Hacker News Reader

Comment by madeofpalk | Hacker News Reader

I'm not as interested in making a technical/legal argument, as I'm just sharing my feelings on the topic (and eventually, what I think the law should be), but during training copies are made of copyrighted material, even if the model doesn't contain exact copies of work. Crawling, downloading, storing (temporarily) for training all involve making copies, and thus are subject to copyright law. Maybe those copies are fair use, maybe it's not (I think it shouldn't be).

My main point is that OpenAI is generating an incredible amount of value all hinging on other people's work at a massive scale, without paying for their materials. Take all the non-public domain work off Netflix and Netflix doesn't have the same value they have today, so Netflix must pay for content it uses. Same goes for OpenAI imho.

View on HN

> Take all the non-public domain work off Netflix and Netflix doesn't have the same value they have today, so Netflix must pay for content it uses. Same goes for OpenAI imho

I'd feel a lot better about that argument if we had sane copyright laws and anything older than 7-10 years was automatically in the public domain. Suddenly, netflix is looking a lot more valuable with just public domain works and there'd be a ton of public domain art to train AI models with. I suspect that the technology would still leave a lot of artists concerned in that situation though, because even once the issue of copyright is largely solved the fact remains that AI enables people who aren't artists to create art.

Assume I agree that copyright holders should be compensated for their works (because I do in some sense).

How would this compensation work? Let's say a portion of profits from LLMs that were trained on copyrighted work should be sent to the copyright holders.

How would we allocate which portion of the profits go to which creators? The only "fair" way here would be if we could trace how much a specific work influenced a specific output but this is currently impossible and will likely remain impossible for quite some time.

This is what licensing negotiations are for. One doesn't get to throw up their hands and say "I don't know how to fairly pay you so I won't pay you at all".

Your argument is ridiculous, because it could identically be applied to "every human artist should have to pay a license to every artist whose work they were inspired by". That would obviously be a horrible future, but megacorps like Disney would love it.

I definitely didn't intend such an implication, and I don't think it follows from what I said.

Calling it copyright today is a misnomer, it's not actually the act of copying the work that's a problem; it should actually be called "Performance Rights" or "Redistribution Rights." The part where this gets complicated is that OpenAI has (presumably, if they haven't that's a different matter) acquired the works through legal means. And having acquired them they're free to do most anything with them so long as they don't redistribute or perform the works.

The big question is where "training an AI on this corpus of works and then either distributing the weights or performing the work via API" fall? Should the weights be considered derivative works? I personally don't think so and although the weights can be used to produce obviously infringing works I don't think this meets the bar of being a redistribution of the work via a funny lossy compression algo like some are claiming. But who knows? Copyright is more political than logical so I think the bend is really gonna be a balance of the tangible IRL harms artists can demonstrate vs. the desires of unrelated industries who wish to leverage this technology and are better for having all this data available.

Every time you view an image online you’re making a copy so that argument is spurious.

Reply on news.ycombinator.com