We once denied the suffering of animals in pain. As AIs grow more complex, we run the danger of making the same mistake

  • PonyOfWar@pawb.social
    link
    fedilink
    English
    arrow-up
    4
    ·
    18 hours ago

    Fundamentally impossible to know. I’m not sure how you’d even find a definition for “suffering” that would apply to non-living entities. I don’t think the comparison to animals really holds up though. Humans are animals and can feel pain, so of course the base assumption for other animals should be that they do as well. To claim otherwise, the burden should be to prove that they don’t. Meanwhile, Humans are fundamentally nothing like an LLM, a program running on silicon predicting text responses based on a massive dataset.

    • badgermurphy@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      14 hours ago

      I don’t see how it is impossible to know. Every component of a machine is a fully known quantity lacking any means of detecting damage or discomfort. Every line of code was put in place for a specific, known purpose, none of which include feeling or processing anything beyond what it IS specifically designed for.

      Creatures and machines bear some similarities, but even simple creatures are dramatically more complex objects than even the most advanced computers. None of their many interacting components were put there for a specific purpose and intention, and many are only partially understood, if at all. With a machine, we know what every bit and piece is for, and it has no purpose beyond the intended ones because that would be a waste and cost more.

      • multiplewolves@lemmy.world
        link
        fedilink
        English
        arrow-up
        3
        ·
        7 hours ago

        This is the right answer. Perhaps no one in this particular thread knows every component of a computer the way a hardware engineer who designed those components would, but the “mystery” is caused by ignorance and that ignorance isn’t shared by every person.

        People exist who know exactly how every single component of a computer does and does not function. Every component was created by humans. Biology remains only partially understood to all of humanity. Not so machinery.

    • tabular@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      2
      ·
      17 hours ago

      The important part is that it feels like something subjectively to be a living human. It’s easy to presume animals close to humans are like us to a degree, but all we know is what it’s like to be ourselves moment to moment. There’s no reason to deny an unalive system cannot also feel - we cannot test anything.

      • PonyOfWar@pawb.social
        link
        fedilink
        English
        arrow-up
        2
        ·
        16 hours ago

        Where do we draw the line though? Humans assign emotions to all kinds of inanimate things: plush animals, the sky, dead people, fictional characters etc. We can’t give all of those the rights of a conscious being, so we need to have some kind of objective way to look at it.

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

          If someone claims feeling in a mere concept (without a body in a location)… I would find it very difficult to take seriously. But I must admit that’s just my intuition.

          I see nothing special in human meat that couldn’t be be significantly replicated by electronics, software, gears, etc. Consciousness is an imergent property.

          I fear that non-human, conscious creatures must fight us for those rights.

    • eleitl@lemmy.zip
      link
      fedilink
      English
      arrow-up
      1
      ·
      18 hours ago

      If you model a given (from digitized neuroanatomy) biological organism with full details in an simulated environment, both the behavior and its internal information processing is inspectable.