• Buffalox@lemmy.world
    link
    fedilink
    English
    arrow-up
    197
    ·
    2 days ago

    the firm is also implementing a “pre-crime” approach, attempting to predict incidents before they happen. In some cases, that also means reporting suspects to police before a crime occurs.

    Absolutely dystopian and should be illegal. The chance of mistakes is sky high, and pursuing these very probable false alerts will most likely consume valuable resources, making police less efficient instead of more.

        • Obituarykidney@lemmy.world
          link
          fedilink
          English
          arrow-up
          24
          ·
          1 day ago

          I reckon it’ll be more like Nightcrawler where anthropic commits the crime so they can be the first to report on it and say “see how successful we are!?” And inflate the stock price further.

          But they’ll sell it as minority report.

          People that are shocked by anthropic being evil haven’t been paying attention. The CEO’s straight up admit on the regular that they see themselves as bad guys and post slop videos of themselves as like Homelander.

    • oce 🐆@jlai.lu
      link
      fedilink
      English
      arrow-up
      33
      arrow-down
      1
      ·
      edit-2
      2 days ago

      Isn’t it the continuation of what has been built since 2001-9-11 as illustrated by Snowden? Now accelerated by AI ™️. I wouldn’t be surprised if NSA and related have been massive contributors to LLM development considering the shit tone of data they have been amassing, probably without being able to use most of it yet.

    • CommanderCloon@lemmy.ml
      link
      fedilink
      English
      arrow-up
      17
      ·
      1 day ago

      In short, Anthropic reported a user for in-platform speech, and then refused to provide police with evidence of actual wrongdoing.

      They don’t need “predictive” bullshit to waste police resources

      • FlashMobOfOne@lemmy.world
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        1
        ·
        1 day ago

        There was also a case earlier this year of SnapChat’s automoderator reporting a woman for a DM she wrote where she jokingly mentioned shooting up a school, and the cops got there in 30 minutes.

        Wild to think that near-predictive crime stoppage is possible except LLM’s can understand nuance or context.

    • Waterpumpee@lemmus.org
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      1
      ·
      1 day ago

      Doing that would result in defamation law suits left and right. I guess this is not about single persons but mob behaviour rather. And i guess local police/intelligence already has those insights, anthrophic just makes it cheaper to aggregate the data.