AI hallucination of Chinese nuclear components almost led to US military attack
arstechnica.com
But the military's overall use of AI seems to be accelerating.
    • Dagnet@lemmy.worldEnglish
      21·
      20 days ago

      “AI said that country has weapons of mass destruction! We need to invade right now!”

    • Viking_Hippie@lemmy.dbzer0.comEnglish
      391·
      20 days ago

      🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈🎈

  • Gerudo@lemmy.zipEnglish
    37·
    20 days ago

    How would AI even have the info to pull together something like this? Are they literally feeding chatgpt military reports?

    • chicken@lemmy.dbzer0.comEnglish
      33·
      20 days ago

      CNN’s report said the analyst in question used a chatbot to analyze intelligence reports regarding the Chinese ship’s manifest, leading to the near-disastrous result

      Last December, the Department of Defense announced it would use Google’s Gemini for Government as the basis for its bespoke “GenAI.mil” platform. Last month, the department added Grok for Government as an option on the platform. Anthropic also offers a customized version of Claude for US spy work.

      Basically

      • tigeruppercut@lemmy.zipEnglish
        17·
        20 days ago

        the department added Grok for Government

        Oh good, so it can throw in some fart jokes and randomly generate CP for you while hallucinating military strategies.

    • limelight79@lemmy.worldEnglish
      19·
      20 days ago

      I assume so, yes. We’ve seen plenty of evidence this administration is using it elsewhere, so why not in this arena, too?

    • Eq0@literature.cafeEnglish
      14·
      20 days ago

      I have some very secret intelligence. What should I do wit it? Feed it to ChatGPT so it can train its next models on it!

  • Senator Collins@aussie.zoneEnglish
    10·
    21 days ago

    Well, there are a lot of these ships going around the world all the time, and very seldom does anything like this happen. I just don’t want people thinking that freighters aren’t safe.

  • Aceticon@lemmy.dbzer0.comEnglish
    6·
    20 days ago

    So AI might actually destroy the World.

    Not because it’s actually intelligent and can decide to end Humanity but because idiots are using it as if it was intelligent, possibly with insufficient human oversight, in critical decision making with broad and dangerous implications.