Also, sycophantic models can potentially harm users by creating a silo or echo chamber for of ideas.

That’s certainly a step in a direction. I tried a few chatbots during the late-2023 onslaught, and they were already too relentless with praise while also being quick to point out wrongthink. It’s hard to take any output seriously if you have any sense of reality.

  • Randomgal@lemmy.ca
    link
    fedilink
    arrow-up
    1
    ·
    20 hours ago

    Fucking skill issue. Just “Focus on the objective information, don’t be conversational.”

  • TranquilTurbulence@lemmy.zip
    link
    fedilink
    English
    arrow-up
    18
    ·
    edit-2
    2 days ago

    You could ask about the most boring thing possible, and the answer is usually something like:

    Great question! [Insert verbose flattery and an answer that goes into way too much detail] Isn’t it fascinating how long it takes for paint to dry on the wall? When was the last time you painted a wall? [Insert a random chatty question that derails the conversation from the original topic.]

    When using ChatGPT, you should totally tweak the settings to make sure your GPT experience does not include any of this nonsense.

  • mbirth@lemmy.ml
    link
    fedilink
    English
    arrow-up
    9
    ·
    2 days ago

    I’ve recently watched this video on YT about some guy that replaced all his social interactions with AI models for for a while. And he, too, said that his ChatGPT “best friend” was too friendly and later in the test preferred some Character.AI “dominatrix” model instead.

    • Pete Hahnloser@beehaw.orgOP
      link
      fedilink
      English
      arrow-up
      1
      ·
      20 hours ago

      I sincerely question whether your standard ChatGPT user knows what a Domme actually is. You don’t get that from movie descriptions. You get something – but nothing approaching BDSM best practices.