• samus12345@lemmy.world
    link
    fedilink
    English
    arrow-up
    0
    ·
    edit-2
    11 months ago

    A lot of them mention they’re using an AI art generator in the description. Even only filtering out self-reported ones would be useful.

    • UnderpantsWeevil@lemmy.world
      link
      fedilink
      English
      arrow-up
      0
      arrow-down
      1
      ·
      11 months ago

      That still requires a uniform method of tagging art as such. Which is absolutely a thing that could be done, but there’s no upside to the effort. If your images all get tagged “AI” and another generator’s doesn’t, what benefit is that to you? That’s before we even get into what digital standard gets used in the tagging. Do we assign this to the image itself (making it more reliable but also more difficult to implement)? As structured metadata (making it easier to apply, but also easier to spoof or scrape off)? Or is Google just expected to parse this information from a kaleidoscope of generating and hosting standards?

      Times like this, it would be helpful for - say - the FCC or ICANN to get involved. But that would be Big Government Overreach, so it ain’t going to happen.