• themurphy@lemmy.ml
    link
    fedilink
    English
    arrow-up
    27
    arrow-down
    2
    ·
    2 days ago

    Well, AI for drugs is not a bad idea. Only because they can go through possible strings very fast, and then come up with some solutions to try.

    AI is not in charge of testing, because how could it.

    But drugs developed by AI, SHOULD be royalty free. But we all know it wont.

        • FatCrab@slrpnk.net
          link
          fedilink
          English
          arrow-up
          1
          ·
          14 hours ago

          Then what’s the point of this article? Machine learning has been a critical aspect of drug discovery for a decade+ now. Anthropic is built on a cornerstone of transformer models. Presumably, they are wanting to implement a transformer based approach to drug discovery otherwise this is absolutely nothing new and that’s a TERRIBLE idea.

          • themurphy@lemmy.ml
            link
            fedilink
            English
            arrow-up
            1
            ·
            10 hours ago

            The point of the article is that they want to develop their own drugs. Not just hand over the model to drug companies for them to use and discover drugs.

            The method is not new. But it’s new that Antrophic will have a drug division.

  • fullsquare@awful.systems
    link
    fedilink
    English
    arrow-up
    21
    ·
    2 days ago

    Yeah they’ll make medbeds any day now. Or maybe it’s for their internal use only, because they’ll noticed bubble and don’t like what they’re seeing. I wonder if they’re foolish enough to vacuum up all data from companies that used their chatbots to come up with something hoping that no one will notice

    Every single bit of that is so fucking stupid, they’re replaying crypto playbook 1:1. Spot an industry that they think they’ll manage in, then try to leech off of it in purest display of rentseeking imaginable, coupled with techbros folk belief that since programming is soo haard, then as they can do that they can do everything else too, all with no plan, no labs, no specialists onboard or faintiest idea what they’re doing

  • partofthevoice@lemmy.zip
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    1
    ·
    2 days ago

    This is funny. Was it too hard to automate everything with AI, that you now need to explore other avenues?

    • CosmoNova@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      23 hours ago

      They‘re following the Musk model of falling up the ladder by declaring ever bigger and more ridiculous goals after the last venture failed. It has proven to be quite effective with investors for some reason.

  • zod000@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    5
    ·
    2 days ago

    Based on the crazy shit that Dario says in public, they probably already are making them and he got first dibs.