Comment on Oops

• ignotum@lemmy.world ⁨2⁩ ⁨days⁩ ago

  So an AI that can identify mushrooms and also tell the user if a mushroom is too similar to a different dangerous mushroom to be identified with a high enough certanity for it to be safe, would be ethical?

  Then how can anyone claim that no such system can ever be created? That makes no sense

  source

  • floquant@lemmy.dbzer0.com ⁨2⁩ ⁨days⁩ ago

    A 2D visual representation is not the same as the real thing

    source

    • ignotum@lemmy.world ⁨2⁩ ⁨days⁩ ago

      So experts cannot identify mushrooms at all by looking at it?

      They might turn it around and look at it from different angles, but then just make an AI that takes in multiple images from different angles, maybe have it ask for different angles if it cannot see everything it needs to see.

      And if the experts use other senses besides vision, like smell and touch, just make an AI that says “it might be X or Y, only way to tell them apart is through the smell, so i can’t be sure”

      source

      • petrol_sniff_king@lemmy.blahaj.zone ⁨17⁩ ⁨hours⁩ ago

        just make an AI that says “it might be X or Y, only way to tell them apart is through the smell, so i can’t be sure”

        I love when idea people tell me to just do this or that as if it’s easy.

        source
  • lIlIlIlIlIlIl@lemmy.world ⁨2⁩ ⁨days⁩ ago

    It’s just anti-AI hate. They’re like flat-earthers

    source

    • FosterMolasses@leminal.space ⁨1⁩ ⁨day⁩ ago

      This is the most pot calling the kettle black statement I have ever seen lmao

      source
    • ignotum@lemmy.world ⁨2⁩ ⁨days⁩ ago

      Yeah, over the top AI hype is annoying, and there are many valid criticisms to be had with regard to how AI is being trained and used (mainly generative AI),
      but all this absolutist anti-AI nonsense beats everything

      source