Jaypg
  • Communities
  • Create Post
  • Create Community
  • heart
    Support Lemmy
  • search
    Search
  • Login
  • Sign Up
Lee Duna@lemmy.nz to Technology@lemmy.worldEnglish · 1 year ago

Nightshade, the free tool that ‘poisons’ AI models, is now available for artists to use

venturebeat.com

external-link
message-square
212
link
fedilink
  • cross-posted to:
  • technology@lemmy.world
566
external-link

Nightshade, the free tool that ‘poisons’ AI models, is now available for artists to use

venturebeat.com

Lee Duna@lemmy.nz to Technology@lemmy.worldEnglish · 1 year ago
message-square
212
link
fedilink
  • cross-posted to:
  • technology@lemmy.world
The tool's creators are seeking to make it so that AI model developers must pay artists to train on data from them that is uncorrupted.
  • Even_Adder@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    1
    arrow-down
    1
    ·
    1 year ago

    Something being derivative doesn’t mean it’s automatically illegal or improper.

    First, copyright law doesn’t prevent you from making factual observations about a work or copying the facts embodied in a work (this is called the “idea/expression distinction”). Rather, copyright forbids you from copying the work’s creative expression in a way that could substitute for the original, and from making “derivative works” when those works copy too much creative expression from the original.

    Second, even if a person makes a copy or a derivative work, the use is not infringing if it is a “fair use.” Whether a use is fair depends on a number of factors, including the purpose of the use, the nature of the original work, how much is used, and potential harm to the market for the original work.

    Even if a court concludes that a model is a derivative work under copyright law, creating the model is likely a lawful fair use. Fair use protects reverse engineering, indexing for search engines, and other forms of analysis that create new knowledge about works or bodies of works. Here, the fact that the model is used to create new works weighs in favor of fair use as does the fact that the model consists of original analysis of the training images in comparison with one another.

    You are expressly allowed to mimic others’ works as long as you don’t substantially reproduce their work. That’s a big part of why art can exist in the first place. You should check out that article I linked.

    • EldritchFeminity@lemmy.blahaj.zone
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      I actually did read it, that’s why I specifically called out MidJourney here, as they’re one I have specific problems with. MidJourney is currently caught up in a lawsuit partly because the devs were caught talking about how they launder artists’ works through a dataset to then create prompts specifically for reproducing art that appears to be made by a specific artist of your choosing. You enter an artist’s name as part of the generating parameters and you get a piece trained on their art. Essentially using LLM to run an art-tracing scheme while skirting copyright violations.

      I wanna make it clear that I’m not on the “AI evilllll!!!1!!” train. My stance is specifically about ethical sourcing for AI datasets. In short, I believe that AI specifically should have an opt-in requirement rather than an opt-out requirement or no choice at all. Essentially creative commons licensing for works used in data sets, to ensure that artists are duly compensated for their works being used. This would allow artists to license out their portfolios for use with a fee or make them openly available for use, however they see fit, while still ensuring that they still have the ability to protect their job as an artist from stuff like what MidJourney is doing.

      • Even_Adder@lemmy.dbzer0.com
        link
        fedilink
        English
        arrow-up
        1
        arrow-down
        2
        ·
        1 year ago

        I actually did read it, that’s why I specifically called out MidJourney here, as they’re one I have specific problems with. MidJourney is currently caught up in a lawsuit partly because the devs were caught talking about how they launder artists’ works through a dataset to then create prompts specifically for reproducing art that appears to be made by a specific artist of your choosing. You enter an artist’s name as part of the generating parameters and you get a piece trained on their art. Essentially using LLM to run an art-tracing scheme while skirting copyright violations.

        I’m pretty sure that’s all part of the discovery from the same case where Midjourney is named as a defendant along with Stability AI, it isn’t its own distinct case. It’s also not illegal or improper to do what they are doing. They aren’t skirting copyright law, it is a feature explicitly allowed by it so that you can communicate without the fear of reprisals. Styles are not something protected by copyright, nor should they be.

        I wanna make it clear that I’m not on the “AI evilllll!!!1!!” train. My stance is specifically about ethical sourcing for AI datasets. In short, I believe that AI specifically should have an opt-in requirement rather than an opt-out requirement or no choice at all. Essentially creative commons licensing for works used in data sets, to ensure that artists are duly compensated for their works being used. This would allow artists to license out their portfolios for use with a fee or make them openly available for use, however they see fit, while still ensuring that they still have the ability to protect their job as an artist from stuff like what MidJourney is doing.

        You can’t extract compensation from someone doing their own independent analysis for the aim of making non-infringing novel works, and you don’t need licenses or permission to exercise your rights. Singling out AI in this regard doesn’t make sense because it isn’t a special system in that regard. That would be like saying dolphin developers have to pay Nintendo every time someone downloads their emulator.

Technology@lemmy.world

technology@lemmy.world

Subscribe from Remote Instance

Create a post
You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !technology@lemmy.world

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


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related news or articles.
  3. Be excellent to each other!
  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, this includes using AI responses and summaries. To ask if your bot can be added please contact a mod.
  9. Check for duplicates before posting, duplicates may be removed
  10. Accounts 7 days and younger will have their posts automatically removed.

Approved Bots


  • @L4s@lemmy.world
  • @autotldr@lemmings.world
  • @PipedLinkBot@feddit.rocks
  • @wikibot@lemmy.world
Visibility: Public
globe

This community can be federated to other instances and be posted/commented in by their users.

  • 3.64K users / day
  • 9.51K users / week
  • 17.1K users / month
  • 37.8K users / 6 months
  • 1 local subscriber
  • 71K subscribers
  • 13.3K Posts
  • 476K Comments
  • Modlog
  • mods:
  • L3s@lemmy.world
  • enu@lemmy.world
  • Technopagan@lemmy.world
  • L4sBot@lemmy.world
  • L3s@hackingne.ws
  • L4s@hackingne.ws
  • BE: 0.19.11
  • Modlog
  • Instances
  • Docs
  • Code
  • join-lemmy.org