• REDACTED@infosec.pub
    link
    fedilink
    English
    arrow-up
    1
    ·
    7 minutes ago

    The idea features heavily in the widely read AI 2027 doomsday scenario of last year, which imagines AI agents designing more and more intelligent versions of themselves, one of which eventually kills all of humanity with a bioweapon in order to make room for more datacentres and solar panels.

    Fucking lunatics

    • fizzle@quokk.au
      link
      fedilink
      English
      arrow-up
      17
      ·
      2 hours ago

      Exactly.

      “We’re working on something so amazing you better hold us back! Honestly we’re going to release our AGI any day now, just a few more rounds of VC and boom.”

    • ryannathans@aussie.zone
      link
      fedilink
      English
      arrow-up
      4
      arrow-down
      3
      ·
      2 hours ago

      Anthropic was started from all the safety engineers that left OpenAI over safety concerns

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

        And then they immediately became as evil as OpenAI. More in fact, if such a thing is possible.

  • inari@piefed.zip
    link
    fedilink
    English
    arrow-up
    8
    ·
    2 hours ago

    Isn’t this basically Anthropic trying to shield itself from new market entrants?

  • db2@lemmy.world
    link
    fedilink
    English
    arrow-up
    21
    arrow-down
    1
    ·
    4 hours ago

    So that means anthropic made a predictive model that actually works and it predicted heads in a basket. “We aren’t the bad guys, pinky promise!”

      • zurohki@aussie.zone
        link
        fedilink
        English
        arrow-up
        5
        ·
        2 hours ago

        This. Big companies only want regulations so that they can be made onorous enough to force smaller companies out of business.

  • sorter_plainview@lemmy.today
    link
    fedilink
    English
    arrow-up
    10
    ·
    3 hours ago

    Unless they developed something radically different than, transformer architecture from Google and mixture of experts from Deepseek, this is just absolute bullshit. Just like they claimed Mythos is something out of this world, but it was just another step in the progress.

  • MyBrainHurts@piefed.ca
    link
    fedilink
    English
    arrow-up
    4
    arrow-down
    2
    ·
    3 hours ago

    1944: Japan says world should have pause on nuclear weapons.

    Like, yes, but at this point, your country opting in or out doesn’t change the destination, only who gets there.

    • shiftymccool@piefed.ca
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 hour ago

      The battle cry of every numb-nutted CEO trying to smash AI into every nook and cranny… “We’re gonna get left behiiiiiiinnnnd duuuurrrr”