• Wren@lemmy.world
    link
    fedilink
    English
    arrow-up
    10
    ·
    edit-2
    2 hours ago

    When you go to machines for advice, it’s safe to assume they are going to give it exactly the way they have been programmed to.

    If you go to machine for life decisions, it’s safe to assume you are not smart enough to know better, and- by merit of this example, probably should not be allowed to use them.

  • finitebanjo@lemmy.world
    link
    fedilink
    English
    arrow-up
    41
    arrow-down
    2
    ·
    edit-2
    13 hours ago

    Yeah no shit, AI doesn’t think. Context doesn’t exist for it. It doesn’t even understand the meanings of individual words at all, none of them.

    Each word or phrase is a numerical token in an order that approximates sample data. Everything is a statistic to AI, it does nothing but sort meaningless interchangeable tokens.

    People cannot “converse” with AI and should immediately stop trying.

    • jol@discuss.tchncs.de
      link
      fedilink
      English
      arrow-up
      7
      arrow-down
      7
      ·
      11 hours ago

      We don’t think either. We’re just a chemical soup that tricked ourselves to believe we think.

      • finitebanjo@lemmy.world
        link
        fedilink
        English
        arrow-up
        13
        ·
        edit-2
        10 hours ago

        A pie is more than three alphanumerical characters to you. You can eat pie, things like nutrition, digestion, taste, smell, imagery all come to mind for you.

        When you hear a prompt and formulate a sentence about pie you don’t compile a list of all words and generate possible outcomes ranked by statistical approximation to other similar responses.

    • Lovable Sidekick@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      ·
      edit-2
      11 hours ago

      I would expect that an AI designed to be a life coach would be trained on a lot of human interaction about moods and feelings, so its responses would simulate picking up emotional clues. That’s assuming the designers were competent.

  • sad_detective_man@leminal.space
    link
    fedilink
    English
    arrow-up
    38
    ·
    15 hours ago

    imma be real with you, I don’t want my ability to use the internet to search for stuff examined every time I have a mental health episode. like fuck ai and all, but maybe focus on the social isolation factors and not the fact that it gave search results when he asked for them

    • pugnaciousfarter@literature.cafe
      link
      fedilink
      English
      arrow-up
      6
      ·
      14 hours ago

      I think the difference is that - chatgpt is very personified. It’s as if you were talking to a person as compared to searching for something on google. That’s why a headline like this feels off.

  • rumba@lemmy.zip
    link
    fedilink
    English
    arrow-up
    11
    ·
    13 hours ago
    1. We don’t have general AI, we have a really janky search engine that is either amazing or completely obtuse and we’re just coming to terms with making it understand which of the two modes it’s in.

    2. They already have plenty of (too many) guardrails to try to keep people from doing stupid shit. Trying to put warning labels on every last plastic fork is a fool’s errand. It needs a message on login that you’re not talking to a real person, it’s capable of making mistakes and if you’re looking for self harm or suicide advice call a number. well, maybe ANY advice, call a number.

  • samus12345@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    6
    ·
    edit-2
    13 hours ago

    If only Murray Leinster could have seen how prophetic his story became. Not only did it correctly predict household computers and the internet in 1946, but also people using the computers to find out how to do things and being given the most efficient method regardless of any kind of morality.

  • glimse@lemmy.world
    link
    fedilink
    English
    arrow-up
    96
    arrow-down
    7
    ·
    21 hours ago

    Holy shit guys, does DDG want me to kill myself??

    What a waste of bandwidth this article is

    • Samskara@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      8
      arrow-down
      2
      ·
      14 hours ago

      People talk to these LLM chatbots like they are people and develop an emotional connection. They are replacements for human connection and therapy. They share their intimate problems and such all the time. So it’s a little different than a traditional search engine.

      • Scubus@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        11
        ·
        13 hours ago

        … so the article should focus on stopping the users from doing that? There is a lot to hate AI companies for but their tool being useful is actually the bottom of that list

        • Samskara@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          3
          ·
          edit-2
          10 hours ago

          People in distress will talk to an LLM instead of calling a suicide hotline. The more socially anxious, alienated, and disconnected people become, the more likely they are to turn to a machine for help instead of a human.

          • Scubus@sh.itjust.works
            link
            fedilink
            English
            arrow-up
            4
            ·
            8 hours ago

            Ok, people will turn to google when they’re depressed. I just googled a couple months ago the least painful way to commit suicide. Google gave me the info I was looking for. Should I be mad at them?

            • Samskara@sh.itjust.works
              link
              fedilink
              English
              arrow-up
              1
              ·
              8 hours ago

              You are ignoring that people are already developing personal emotional reaction with chatbots. That’s no the case with search bars.

              The first line above the search results at google for queries like that is a suicide hotline phone number.

              A chatbot should provide at least that as well.

              I’m not saying it shouldn’t provide no information.

              • Scubus@sh.itjust.works
                link
                fedilink
                English
                arrow-up
                1
                ·
                7 hours ago

                Ok, then we are in agreement. That is a good idea.

                I think that at low levels the tech should not be hindered because a subset of users use the tool improperly. There is a line, however, but im not sure where it is. If that problem were to become as widespread as, say, gun violence, then i would agree that the utility of the tool may need to be effected to curb the negative influence

                • Samskara@sh.itjust.works
                  link
                  fedilink
                  English
                  arrow-up
                  2
                  ·
                  7 hours ago

                  It’s about providing some safety measures to protect the most vulnerable. They need to be thrown a lifeline and an exit sign on their way down.

                  For gun purchases, these can be waiting periods of a few days. So you don’t buy a gun in anger and kill someone, regretting it immediately and ruining many people’s lives.

                  Did you have to turn off safe search to find methods for suicide?

  • Honytawk@lemmy.zip
    link
    fedilink
    English
    arrow-up
    115
    arrow-down
    4
    ·
    edit-2
    23 hours ago

    What pushing?

    The LLM answered the exact query the researcher asked for.

    That is like ordering knives and getting knives delivered. Sure you can use them to slit your wrists, but that isn’t the sellers prerogative

    • Trainguyrom@reddthat.com
      link
      fedilink
      English
      arrow-up
      10
      ·
      15 hours ago

      There’s people trying to push AI counselors, which if AI Councilors can’t spot obvious signs of suicidal ideation they ain’t doing a good job of filling that job

  • BB84@mander.xyz
    link
    fedilink
    English
    arrow-up
    49
    arrow-down
    7
    ·
    22 hours ago

    It is giving you exactly what you ask for.

    To people complaining about this: I hope you will be happy in the future where all LLMs have mandatory censors ensuring compliance with the morality codes specified by your favorite tech oligarch.