6 ms·

I have absolutely no problem with generative models, not even for producing 'illegal' images and texts, but the copyright infringement by big players is disgust

by tpxl 4y ago

I have absolutely no problem with generative models, not even for producing 'illegal' images and texts, but the copyright infringement by big players is disgusting. Microsoft (copilot), openai, Adobe, ... etc will wholesale breach copyright and not give a damn, but god forbid if you copy a line of their code they will use all their might to crush you. If they want to use copyrighted stuff for their models, they should license it.

andybak 4y ago
This comment is a bit hard to unpick. It's not clear that training violates copyright. However plagarism (created by any means AI or not) is already covered by existing IP law. (The difficult part is telling whether AI generated stuff infringes on the copyright of stuff in the training data. And the fact the onus is on the end-user of AI tools sucks...) However - if you're arguing that copyright law should be extended to prevent use in training - that's a much tougher sell and one where I'm not 100% sure of my position.
- tpxl 4y ago
  > It's not clear that training violates copyright What is perfectly clear are two things: 1. Github Copilot will sometimes produce non-trivial inputs it was trained on as outputs 2. Those outputs will not conform to the license That is, in my mind, a clear violation. This could easily be solved by making training (or transformation or whatever) explicit in the license. Edit: I single out Github copilot because such examples were found. I'm not sure you can find such examples for things such as Stable Diffusion or DALL-E, but the operating principle is the same.
  - wongarsu 4y ago
    > I'm not sure you can find such examples for things such as Stable Diffusion or DALL-E DALL-E 1 will sometimes replicate watermarks (without being promted for it [1]). I'm not sure anyone has ever shown a 1-1 replication like what's shown with Copilot, but it's easy to get image models to create clearly derivative works (just ask for the Mona Lisa, or even a well-known modern work). 1: https://ibb.co/gz5RDkB https://ibb.co/gz5RDkB
- ginko 4y ago
  >It's not clear that training violates copyright. Training a model essentially creates a derivative work from the input data. GPL licensed code requires keeping the same license for derivative work. BSD-style licenses at the very least require retaining the copyright statement. It feels pretty clear to me that the trained model and code generated from it would have to adhere to the original license. In a way ML is a bit like lossy compression. Would you think JPEG encoding and decoding an image strips the result of its original copyright?
  - andybak 4y ago
    I agree that this is one likely interpretation but it's not the only one and it's yet to be tested in the courts.
  - wongarsu 4y ago
    There is a point where a derivative work is so transformative that the original copyright no longer matters. In the US you would argue via fair use, where the precedent is "The more transformative the new work, the less will be the significance of other factors, like commercialism, that may weigh against a finding of fair use" [1], and other jurisdictions have equivalent ways to excempt works that are different or transformative enough. So the big question is how transformativ a given model is. And at least for language models I'm not sure we even understand Transformer models enough to give a qualified answer. 1: https://en.wikipedia.org/wiki/Transformative_use https://en.wikipedia.org/wiki/Transformative_use
    - staindk 4y ago
      I think this is true, but currently you can coax e.g. stable diffusion to generate art in the style of a specific person, and if you do things right, it will even have a smudge-y version of their monogram/signature in the bottom right. Pretty clear indication that something is not yet 'transformative enough', IMO. Same happens with Copilot and such - you can get it to spit out tens of lines of code verbatim pulled from someone else's project, with comments and all. These kinds of things just don't sit right with me.
archontes 4y ago
It isn't copyright infringement to train an AI on information that you obtained legally. Example: If I put my book on mybook.com, and you download it (legally), you can read it, learn from it, and produce works in a similar style, all without my consent, and copyright offers no tools to restrict that. The only tool copyright offers to protect against that is distribution.
- staticman2 4y ago
  A.I.'s are subject to copyright (unlike a human being like yourself ) so the A.I. which has been taught is an infringement of the copyright of the artist because the A.I. is a derivitive copyrightable work. Since copyright doesn't govern your brain's "wetware", the comparison to human behavior is irrelevant.
  - archontes 4y ago
    Two issues with that. Training an AI has thus far been regarded in courts as non-creative, and thus the AI model/weights are not copyrightable, nor does the model sufficiently represent the original work, so it's not a derivative work.
    - staticman2 4y ago
      I think you are confusing the issue of whether the A.I. output is creative with whether training the A.I. itself is creative enough for the A.I. software to be copyrightable. The law has a very broad definition of creativity when a human is involved, even a human taking a picture of the Mona Lisa is considered creative. Since humans train the A.I., and it takes creativity to design the training scheme, the A.I. would appear to me to clearly be a creative work and copyrightable. You brought up a seperate issue of whether the model infringes the original work. And whether it's okay because the model doesn't resemble the original. I don't think that's precisely the accurate legal standard. What's relavant is whether the copying into the model constitutes fair use. One of the considerations of fair use is whether the use is transformative. But that's only one consideration a court will look at in determining whether the A.I. company can succeed in a fair use defense. I don't believe this is settled law. But if there's any broad point I want to make, it's that the law does not ever consider a machine or piece of software "like a human" or say "If it's okay for a human to do it it's okay for a piece of software to do it."

