1095
This is ridiculous (thelemmy.club)
submitted 1 week ago by to c/fuck_ai@lemmy.world
175 comments fedilink hide all child comments
 
you are viewing a single comment's thread
view the rest of the comments
[–] 14 points 1 week ago* (last edited 1 week ago) (4 children)

I think this labeling was probably done by AI.

  • permalink
  • fedilink
  • source
  • hideshow 4 child comments
    •
  • [–] 14 points 1 week ago* (3 children)

    Lots of “AI detector” tools for both images and text are AI, yes. People using them don’t realize the irony because everyone wants so badly to believe there’s a way to stop all the AI fakers.

    edit: please, someone come up with an actually reliable way to detect AI art, something that’s not based on probabilities that will be skewed by real artists having their art used to train AI

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 3 child comments
    •
  • [–] 2 points 1 week ago (2 children)

    For images, some photoshop forensics and a trained eye. Text is a bit trickier.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 7 points 1 week ago (1 child)

    Text is trickier but for now the major signs are

    • corporate speech
    • "it's not X — it's Y"
    • em dashes used in the wrong place

    There are others but i can think only of those rn

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 4 points 1 week ago*
    • lists of three
    • nattering on about a topic, but always feeling like there's not much substance
    • proper headings inside social media posts (very few human posters do this)
    • an emoji in each list item or heading (typically in the front)
  • permalink
  • fedilink
  • source
  • parent