• coyotino [he/him]@beehaw.orgOP
    link
    fedilink
    English
    arrow-up
    2
    ·
    16 hours ago

    i guess my point is that I understand why the researchers are doing it - the UN gave them money to research ways the UN could use AI, so that is what they did. It’s not like the research is unethical in the sense that it directly harms participants. Maybe it’s a dumb waste of money, but at that point, the question is more for the UN leaders that said “we should give someone money to research AI”. And I don’t know that 404 Media has the pull to interview those people.

    • sleepundertheleaves@infosec.pub
      link
      fedilink
      arrow-up
      1
      ·
      edit-2
      9 hours ago

      the UN gave them money to research ways the UN could use AI, so that is what they did

      No, no, no, that’s not an excuse.

      If they were, in good faith, researching ways the UN could use AI, this fucking horrible idea would have been thrown out in the first round of brainstorming.

      This is a horrible idea. This is a stupid idea. We live in a world where most of the privileged wealthy West is desperate to pretend that refugees aren’t real, or don’t matter, or deserve to live in poverty. And creating fake AI refugees just gives the privileged wealthy West another way to excuse themselves, by dismissing what the AI says is fake, by telling themselves there aren’t any real people in situations that bad.

      If you’re getting to the point where you’re implementing an obviously horrible idea and asking for public feedback on it, you don’t get to blame the people who told you to come up with ideas. You should have thrown that bad idea out. You should not have implemented it. That’s on you.

    • Asetru@feddit.org
      link
      fedilink
      arrow-up
      3
      ·
      14 hours ago

      the UN gave them money to research ways the UN could use AI, so that is what they did.

      That’s kind of my point… They didn’t. To research ways the un could use ai, you could have workshops and interviews with various groups, experts and non-experts alike. You don’t just pick one, utterly insane use case (that is called out beforehand as such) and implement that. You do research on the options and pick either the best ones or, if there’s no good one, none!

      To come up with a research project, it has to go through various pitches, drafts and proposals. I can’t imagine every single control instance failing so utterly that this kind of project with this high school level of arguing (“well, we could do this, so why wouldn’t we?”) passes each of them. There has to be a better reason why they did this. And if there really isn’t, a lot of people should ask themselves what the fuck they’re getting paid for if they let this happen - and some other people if they’re the ones who should fire the former.