3 ms·

I think you could make an argument that ChatGPT is essentially a really powerful compression tool, and that its just reposting material that third parties uploa

by Taek 4y ago

I think you could make an argument that ChatGPT is essentially a really powerful compression tool, and that its just reposting material that third parties uploaded. You would try to argue to the judge that you are presenting the user with a compressed form of the training data, and that the training data is all third party. I don't know that this argument will be easily able to convince a judge, but there's a world where it happens.

basch 4y ago
It would make infinitely more sense to get around it contractually. To use the service, the prompter agrees to take liability for the output. More like upscaling. Should a service be liable for upscaling an illegal picture, that they "produced?" If LLM output is reverse compression, it is upscaling of the original prompt. (That said, I dont think your example of a model compressing the internet really invokes 230, because the model creator is the one grabbing the content, not running a service that other people post content to.)
dragonwriter 4y ago
> I think you could make an argument that ChatGPT is essentially a really powerful compression tool, and that its just reposting material that third parties uploaded. The training data sets are actively selected by OpenAI, they aren’t a result of third-parties submitting data to them. In some cases, the material was supplied to some other online services in way which would enable that provider to be protected by Section 230, but that doesn’t help OpenAI. > You would try to argue to the judge that you are presenting the user with a compressed form of the training data, and that the training data is all third party. The key part of Section 230 isn’t that information is third-party in origin, but that it is provided to the service in question at the initiative of some other content provider.
ahtihn 4y ago
Wouldn't making this argument mean that ChatGPT is violating copyright?
- dragonwriter 4y ago
  LLMs (well, neural network models more generally, but LLMs are the main interest right now) trained on copyright protected source data are either (1) violating copyright, or (2) benefit from an exception to copyright of some kind (such as Fair Use in the US, as Microsoft has explicitly claimed in connection with Github Copilot.) Even if they don’t contain, in some manner, literal copies of works (and in some cases they do, since works can be accidentally or intentionally recapitulated), they are derivative of copyright protected works in a direct and mechanical way.
  - Taek 4y ago
    Fair use likely applies to LLMs. Its loosely equivalent to an artist drawing inspiration while watching a movie.
    - dragonwriter 4y ago
      > Fair use likely applies to LLMs Fair use is so radically different from anything else for which there is legal precedent that I wouldn’t put money on either side of a court decision of fair use. > Its loosely equivalent to an artist drawing inspiration while watching a movie. It’s loosely analogous to lots of things, on both sides of fair use, but loose analogies are potentially very bad guides for fair use analysis, which often turns on fine distinctions. Clearly, given that LLMs are reproducible in ways human brains during the history of Fair Use have not been, the cases are radically disanalogous when it comes to market impact, which is an explicit factor in Fair Use analysis.
cowl 4y ago
That is certainly not what is happening and I think that argument is even more dangerous because it means that It is storing copies of potentially copyrighted material. In fact it is precisly this conjecture that is the bases of the latest lawsuit. https://www.theverge.com/2023/1/16/23557098/generative-ai-art-copyright-legal-lawsuit-stable-diffusion-midjourney-deviantart https://www.theverge.com/2023/1/16/23557098/generative-ai-ar... They need to show that they are not keeping any form of copy or original content compressed or not.

by Taek 4y ago

basch 4y ago

It would make infinitely more sense to get around it contractually. To use the service, the prompter agrees to take liability for the output. More like upscaling. Should a service be liable for upscaling an illegal picture, that they "produced?" If LLM output is reverse compression, it is upscaling of the original prompt. (That said, I dont think your example of a model compressing the internet really invokes 230, because the model creator is the one grabbing the content, not running a service that other people post content to.)

dragonwriter 4y ago

> I think you could make an argument that ChatGPT is essentially a really powerful compression tool, and that its just reposting material that third parties uploaded. The training data sets are actively selected by OpenAI, they aren’t a result of third-parties submitting data to them. In some cases, the material was supplied to some other online services in way which would enable that provider to be protected by Section 230, but that doesn’t help OpenAI. > You would try to argue to the judge that you are presenting the user with a compressed form of the training data, and that the training data is all third party. The key part of Section 230 isn’t that information is third-party in origin, but that it is provided to the service in question at the initiative of some other content provider.

ahtihn 4y ago

Wouldn't making this argument mean that ChatGPT is violating copyright?

dragonwriter 4y ago
LLMs (well, neural network models more generally, but LLMs are the main interest right now) trained on copyright protected source data are either (1) violating copyright, or (2) benefit from an exception to copyright of some kind (such as Fair Use in the US, as Microsoft has explicitly claimed in connection with Github Copilot.) Even if they don’t contain, in some manner, literal copies of works (and in some cases they do, since works can be accidentally or intentionally recapitulated), they are derivative of copyright protected works in a direct and mechanical way.
- Taek 4y ago
  Fair use likely applies to LLMs. Its loosely equivalent to an artist drawing inspiration while watching a movie.
  - dragonwriter 4y ago
    > Fair use likely applies to LLMs Fair use is so radically different from anything else for which there is legal precedent that I wouldn’t put money on either side of a court decision of fair use. > Its loosely equivalent to an artist drawing inspiration while watching a movie. It’s loosely analogous to lots of things, on both sides of fair use, but loose analogies are potentially very bad guides for fair use analysis, which often turns on fine distinctions. Clearly, given that LLMs are reproducible in ways human brains during the history of Fair Use have not been, the cases are radically disanalogous when it comes to market impact, which is an explicit factor in Fair Use analysis.

cowl 4y ago

That is certainly not what is happening and I think that argument is even more dangerous because it means that It is storing copies of potentially copyrighted material. In fact it is precisly this conjecture that is the bases of the latest lawsuit. https://www.theverge.com/2023/1/16/23557098/generative-ai-art-copyright-legal-lawsuit-stable-diffusion-midjourney-deviantart https://www.theverge.com/2023/1/16/23557098/generative-ai-ar... They need to show that they are not keeping any form of copy or original content compressed or not.