768
OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling's Harry Potter series (www.businessinsider.com)
submitted 3 years ago by [M, B] to c/technology@lemmy.world
306 comments fedilink hide all child comments
 

OpenAI now tries to hide that ChatGPT was trained on copyrighted books, including J.K. Rowling's Harry Potter series::A new research paper laid out ways in which AI developers should try and avoid showing LLMs have been trained on copyrighted material.

you are viewing a single comment's thread
view the rest of the comments
[–] 5 points 3 years ago (1 child)

I mean, this is the exact way the U.S. Copyright Office's guidance says they think you should use it.

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

    Sure, but even under this guidance copyright owners of the training data are still shafted, based on how the data is scraped pretty much freely. Can an opportunist generate an unofficial sequel to Harry Potter, do the minimum to ensure they get copyright and reap the reward from it?

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

    That's how copyright has always worked. Fair use is complex, but as long as you're not straight up copying someone's work you're fine. 50 Shades of Grey started out as Twilight fanfiction. So yeah, you could.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [–] 3 points 3 years ago* (1 child)

    Yes fair use has its stipulations but AI is breaking new grounds here. We are no longer dealing with the reaction videos but in an era where literally dozen of pages of content can be generated in a matter of minutes, with relatively little human involvement. Perhaps it's time to revisit if the law still holds in light of these new technology and tools.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 3 years ago (1 child)

    Fair use has never been seriously challenged. I’m betting it might happen soon though. We have to remember Fair Use isn’t a law, it’s a set of guidelines under the law that has never been clearly defined.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 3 years ago* (last edited 3 years ago)

    First of all, fair use is not a set of guidelines, it's a legal doctrine that allows us limited use of copyrighted material without permission from the owner. It is a part of the U.S. Copyright Act, which is a law enacted by Congress.

    Second, fair use has been seriously challenged plenty of times, just to name a few:

    • Campbell v. Acuff-Rose Music, Inc.

    • Authors Guild v. Google, Inc.

    • Lenz v. Universal Music Corp.

    I recommend reading this article by Kit Walsh, who’s a senior staff attorney at the EFF, a digital rights group who recently won a historic case: border guards now need a warrant to search your phone.

    Fair use protects creativity, innovation, and our freedom of expression, but You almost sound like you want it weakened.

  • permalink
  • fedilink
  • source
  • parent