• Echo Dot@feddit.uk
    link
    fedilink
    English
    arrow-up
    14
    arrow-down
    2
    ·
    21 小时前

    I would have no problem with AI if it could be useful.

    The problem is no matter how many times I’m promised otherwise it cannot automate my job and talk to the idiots for me. It just hallucinates a random gibberish which is obviously unhealthful.

    • Electricd@lemmybefree.net
      link
      fedilink
      English
      arrow-up
      1
      arrow-down
      3
      ·
      edit-2
      6 小时前

      Prompt or model issue my dude

      Or you’re one of the few who have a pretty niche job

      Just things like different words or vocabulary, or helping with some code related knowledge, Linux issues… or even random known knowledge that you happen not to know

      • CheeseNoodle@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        36 分钟前

        I think LLMs are also just genuinely not as universally useful as expected. Everyone thinks it can automate every job except their own not because everyone thinks their job is special but because they know all the intricate parts of that job that LLMs are still really bad at.

        For instance AI could totally do my job at a surface level but it quickly devolves into deal breaking caveats which I am lumping into very broad categories to save time:

        • Output would look good (great even) but not actually be useable in most applications
        • Output cannot even begin to be optimized in the same way humans can optimize it
        • It would take more effort for to fix these than to just do the whole thing on my own to begin with
    • Regrettable_incident@lemmy.world
      link
      fedilink
      English
      arrow-up
      4
      ·
      16 小时前

      I’ve found it useful for a few things. I had had a song intermittently stuck in my head for a few years and had unsuccessfully googled it a few times. Couldn’t remember artist, name, lyrics (it was in a language I don’t speak) - and chatGPT got it in a couple of tries. Things that I’m too vague about to be able to construct a search prompt and want to explore. Stuff like that. I just don’t trust it with anything that I want actual facts for.

      • kossa@feddit.org
        link
        fedilink
        English
        arrow-up
        4
        ·
        edit-2
        5 小时前

        Yep, preparing a proper search is my use case. Like “how is this special screw called?”. I can describe the screw and tell the model to provide me a list of how that screw could be called.

        Then I can search for the terms of that list and one of the terms is the correct one. It’s way better than hoping that somebody described the screw in the same words in some obscure forum.

        But, is it worth to burn the planet, make RAM, GPUs, hard drives unaffordable for everybody and probably crash the world economy for a better screw search? I doubt it.

    • architect@thelemmy.club
      link
      fedilink
      English
      arrow-up
      4
      ·
      19 小时前

      It’s really good at answering customer questions for me, to be honest.

      But, I still have to okay it. Just in case. There’s no trust.

      However that still does take a lot less bandwidth for me because I’m not good at the customer facing aspects of my business.

    • Soup@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      19 小时前

      I still would, as the increased productivity, once again, does not lead to reduced hours. Always more productive, always locked into a bullshit schedule.