all alone and allowed eons more for mathematicians to ponder. Give it enough time and we would figure it out ourselves, and maybe even force ourselves to invent new techniques that might otherwise go undiscovered. I think long term AI is an unnecessary crutch that will cause us to atrophy. AI proofs will prime mathematicians to think a certain way. We learn things on dead end paths.

  • Wirlocke ( Wirlocke@lemmy.blahaj.zone ) 
    link
    fedilink
    English
    arrow-up
    1
    ·
    2 hours ago

    For the most part the AI math proofs discovered so far have only served to show us what we’ve been blinded by precedent and bias to see.

    The pattern forming seems to be that AI attacks a problem in a clunky brute force manner until it technically disproves a widely held conjecture or hypothesis, then that invigorates real mathematicians to both improve upon it and redo what the AI did in a way that provide more meaning and connects to other concepts in mathematics.

    I think this is the most optimistic outcome for AI in mathematics. Having AI break down unintuitive barriers and having mathematicians explore afterwards, acting more like philosophers than engineers.

    • eah ( eah@programming.dev ) OP
      link
      fedilink
      English
      arrow-up
      1
      ·
      13 minutes ago

      So that’s the situation right now. Will that always be the case or will AI output eventually be literally indistinguishable from what a human would produce? If you give even a single inch to this concept that AI is going to help, you may eventually find that the humans don’t add anything.

    • factually incorrect. search engines are used to provide you the information you need to perform thinking on. ai does the search, condenses the data, and provides “insight”. whatever is the quality, most people are abusing that to outsource their thoughts. any research or thinking or calculation most people will outsource to ai. maybe not your friends in your online circle, but most people irl definitely do.