this post was submitted on 24 Oct 2023
66 points (80.0% liked)

Technology

59440 readers
4492 users here now

This is a most excellent place for technology news and articles.

Our Rules

  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots

founded 1 year ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
66
This new data poisoning tool lets artists fight back against generative AI (www.technologyreview.com)
submitted 1 year ago by [email protected] to c/[email protected]
9 comments fedilink hide all child comments
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 10 points 1 year ago (2 children)

If it works anything like the other supposed AI image protector tool I'm aware of (Glaze) then it's not gonna look great and I would not call it a practical way to go. Everything I've seen run through glaze looks objectively worse than the original.

Also in the long run this is just an arms race and it's just a matter of time before models learn to subvert these kinds of tools. And if that's the case that means every time someone figures out how to get over these hurdles, anyone looking to protect their images will have to go back and replace every online instance of those images when the protection tool comes out with a fix. Back and forth forever.

And that's just ridiculous and basically impossible when you realize that stuff gets reposted all over the net all the time and can't be controlled.

[–] [email protected] 14 points 1 year ago* (last edited 1 year ago) (1 children)

every time someone figures out how to get over these hurdles, anyone looking to protect their images will have to go back and replace every online instance of those images when the protection tool comes out with a fix.

And if those older versions got downloaded and saved by a trainer there's nothing at all they can do to replace those.

This all feels a lot like the DRM treadmill, which has never done much to actually prevent piracy. Just made things annoying for everyone else.

[–] [email protected] 2 points 1 year ago

Yep totally agree. It's a pointless effort to try to combat the issue of AI this way.

[–] [email protected] 6 points 1 year ago

Zhao’s team also developed Glaze,

from the article, so it's likely they run on similar principles.