Researchers convinced ChatGPT to do things it normally wouldn’t with basic psychology.

Researchers convinced ChatGPT to do things it normally wouldn’t with basic psychology.

  • troed@fedia.io
    56·
    1 month ago

    All the hatred against LLMs really misses one of the huge and quite unexpected findings - like this article. These LLMs “function” very similar to human brains.

    • usernameusername@sh.itjust.worksEnglish
      41·
      1 month ago

      Yes I definitely do believe that LLMs are very close to a reverse engineering of the human brain and that human brains work based on language