• magdalar@lemmy.world
    link
    fedilink
    English
    arrow-up
    17
    ·
    2 days ago

    I worry the same thing happening in all fields, especially medical. we train the models on experts, then give the newbies the model… so when and how do we improve and retrain these models, and ok what data?

    • MangoCats@feddit.it
      link
      fedilink
      English
      arrow-up
      2
      ·
      2 days ago

      The thing about medical is that it’s so big - no one expert can cover even a small fraction of “medical”. So, unless you have a whole team of specialists there on-hand, you just might benefit from having some “automated experts” available when you need the pattern recognition they can offer.

      • minfapper@piefed.social
        link
        fedilink
        English
        arrow-up
        4
        ·
        1 day ago

        The thing about software is that it’s so big - no one expert can cover even a small fraction of “software”. If you’re a kernel driver developer, you might benefit from having some “automated experts” available in CSS/HTML for the pattern recognition they can offer.

        I find it amusing how most experts think AI is great for every field except theirs.

        • MangoCats@feddit.it
          link
          fedilink
          English
          arrow-up
          1
          ·
          22 hours ago

          Personally, I find AI/LLM agents to be very useful in software development, dramatically improving the current imperfect practice of the art to something - not perfect - but quite a bit better than was practical / possible even one year ago.

          The MD I worked for spent six weeks researching problems a friend of his had before the friend died, if he could have done that same research in a day instead of six weeks, he (or any other M.D. doing the research) could have saved his friends life.

      • gothic_lemons@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        2 days ago

        Except the medical expert in field A can’t be the human in the loop to verify the AI medical slopPut for field B is correct, cuz the medical expert isn’t an expert in field B. A team of human specialists will be better in the long run.

        • MangoCats@feddit.it
          link
          fedilink
          English
          arrow-up
          1
          ·
          2 days ago

          A team of human specialists will be better in the long run.

          For the 0.1% of the population who can afford to assemble a team of human specialists in a timely fashion when they need them…

          For the rest of us, cyborg M.D.s with brain augmentation (they called this a PubMed search in the 1990s, it’s faster and more effective today) are better than the un-augmented M.D.s we can afford.