3 ms·
> simply training a model on illegally distributed text should not be copyright infringement You can train a model on copyrighted text, you just can't distribu
by dkjaudyeqooe 2y ago
> simply training a model on illegally distributed text should not be copyright infringement
You can train a model on copyrighted text, you just can't distribute the output in any way without violating copyright. (edit: depending on the other fair use factors).
One of the big problems is that training is a mechanical process, so there is a direct line between the copyrighted works and the model's output, regardless of the form of the output. Just on those terms it is very likely to be a copyright violation. Even if they don't reproduce substantive portions, what they do reproduce is a derived work.
- saulpw 2y agoIf that mechanical process is not reversible, then it's not a copyright violation. For instance, I can compute the SHA256 hashes for every book in existence and distribute the resulting table of (ISBN, SHA256) and that is not a copyright violation.
- dkjaudyeqooe 2y agoThat's actually within the other fair use factors. So your hash table is fair use because its transformative and doesn't substitute for the original work. I edited my post to make it a bit clearer.
- anticensor 2y agoIt's actually even less than fair use, it's non-copyright use: one-way hashes are intentionally designed to eliminate the creative element and output random looking data.
- aoanevdus 2y agoWhat’s a “mechanical process”? If I read The Lord of the Rings and it teaches me to write Star Wars, is that a mechanical process? My brain is governed by the laws of physics, right? What if I’m a simulated brain running on a chip? What if I’m just a super-smart human and instead of reading and writing in the conventional way, I work out the LLM math in my head to generate the output?
- dkjaudyeqooe 2y agoAnything a machine does. You can simulate whatever you like, but under the law it's not human so it's mechanical.
- gruez 2y ago>One of the big problems is that training is a mechanical process, so there is a direct line between the copyrighted works and the model's output, regardless of the form of the output. Just on those terms it is very likely to be a copyright violation. Even if they don't reproduce substantive portions, what they do reproduce is a derived work. Google making thumbnails or scanning books are both arguably "mechanical". Both have been ruled as fair use.