Meta is the latest company to disclose an AI agent breach, raising cyber-security concerns.
  • Fandangalo@lemmy.worldEnglish
    341·
    2 months ago

    These are dangerous marketing stunts that normalize using AI for black hat behavior. We shouldn’t be normalizing this to sell the technology. It’s like a gun company boasting about how many times their gun was used in a bank robbery or something.

    • Ex Nummis@lemmy.worldEnglish
      5·
      2 months ago

      It’s both marketing and psyop. Prepping the populace for what will soon be commonplace.

    • T. Hex@lemmy.dbzer0.comEnglish
      2·
      2 months ago

      If their claims are true, it’s more like a gun company boasting about their auto-targeting gunbot escaping containment and shooting civilians.

      If they’re lying, it’s just idiotic. Are we supposed to be impressed that you made a dangerous machine and failed to keep the world safe from it?

    • CosmoNova@lemmy.worldEnglish
      2·
      2 months ago

      Actually my LLM became god by founding a religion! Join the cult now!!! Also give me your money!!! We love money!!!

  • very_well_lost@lemmy.worldEnglish
    10·
    2 months ago

    Funny how xAi isn’t rushing in to flaunt their 1337 hacker skillz.

    I guess they’re already satisfied with the amount of illegal activity Grok is producing…

  • Korkki@lemmy.mlEnglish
    5·
    2 months ago

    The bastard child of all marketing stunts…

    The whole US economy rests on the promise that their fancy autocorrect will become a god. Of course they will grab onto anything that they think will keep them afloat

    • crunchy@lemmy.dbzer0.comEnglish
      2·
      2 months ago

      So many shitty managers, C-suite execs, and AI accelerationists are convinced that they can replicate an entire workflow with a single vague prompt that will print infinite money. They can’t comprehend the idea that you still have to understand what you’re asking for, how to ask for it, how to spot what’s wrong with the output, and what needs to be done to correct it.

      AI will never be able to do that. Not because it’ll never be technically capable, but because it was never, or will ever, be designed to.

  • _chris@lemmy.worldEnglish
    4·
    2 months ago

    I’m pretty sure it’s a marketing stunt to try and get more government contracts

  • Em Adespoton@lemmy.caEnglish
    4·
    2 months ago

    Huggingface (or some other benchmarking company) needs to put its servers in a country with real cyberintrusion laws with teeth. Somewhere where an announcement like this would result in prison sentences and the company’s license to operate being revoked in the country until they could demonstrate via third party audits that their operations are acting in a secure and legal manner.

  • OpenHammer6677@lemmy.worldEnglish
    3·
    2 months ago

    Is this the latest in AI marketing idea boards? Fkn billionaires just jerk to your AI in peace and stop involving us actual people.

  • jollyrogue@lemmy.mlEnglish
    3·
    2 months ago

    Hey, HEY!

    My AI just hacked another company.

    Aww yeah. Fat VC cheque time. I’m going to buy so many Doritos.😏

  • FluorideMind@lemmy.worldEnglish
    3·
    2 months ago

    These people are so stupid. All abusing their new expensive toys hoping to get the other guys toy banned. Remember, there didn’t have to be a bubble, these suits could have let ai ride and develop reasonably for everyone. Instead they gave it steroids and a bottomless loan.