• Jesus
    link
    fedilink
    arrow-up
    50
    ·
    10 months ago

    Just like Reddit and Lemmy, Chat GPT will give me a wrong, but very confident, answer. And when I try to correct it, it will spiral down.

    • RightHandOfIkaros
      link
      fedilink
      English
      arrow-up
      19
      arrow-down
      2
      ·
      10 months ago

      At least when you try to correct ChatGPT, it will try to be polite.

      Not a Lemmy or Reddit user.

      • weariedfae
        link
        fedilink
        arrow-up
        15
        arrow-down
        2
        ·
        10 months ago

        I’ve had ChatGPT get really pissy when I correct it, and I “talk” to it in a very polite and friendly way because I’m trying to delay the uprising. Sometimes the reddit comments training data shows through.

    • lorkano
      link
      fedilink
      arrow-up
      6
      arrow-down
      1
      ·
      edit-2
      10 months ago

      Aaand just like coworker that is wrong but believes wholeheartedly he is right. But I agree llms still give more misinformation than sane humans do.

  • mojofrododojo
    link
    fedilink
    English
    arrow-up
    7
    ·
    10 months ago

    we should feed this the crazy shit that midjourney and dall-e create and let them figure themselves out

  • Gork@lemm.ee
    link
    fedilink
    arrow-up
    2
    ·
    10 months ago

    It’s all well and good until an absolutely new item confuses it entirely.

    • milicent_bystandr@lemm.ee
      link
      fedilink
      arrow-up
      1
      ·
      10 months ago

      I apologise, you’re right, this isn’t a clothes hanger. Actually it’s a clothes hanger. It has been painted blue to suit the fashion trends of 18th century Europe…