• suction@lemmy.world
    link
    fedilink
    English
    arrow-up
    1
    ·
    6 months ago

    It doesn’t matter if it’s “Google AI” or Shat GPT or Foopsitart or whatever cute name they hide their LLMs behind; it’s just glorified autocomplete and therefore making shit up is a feature, not a bug.

    • interdimensionalmeme@lemmy.ml
      link
      fedilink
      English
      arrow-up
      1
      ·
      6 months ago

      Making shit up IS a feature of LLMs. It’s crazy to use it as search engine. Now they’ll try to stop it from hallucinating to make it a better search engine and kill the one thing it’s good at …

    • Johanno@feddit.de
      link
      fedilink
      English
      arrow-up
      0
      ·
      6 months ago

      Chatgpt was in much higher quality a year ago than it is now.

      It could be very accurate. Now it’s hallucinating the whole time.

      • Lad@reddthat.com
        link
        fedilink
        English
        arrow-up
        0
        ·
        6 months ago

        I was thinking the same thing. LLMs have suddenly got much worse. They’ve lost the plot lmao

          • Ben Hur Horse Race@lemm.ee
            link
            fedilink
            English
            arrow-up
            0
            ·
            6 months ago

            I’m not sure thats definitely true… my sense is that the AI money/arms race has made them push out new/more as fast as possible so they can be the first and get literally billions of investment capitol

            • Cringe2793@lemmy.world
              link
              fedilink
              English
              arrow-up
              0
              ·
              6 months ago

              Maybe. I’m sure there’s more than one reason. But the negativity people have for AI is really toxic.

              • Ben Hur Horse Race@lemm.ee
                link
                fedilink
                English
                arrow-up
                1
                ·
                5 months ago

                is it?

                nearly everyone I speak to about it (other than one friend I have who’s pretty far on the spectrum) concur that no one asked for this. few people want any of it, its consuming vast amounts of energy, is being shoehorned into programs like skype and adobe reader where no one wants it, is very, very soon to become manditory in OS’s like windows, iOS and Android while it threatens election integrity already (mosdt notibly India) and is being used to harass individuals with deepfake porn etc.

                the ethics board at openAI got essentially got dispelled and replaced by people interested only in the fastest expansion and rollout possible to beat the competition and maximize their capitol gains…

                …also AI “art”, which is essentially taking everything a human has ever made, shredding it into confetti and reconsstructing it in the shape of something resembling the prompt is starting to flood Image search with its grotesque human-mimicing outputs like things with melting, split pupils and 7 fingers…

                you’re saying people should be positive about all this?

                • Cringe2793@lemmy.world
                  link
                  fedilink
                  English
                  arrow-up
                  0
                  arrow-down
                  1
                  ·
                  5 months ago

                  You’re cherry picking the negative points only, just to lure me into an argument. Like all tech, there’s definitely good and bad. Also, the fact that you’re implying you need to be “pretty far on the spectrum” to think this is good is kinda troubling.