by tpxl 4y ago

andybak 4y ago

This comment is a bit hard to unpick. It's not clear that training violates copyright. However plagarism (created by any means AI or not) is already covered by existing IP law. (The difficult part is telling whether AI generated stuff infringes on the copyright of stuff in the training data. And the fact the onus is on the end-user of AI tools sucks...) However - if you're arguing that copyright law should be extended to prevent use in training - that's a much tougher sell and one where I'm not 100% sure of my position.

tpxl 4y ago
> It's not clear that training violates copyright What is perfectly clear are two things: 1. Github Copilot will sometimes produce non-trivial inputs it was trained on as outputs 2. Those outputs will not conform to the license That is, in my mind, a clear violation. This could easily be solved by making training (or transformation or whatever) explicit in the license. Edit: I single out Github copilot because such examples were found. I'm not sure you can find such examples for things such as Stable Diffusion or DALL-E, but the operating principle is the same.
- wongarsu 4y ago
  > I'm not sure you can find such examples for things such as Stable Diffusion or DALL-E DALL-E 1 will sometimes replicate watermarks (without being promted for it [1]). I'm not sure anyone has ever shown a 1-1 replication like what's shown with Copilot, but it's easy to get image models to create clearly derivative works (just ask for the Mona Lisa, or even a well-known modern work). 1: https://ibb.co/gz5RDkB https://ibb.co/gz5RDkB
ginko 4y ago
>It's not clear that training violates copyright. Training a model essentially creates a derivative work from the input data. GPL licensed code requires keeping the same license for derivative work. BSD-style licenses at the very least require retaining the copyright statement. It feels pretty clear to me that the trained model and code generated from it would have to adhere to the original license. In a way ML is a bit like lossy compression. Would you think JPEG encoding and decoding an image strips the result of its original copyright?
- andybak 4y ago
  I agree that this is one likely interpretation but it's not the only one and it's yet to be tested in the courts.
- wongarsu 4y ago
  There is a point where a derivative work is so transformative that the original copyright no longer matters. In the US you would argue via fair use, where the precedent is "The more transformative the new work, the less will be the significance of other factors, like commercialism, that may weigh against a finding of fair use" [1], and other jurisdictions have equivalent ways to excempt works that are different or transformative enough. So the big question is how transformativ a given model is. And at least for language models I'm not sure we even understand Transformer models enough to give a qualified answer. 1: https://en.wikipedia.org/wiki/Transformative_use https://en.wikipedia.org/wiki/Transformative_use
  - staindk 4y ago
    I think this is true, but currently you can coax e.g. stable diffusion to generate art in the style of a specific person, and if you do things right, it will even have a smudge-y version of their monogram/signature in the bottom right. Pretty clear indication that something is not yet 'transformative enough', IMO. Same happens with Copilot and such - you can get it to spit out tens of lines of code verbatim pulled from someone else's project, with comments and all. These kinds of things just don't sit right with me.

archontes 4y ago

It isn't copyright infringement to train an AI on information that you obtained legally. Example: If I put my book on mybook.com, and you download it (legally), you can read it, learn from it, and produce works in a similar style, all without my consent, and copyright offers no tools to restrict that. The only tool copyright offers to protect against that is distribution.

staticman2 4y ago
A.I.'s are subject to copyright (unlike a human being like yourself ) so the A.I. which has been taught is an infringement of the copyright of the artist because the A.I. is a derivitive copyrightable work. Since copyright doesn't govern your brain's "wetware", the comparison to human behavior is irrelevant.
- archontes 4y ago
  Two issues with that. Training an AI has thus far been regarded in courts as non-creative, and thus the AI model/weights are not copyrightable, nor does the model sufficiently represent the original work, so it's not a derivative work.
  - staticman2 4y ago
    I think you are confusing the issue of whether the A.I. output is creative with whether training the A.I. itself is creative enough for the A.I. software to be copyrightable. The law has a very broad definition of creativity when a human is involved, even a human taking a picture of the Mona Lisa is considered creative. Since humans train the A.I., and it takes creativity to design the training scheme, the A.I. would appear to me to clearly be a creative work and copyrightable. You brought up a seperate issue of whether the model infringes the original work. And whether it's okay because the model doesn't resemble the original. I don't think that's precisely the accurate legal standard. What's relavant is whether the copying into the model constitutes fair use. One of the considerations of fair use is whether the use is transformative. But that's only one consideration a court will look at in determining whether the A.I. company can succeed in a fair use defense. I don't believe this is settled law. But if there's any broad point I want to make, it's that the law does not ever consider a machine or piece of software "like a human" or say "If it's okay for a human to do it it's okay for a piece of software to do it."