• Maxxie@piefed.blahaj.zone
    link
    fedilink
    English
    arrow-up
    5
    ·
    6 days ago

    The article touches on it, but I cannot see how would you fundamentally solve the safety issue. It will be heavy (frame and batteries) and strong (to be useful as well as move around all these batteries), comparable to a human.

    Who in their right mind would have a thing that can ran out of power and crush your dog, choke a baby when it though it was a pillow or just like… grab a knife and stab someone cause today it feels like Mechahitler.

    • shalafi@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      1
      ·
      6 days ago

      I can program my firewall to shut down gracefully when the UPS signals power is down to 2 minutes. That bit is a solved problem.

    • tomiant@programming.dev
      link
      fedilink
      English
      arrow-up
      1
      ·
      6 days ago

      Just make it light and weak. Problem solved. But then how would we make the robot safe from humans? Humans are heavy and strong, and could pass out, and crush the robot.

      • Maxxie@piefed.blahaj.zone
        link
        fedilink
        English
        arrow-up
        1
        ·
        6 days ago

        humanoid robot can’t be light and weak in the current century, too much stuff is involved in being a humanoid

        But then how would we make the robot safe from humans? Humans are heavy and strong, and could pass out, and crush the robot.

        I think an average person is way more invested in their relatives than in a roomba, that’s the fucked up world we live in :(

    • morphballganon@lemmynsfw.com
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      1
      ·
      6 days ago

      “The fascists’ version of AI is another fascist, so all AI must also be fascists”

      Bruh…

      They could be made to recognize when power is low and move to a safe position to avoid crushing/falling.

      But if you write them off because LLMs suck right now (which is completely irrelevant), then yeah, we’ll never get there. Thanks.

    • FizzyOrange@programming.dev
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      1
      ·
      6 days ago

      Well, it depends how reliable it is. Anyone who has lived with a flatmate has lived with a thing that could potentially kill your dog/baby/you because of whatever they were feeling. It’s mostly fine because most people are unlikely to actually do that.

      If they develop AI that is very reliable then sure, I’d be ok with it. That’s a big if, but it’s premature to say it’s never going to happen.

        • FizzyOrange@programming.dev
          link
          fedilink
          English
          arrow-up
          3
          ·
          6 days ago

          Right… but I don’t think anyone is talking about putting today’s robots in their home. That’s obviously crazy. They can’t really do anything useful.

          This is a hypothetical question about 10-50 years from now.

          • Nightwatch Admin@feddit.nl
            link
            fedilink
            English
            arrow-up
            1
            arrow-down
            1
            ·
            6 days ago

            Hypothetically speaking… ok. But unless someone invents portable anti-grav, bipedal humanoid robots are bad design even in 50 years. They will still be top heavy because of batteries, and electro motors in their limbs with sufficient power, and last but not least a proper heavy head.