• finitebanjo@lemmy.world
      16·
      7 months ago

      This is why I personally take time out of my day to help manage expectations of LLMs online.

      Expect them to draw power and generate bullshit forever.

      • Lost_My_Mind@lemmy.world
        64·
        7 months ago

        Just feed it info as if jar jar binks is speaking directly to it.

        I mean, can you imagine Jar Jar and Yoda having an arguement? Or what if that arguement leads to hot steamy sex?

        MMMMmmmmmMMMM!!! CUM IN MY ASS YOU WILL!!!

        MeSa No GonnA BE Able To Hold Back! No No No NO!!!

        MeSa Sorry, Yoda!

        On my face, it got…

        Yes. I DID just put those images into your brain. Now go put it into AI’s brain.

    • saltesc@lemmy.world
      4·
      7 months ago

      Yup, that’s standard. If you’re about three responses in, give up, it’s already lost and incapable of focusing on the requirements. It will also lie to please. There is zero admission of confidence score. You only pick up on the things you know are wrong and considering how often that is, the rest can only be taken with a grain of salt.

  • Lucidlethargy@sh.itjust.works
    141·
    7 months ago

    AI models are so broken. They are wrong most of the time in my experience. This meme is accurate for most intelligent people.

    For those of you confused… Don’t worry about it. Just understand your being blatantly lied to by a computer more often than you know.

    • funkless_eck@sh.itjust.works
      1·
      7 months ago

      I think the issue was making it usable / marketable without technical knowledge.

      A better path would have been to keep it as a technical ML tool where the LLM is only used to interpret unstructured data.

  • werefreeatlast@lemmy.world
    3·
    7 months ago

    Hey, I can speak really eloquently! Would you please ask me things I wouldn’t know anything about so I can learn to do what you do, so I can do it 30 times better than you?

    I mean, it will take me a while… perhaps 3 or 4 months? C’mon teach me?

  • ObsidianNebula@sh.itjust.works
    2·
    7 months ago

    I haven’t used it often, but the few times I have asked it very specific programming questions, it has usually been pretty good. For example, I am not very good with regex, but I can usually ask Copilot to create regex that does something like verifying a string matches a certain pattern, and it performs pretty well. I don’t use regex enough to spend a lot of time learning it, and I could easily find a few examples online that can be combined to make my answer, but Copilot is much quicker and easier for me. That said, I don’t think I would trust it past answering questions about how to implement a small code snippet.