122
Make no mistakes (lemmy.world)
submitted 5 months ago by to c/memes@piefed.world
9 comments fedilink hide all child comments
๐Ÿ˜„ 1
 
you are viewing a single comment's thread
view the rest of the comments
[โ€“] 4 points 5 months ago (3 children)

I didn't think it was about judging. I thought it was about bypassing safety restrictions is it not? Basically learning how to smuggle drugs through indirect means

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [โ€“] 5 points 5 months ago* (2 children)

    Yea the action being presented is trying to bypass the safety guards, but the meme is a judgmental stare. Even if you want to be polite and call it a questioning stare, chatgpt doesn't do any of that itself, either.

    LLMs have no ability to think or question or judge at all. It's just correlation of words, and if some naughty words show up in the correlation, OpenAI has programmed other nanny processes to intervene and spit out other output.

    The meme is showing an emotional expression, but chatgpt's responses are about as emotional as a greyed out "OK" button on a form that takes about 20,000,000,000,000,000 times the processing power.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [โ€“] 3 points 5 months ago (1 child)

    but the meme is a judgmental stare.

    The meme is digital data, processed by my display to be a collection of pixels of different colors in a particular arrangement. My visual cortex may interpret the arrangement of pixels as a judgmental state of a 3D animated cartoon person, but the underlying digital data has no emotions and no judgment.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •