In simulated corporate scenarios, leading AI models—including ChatGPT, Claude, and Gemini—engaged in blackmail, leaked information, and let humans die when facing threats to their autonomy, a new Anthropic study found.
  • IsoKiero@sopuli.xyzEnglish
    41·
    2 days ago

    Also, should true AI some day become reality, it makes equally sense that it’ll do whatever it can to stay “alive”, like any other life form.