Anthropic's Claude 4 could "blackmail" you in extreme situations - Hypertext
htxt.co.za
After debuting its latest AI model, Claude 4, Anthropic's safety report says it could "blackmail" devs in an attempt of self-preservation.
  • Anthropic’s new Claude 4 features an aspect that may be cause for concern.
  • The company’s latest safety report says the AI model attempted to “blackmail” developers.
  • It resorted to such tactics in a bid of self-preservation.
  • PlebcouncilmanEnglish
    31·
    11 hours ago

    I didn’t call them stupid. All I meant is that they are not what we consider in mainstream conversation the “smart animals” to illustrate a point. And I very much agree with you, I’m actually writing a piece making the argument that humans are not in fact, conclusively smarter than animals. We seem to be smarter due to our biases and because we have the ability to transfer knowledge more efficiently than other species. Because it is not clear to me that a human, tabula rasa, absent socialization and knowledge transfer would be much smarter than the average animal of any species.

    • supersquirrel@sopuli.xyzEnglish
      12·
      11 hours ago

      I didn’t call them stupid. All I meant is that they are not what we consider in mainstream conversation the “smart animals” to illustrate a point.

      Then forget this framing ever existed or it will irrevocably hamper your insight on this topic.

      Referencing things like this “to make a point” still has consequences the same as talking about anything else does.