• FaceDeer@fedia.io
    link
    fedilink
    arrow-up
    11
    arrow-down
    7
    ·
    2 months ago

    As recent advances in AI have shown, humans are really quite predictable when you throw enough data and compute at the problem. At some point the algorithm will be sophisticated enough that it’ll be able to get to know you better than you know yourself, and will be able to provide you with things you had no idea were what you really wanted.

    Interesting times.

      • FaceDeer@fedia.io
        link
        fedilink
        arrow-up
        9
        arrow-down
        1
        ·
        2 months ago

        Yes, but recent advances have really rubbed it in our faces in ways that are a lot harder to deny. Humans haven’t become fundamentally more or less predictable over time but recent advances have shown how predictable we are.

        • MajorHavoc@programming.dev
          link
          fedilink
          arrow-up
          2
          ·
          2 months ago

          Yep. I learned from an algorithm that I might enjoy music by “The Beatles”. The algorithm was quite correct, but I think my having simple tastes, and the Beatles having amazing music is due most of the credit.

      • Chris@feddit.uk
        link
        fedilink
        English
        arrow-up
        7
        ·
        2 months ago

        Yes, I heard/saw/read that this is exactly what Amazon do, some years back now. They know who you are, what stage of life you are at, and they know what you want before you do.

    • Schmeckinger@lemmy.world
      link
      fedilink
      arrow-up
      6
      ·
      2 months ago

      Yeah algorithms keep throwing stuff at me I would probably like to watch, but I don’t click on it to not get even more brain damage.

    • MajorHavoc@programming.dev
      link
      fedilink
      arrow-up
      2
      ·
      edit-2
      2 months ago

      I had this exact experience with music algorithm recommendations:

      The algorithm analyzed all the songs I asked it to play, and concluded (correctly) that I might enjoy listening to the Beatles. (True story.)

      (Now a bit of sarcasm:) I look forward to future insights, in other art forms, such as perhaps the writings of Shakespeare or the paintings of Leonardo Da Vinci.

    • Krauerking@lemy.lol
      link
      fedilink
      arrow-up
      1
      ·
      2 months ago

      Yeah doubtful. I think it finds something you will engage in and push on it over and over again until people get normalized to it.

      I think it’s more like cold reading from a psychic. It’s gonna use generic generalized data about the big identifiers for you like age and gender and as you respond try to change its answer to what it needs to based on what you gave it.

      That’s not new or magical in any way. And it can be really wrong about the broad stuff if you don’t fit in with generic identifying groups related to you.

      It really just feels like a sales pitch for the middle class to buy more stuff.

    • queermunist she/her@lemmy.ml
      link
      fedilink
      arrow-up
      1
      ·
      2 months ago

      That is not what happened.

      Humans aren’t static. You don’t actually have these secret hidden likes AI can discover, instead, you grow to like the stuff that becomes familiar. You’re being trained.

    • Ephera@lemmy.ml
      link
      fedilink
      arrow-up
      1
      ·
      2 months ago

      Problem is that none of the algorithms actually care about showing you things you like.

      Ads try to sell you on things that you wouldn’t otherwise buy. Occasionally, they may just inform you about a good product that you simply didn’t know about, but there’s more money behind manipulating you into buying bad products, because it’s got a brand symbol.

      And content recommendation algorithms don’t care about you either. They care about keeping you on the platform for longer, to look at more ads.
      To some degree, that may mean showing you things you like. But it also means showing you things that aggravate you, that shock you. And the latter is considered more effective at keeping users engaged.