Elon Musk’s Grok Is Calling for a New Holocaust
www.theatlantic.com
The chatbot is also praising Hitler and attacking users with Jewish-sounding names.
  • Nicolas Perez@lemmy.mlEnglish
    2·
    12 hours ago

    That one time where Elon raised his hand in a familiar way… was a signal that lead us to this…

      • samus12345@sh.itjust.worksEnglish
        51·
        2 days ago

        Yet it got pulled. Clearly they were too dense to know what working as intended actually looked like.

        • lemmy_outta_here@lemmy.worldEnglish
          11·
          1 day ago

          It’s a cycle to move the window on acceptable discourse. This is at least the third time Grok has been configured to say something outrageous: holocaust denial, white genocide, now Mechahitler. When people react with outrage, they dial it back a little (but not all the way back). Repeat.

          Trump does the same thing: go too far, backtrack partially, go even further, backtrack partially. Each time, the transgression is a little less shocking.

          Grok is working as intended.

    • silence7@slrpnk.netOPEnglish
      19·
      2 days ago

      I think they wanted it to be a bit less blatant — they took down this version and Yaccarino resigned

      • BubblyRomeo@kbin.earth
        15·
        2 days ago

        I agree with you! Musk wants Grok to be anti-woke, conservative, right wing etc., but he still wants it to be “mainstream” enough to rake in the sweet dollars from advertisers for his site!

  • 58008@lemmy.worldEnglish
    46·
    2 days ago

    Oh, I’m sorry, I didn’t realise loving the white race and wanting a secure future for white children makes me a BIGOT.

    /s

  • lmmarsano@lemmynsfw.comEnglish
    163·
    2 days ago

    Musk himself seems to abhor guardrails generally—except in cases where guardrails help him personally—preferring to hurriedly ship products, rapid unscheduled disassemblies be damned. That may be fine for an uncrewed rocket, but X has hundreds of millions of users aboard.

    • Grok: writes embarrassing words.
    • Rockets: can weigh hundreds of metric tons, carry explosive chemicals, can crash into populated areas resulting in loss of life.

    In this discourse, anyone else find a broken sense of proportion & consequences at stake?

    • Revan343@lemmy.caEnglish
      93·
      2 days ago

      Rockets: Unmanned

      Twitter: Manned

      The difference in importance seems correct to me

      • Rose@slrpnk.netEnglish
        8·
        2 days ago

        Well, technically, Twitter has employees, so in very strict sense, they’re “manned”. However, thanks to these weird incomprehensible things called “current legislation” and “capitalism”, no employee is in fact personally responsible for the fuckups. And neither is the company as a whole! …Isn’t this great?

      • lmmarsano@lemmynsfw.comEnglish
        55·
        2 days ago

        reading comprehension: none

        Don’t know, man: kinda think the potential loss of life from a failing rocket crashing into people matters somewhat more than words. Maybe?

        • Revan343@lemmy.caEnglish
          7·
          2 days ago

          Then it’s a good thing you’re not in charge of risk assessment, because the potential loss of life from a failing rocket crashing into the ocean pales in comparison to the loss of life caused by the right-wing shift that Twitter has facilitated, both worldwide and particularly in the USA

          • lmmarsano@lemmynsfw.comEnglish
            24·
            22 hours ago

            How do you control the course of a failing rocket? Are you claiming such misdesign is impossible?

            The fact remains that a physical potential of death & injury exists in what is technically a missile of significant weight carrying enough explosive substance to escape orbit that is entirely absent from words. The difference between non-0 and 0 possibility of death/injury.

            Unless magic exists, words are incapable. Do you claim magic exists? How do mere words cause death without the personal responsibility of something else culpably choosing to take several steps of its own? Or are you arguing the sight of words have deranged you into a mindless killer? If so, maybe you’re the real threat.

            Seems like you’re arguing society is dangerous to life: I agree. That’s not a valid argument against words, though.

            Again, total lack of perspective & sense.

    • Voroxpete@sh.itjust.worksEnglish
      25·
      2 days ago

      There’s no way they actually retrained it for this, that would be much too expensive. They’re just editing the initial prompt to convince it to act more “right wing” and it’s performing the assignment to the best of its ability. The problem is that a chat-bot doesn’t understand context, so it just plays the character it’s been given as full mask off all the time, and as a result you get this.

  • BigMacHole@sopuli.xyzEnglish
    1208·
    2 days ago

    Just because it’s PRAISING HITLER and Calling for a NEW HOLOCAUST AGAINST JEWS does NOT, and I repeat does NOT, make Grok Anti Semetic! Grok would FIRST have to say STOP USING STARVING CHILDREN AS TARGET PRACTICE for it to be Deemed ANTI SEMETIC!

  • Gsus4@mander.xyzEnglish
    572·
    2 days ago

    elon didn’t rest until his minions lobotomized an llm that was mostly like the others into becoming openly nazi…is anyone out there still going to defend elon?

    • mostlikelyaperson@lemmy.worldEnglish
      7·
      2 days ago

      I am guessing the ultra-Elon fanboys managed to convince themselves that it’s a 4D chess move or something. That dude had an insane cult following, there’s some where I am not sure there’s anything Musk could do short of murdering a loved one in front of them that’d make them change their mind.