• Tracaine@lemmy.world
    link
    fedilink
    English
    arrow-up
    14
    ·
    25 days ago

    So if I tell it that I’m literally a slave who is beaten and sexually violated daily, it’s still going to tell me to stay in that situation? That seems helpful.

      • squaresinger@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        25 days ago

        Have a look at relationship subreddits. They are full of people who have been manipulated and gaslit for years or decades who have no idea what is actually normal. For people like that a reality check is really helpful or even vital.

  • kamenLady.@lemmy.world
    link
    fedilink
    English
    arrow-up
    13
    ·
    25 days ago

    The US company said new ChatGPT behaviour for dealing with “high-stakes personal decisions” would be rolled out soon.

    Fuck we’ve been in a black mirror episode all along

  • flamingo_pinyata@sopuli.xyz
    link
    fedilink
    English
    arrow-up
    12
    arrow-down
    1
    ·
    25 days ago

    Does anyone else think that LLMs have been regressing in the last 2 years. At least in the area of deep questions. Initially you could push the limits and actually bounce crazy ideas off of them and get some creative results.

    Now they basically they always give you a canned answer that sounds like it was pre-approved by a committee.

    • Repple (she/her)@lemmy.world
      link
      fedilink
      English
      arrow-up
      8
      ·
      25 days ago

      I believe one of Apple’s papers covered this, that chain of thought and other techniques that increase accuracy decrease perceived creativity. Memory is fuzzy though

    • Grimy@lemmy.world
      link
      fedilink
      English
      arrow-up
      6
      ·
      25 days ago

      The original 3.5 during the first month was peak creativity. The censorship really does ruin them.