• hoshikarakitaridia@lemmy.world
    link
    fedilink
    arrow-up
    14
    arrow-down
    7
    ·
    2 days ago

    Because in a lot of applications you can bypass hallucinations.

    • getting sources for something
    • as a jump off point for a topic
    • to get a second opinion
    • to help argue for r against your position on a topic
    • get information in a specific format

    In all these applications you can bypass hallucinations because either it’s task is non-factual, or it’s verifiable while promoting, or because you will be able to verify in any of the superseding tasks.

    Just because it makes shit up sometimes doesn’t mean it’s useless. Like an idiot friend, you can still ask it for opinions or something and it will definitely start you off somewhere helpful.

    • WalnutLum@lemmy.ml
      link
      fedilink
      arrow-up
      12
      arrow-down
      2
      ·
      2 days ago

      All LLMs are text completion engines, no matter what fancy bells they tack on.

      If your task is some kind of text completion or repetition of text provided in the prompt context LLMs perform wonderfully.

      For everything else you are wading through territory you could probably do easier using other methods.

    • ms.lane@lemmy.world
      link
      fedilink
      English
      arrow-up
      11
      ·
      2 days ago

      Also just searching the web in general.

      Google is useless for searching the web today.

      • fibojoly@sh.itjust.works
        link
        fedilink
        arrow-up
        2
        arrow-down
        1
        ·
        2 days ago

        Not if you want that thing that everyone is on about. Don’t you want to be in with the crowd?! /s

      • dev_null@lemmy.ml
        link
        fedilink
        arrow-up
        2
        ·
        2 days ago

        Yes, but for some tasks mistakes don’t really matter, like “come up with names for my project that does X”. No wrong answers here really, so an LLM is useful.

          • dev_null@lemmy.ml
            link
            fedilink
            arrow-up
            3
            ·
            edit-2
            2 days ago

            And yet virtually all of software has names that took some thought, creativity, and/or have some interesting history. Like the domain name of your Lemmy instance. Or Lemmy.

            And people working on something generally want to be proud of their project and not name it the first thing that comes to mind, but take some time to decide on a name.

            • Rekorse@sh.itjust.works
              link
              fedilink
              arrow-up
              1
              ·
              2 days ago

              Wouldnt they also not want to take a random name off an AI generated list? How is that something to be proud of? The thought, creativity, and history behind it is just that you put a query into chatgpt and picked one out of 500 names?

              Maybe its just a difference of perspective but thats not only not a special origin story for a name, its taking from others in a way you won’t be able to properly credit them, which is essential to me.

              I would rather avoid the trouble and spend the time with a coworker or friend throwing ideas back and forth and building an identity intentionally.

              I suppose AI could be nice if I was alone nearly all the time.

              • dev_null@lemmy.ml
                link
                fedilink
                arrow-up
                1
                ·
                edit-2
                1 day ago

                The process of throwing ideas back and forth usually doesn’t include just choosing one, but generating ideas as jumping off points, usually with some existing concept in mind. Talking with friends, looking at other projects, searching for inspiration online and in the real world, and now also generating some more ideas with an LLM to add to the mix. Using one source and just picking a suggestion probably won’t get you a good result.