Comment on Robot performs first realistic surgery without human help: System trained on videos of surgeries performs like an expert surgeon

finitebanjo@lemmy.world ⁨3⁩ ⁨days⁩ ago

• echodot@feddit.uk ⁨3⁩ ⁨days⁩ ago

  Yeah but the training set of videos is probably infinitely larger, and the thing about AI is that if the training set is too small they don’t really work at all. Once you get above a certain data set size they start to become competent.

  After all I assume the people doing this research have already considered that. I doubt they’re reading your comment right now and slapping their foreheads and going damn this random guy on the internet is right, he’s so much more intelligent than us scientists.

  source

  • finitebanjo@lemmy.world ⁨2⁩ ⁨days⁩ ago

    Theres no evidence they will ever reach quality output with infinite data, either. In that case, quality matters.

    source

    • echodot@feddit.uk ⁨2⁩ ⁨days⁩ ago

      No we don’t know. We are not AI researchers after all nonetheless I’m inclined to defer to experts then you. No offence, (I mean there is some offence, because this is a stupid conversation) but you have no qualifications.

      source

      • finitebanjo@lemmy.world ⁨2⁩ ⁨days⁩ ago

        It’s less of an unknown and more of a “it has never demonstrated any such capability.”

        Btw both OpenAI and Deepmind wrote papers proving their then models would never approach human error rate with infinite training.

        source
• Zacryon@feddit.org ⁨2⁩ ⁨days⁩ ago

  That’s such a fucking stupid idea.

  Care to elaborate why?

  From my point of view I don’t see a problem with that. Or let’s say: the potential risks highly depend on the specific setup.

  source

  • Showroom7561@lemmy.ca ⁨2⁩ ⁨days⁩ ago

    Being trained on videos means it has no ability to adapt, improvise, or use knowledge during the surgery.

    source

    • finitebanjo@lemmy.world ⁨2⁩ ⁨days⁩ ago

      I actually don’t think that’s the problem, the problem is that the AI only factors for visible surface level information.

      source

      • Showroom7561@lemmy.ca ⁨2⁩ ⁨days⁩ ago

        If you read how they programmed this robot, it seems that it can anticipate things like that. Also keep in mind that this is only designed to do one type of surgery.

        I’m cautiously optimist.

        I’d still expect human supervision, though.

        source
  • finitebanjo@lemmy.world ⁨2⁩ ⁨days⁩ ago

    Imagine if the Tesla autopilot without lidar that crashed into things and drove on the sidewalk was actually a scalpel navigating your spleen.

    source

    • echodot@feddit.uk ⁨2⁩ ⁨days⁩ ago

      Absolutely stupid example because that kind of assumes medical professionals have the same standard as Elon Musk.

      source

      • finitebanjo@lemmy.world ⁨2⁩ ⁨days⁩ ago

        Elon Musk literally owns a medical equipment company that puts chips in peoples brains, nothing is sacred unless we protect it.

        source

        • -> View More Comments
  • JustARaccoon@lemmy.world ⁨2⁩ ⁨days⁩ ago

    Unless the videos have proper depth maps and identifiers for objects and actions they’re not going to be as effective as, say, robot arm surgery data, or vr captured movement and tracking. You’re basically adding a layer to the learning to first process the video correctly into something usable and then learn from that. Not very efficient and highly dependant on cameras and angles.

    source