• RedWizard [he/him, comrade/them]@hexbear.net
    link
    fedilink
    English
    arrow-up
    28
    arrow-down
    1
    ·
    2 days ago

    “We don’t believe robots will make people unemployed, but rather they’ll boost efficiency or take on tasks humans are unwilling to do - like exploring the vast universe or the ocean depths where people can’t go. Machines can assist us in that exploration,” Liang said.

    Then why robots human shaped? I mean that seriously, would it simply not be better to design robots for the task at hand, instead of designing robots that are human shaped?

    • ☆ Yσɠƚԋσʂ ☆@lemmygrad.mlOP
      link
      fedilink
      English
      arrow-up
      18
      ·
      2 days ago

      There are a lot of significant advantages to a single general-purpose platform versus designing custom robots for every task. You can use the same group of robots and assign them different jobs as needed. These robots can also adapt to new tasks and workflows as they emerge. Additionally, you can use the same set of replacement parts that work on any robot. Production costs are also much lower since you won’t need lots of factories for different robot types.

      • RedWizard [he/him, comrade/them]@hexbear.net
        link
        fedilink
        English
        arrow-up
        19
        arrow-down
        1
        ·
        edit-2
        2 days ago

        Yes, but in the context of what they’re saying, exploring space and the depths of the ocean, why humanoid robots? The world has specialized machines and platforms for that very thing right now.

        • Nacarbac [any]@hexbear.net
          link
          fedilink
          English
          arrow-up
          12
          ·
          edit-2
          2 days ago

          I think the text is misrepresenting - he didn’t specify humanoid in the quote. But running with it to try and speculate, hm, a humanoid bodyplan would be compatible with (the few) existing technologies and infrastructure - and as long as it isn’t limited to following human functionality for the limbs (especially in space or underwater) then it’s just a multiarmed drone that happens to fit in an acceleration couch or whatever. Not sure that’s enough of a benefit unless you intend actual humans to follow after.

          I guess it’d also be a good visual stunt to have a humanoid do stuff? At least for orbital and underwater stuff it could even be teleoperated in an immersive VR style - signal lag would make it a bit dreamlike, but having a humanoid automata build a sandcastle on the Moon or explore a habitat being printed by specialised drones would be extremely cool.

        • ☆ Yσɠƚԋσʂ ☆@lemmygrad.mlOP
          link
          fedilink
          English
          arrow-up
          9
          ·
          2 days ago

          As @Nacarbac@hexbear.net notes, the article not saying that it’s all going to be humanoid robots. That said, everything I noted above still applies. General purpose robots are versatile, and they’re going to be useful in many different context. Think of it this way, why do we make generic computer chips instead of making specialized chips for different kinds of tasks which could be much more optimized. We do that for some cases of course, but the backbone of computing is a generic chip that you can produce at scale and use for all kinds of tasks. Same logic applies here.

        • fox [comrade/them]@hexbear.net
          link
          fedilink
          English
          arrow-up
          12
          ·
          2 days ago

          Ocean? Fish robot. Everything in the ocean goes fish shaped if it needs to move through water quickly. Space? Probably a tapered cylinder or something. Maybe a fish shaped again.

    • ExotiqueMatter@lemmygrad.ml
      link
      fedilink
      English
      arrow-up
      8
      ·
      edit-2
      2 days ago

      Because with humanoid robots you don’t need to redesign the infrastructure to change between robots and human workers. There is less births and more workers are retiring than coming in? Start putting robots on your factory floor. University graduations are up and more peoples need jobs? Retire some robots and hire humans.

      It makes the composition of the workforce flexible, able to change in a few hours or days. Refitting a factory to add more non-humanoid robots or to remove them and make workstations meant for humans could take months or years because non-humanoid robots would need vastly different infrastructure than what humans need to work.