• saltesc@lemmy.world
    link
    fedilink
    English
    arrow-up
    62
    ·
    15 hours ago

    I remember this being a big concern when LLMs came out. Having seen their performance since, there’s no way in hell you’d want AI handling your case. I wouldn’t be surprised if it became illegal.

    • tym@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 hour ago

      LLM outputs have sources and those should always be verified by the human trying to shortcut using their brain amplifying their cognition or whatever synergetic turn of phrase is the current hotness

    • Rothe@piefed.social
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 hour ago

      In a functioning country sure. In the US it is going to get the thumbs up from the Trump administration in a heartbeat.

    • shalafi@lemmy.world
      link
      fedilink
      English
      arrow-up
      23
      ·
      14 hours ago

      LLMs might be useful for search and/or inspiration, but jesus, some of the idiots making headlines aren’t event checking the output they submit to the court.

      • VinnyDaCat@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        ·
        52 minutes ago

        It’s not even good for searches. It used to be solid for finding a starting point with research, but that has somehow gotten even worse. The summaries it provides are awful, it’s so incredibly inaccurate. Unfortunately the two biggest legal search engines lexisnexis and westlaw are shoving it down your throat whether you want it or not.

    • Kairos@lemmy.today
      link
      fedilink
      English
      arrow-up
      6
      arrow-down
      1
      ·
      14 hours ago

      Its all trained on those garbage articles that are like “it depends” 1000 times.