401–410 of 599 posts
I think this will be a bigger issue than some people think. Maybe there's a market for 'clean' training data that doesn't include potential copyright claims. Just public domain works. We'll know it's an AI because it talks like a late 18th century/early 19th century writer?
Earlier quoted context omitted.
Copyright is much newer than books.
So is internet and rapid copy-sharing of books. I personally feel our copyright laws are too rigid, but that doesn't mean copyright shouldn't exist. After x years, any book should be free to read, after y years, it should be free to be incorporated into AI models, after z years it should be in the public domain.
I don't get your point. Because we can, we shouldn't...?
Earlier quoted context omitted.
> Possession of copyrighted material in itself is not illegal. The means of procurement matters. If they are in possession of copyrighted material because someone without the proper rights gave it to them illegally, then the possession itself is also illegal. It's illegal to own knowingly stolen property in all 50 US states and most countries, and while we could argue to the end of days about whether copying a file t…
> If they are in possession of copyrighted material because someone without the proper rights gave it to them illegally, then the possession itself is also illegal. No, they aren’t. > It's illegal to own knowingly stolen property in all 50 US states While copyright violation is often metaphorically (or hyperbolicly) referred to as stealing, copyright violation isn't theft and a copy created in violation of copyright…
Very convincing argument. Also, that's maybe the one part of this discussion that can't be debated. Possession of illegally obtained property, intellectual or otherwise, is illegal. Always has been, always will be. It's bizarre for you to be claiming otherwise.
> While copyright violation is often metaphorically (or hyperbolicly) referred to as stealing [...]
You are making a pedantic argument about the term "stealing," which is annoyingly pointless given the rest of that sentence (which you conveniently didn't quote) acknowledges the debate about the term. However, there's no debate to be had. The courts have clarified that violating intellectual property is still a denial of owed compensation (theft), but instead prefer the term "infringe" to make clear the distinction between violating physical rights (criminal) and violating intellectual rights (civil).
It's still a violation of copyright to be in possession of works obtained via illegal reproduction. You have zero fair use protections for illegally reproduced content. You are still breaking the law. You are still stealing via denial of compensation. The courts have already clarified all of this. Your pedantry doesn't change any of that.
Earlier quoted context omitted.
I agree it's not an entirely new issue. But it's a little different from search results. Say I use the generative paint brush in photoshop. It reproduces a portion of the copyrighted work. I then use the image on an advertising campaign, other merchandise, or post the final product as my own work. Would I be responsible? Would Adobe? Given that retraining these models is not simple, or cheap, would this be just 'cost…
In that particular case if you have an enterprise licence, Adobe have accepted responsibility: >Adobe is so confident its Firefly generative AI won’t breach copyright that it’ll cover your legal bills The offer is available only to users of its enterprise Firefly product, which launches today https://www.fastcompany.com/90906560/adobe-feels-so-confiden...
Smells to me like a bad case of corporate marketing bullshit.
Earlier quoted context omitted.
I agree it's not an entirely new issue. But it's a little different from search results. Say I use the generative paint brush in photoshop. It reproduces a portion of the copyrighted work. I then use the image on an advertising campaign, other merchandise, or post the final product as my own work. Would I be responsible? Would Adobe? Given that retraining these models is not simple, or cheap, would this be just 'cost…
This already sort of happened in 2008 when Chuck Close forced someone to stop distributing a Photoshop plug-in that imitated his style. https://hyperallergic.com/54104/my-chuck-close-problem/
I'm not a lawyer but I'm pretty sure this was a trade mark issue. He just had to avoid marketing his product using someone else's name.
> The complaint lays out in steps why the plaintiffs believe the datasets have illicit origins — in a Meta paper detailing LLaMA, the company points to sources for its training datasets, one of which is called ThePile, which was assembled by a company called EleutherAI. ThePile, the complaint points out, was described in an EleutherAI paper as being put together from “a copy of the contents of the Bibliotik private t…
> If you downloaded a book from that website, you would be sued and found guilty of infringement. How often does this actually happen? You might get handed an infringement notice, and your ISP might terminate your service if you're really egregious about it, but I haven't ever heard of someone actually being sued for downloading something.
The best thing that could happen to humanity is if OpenAI is sued comically into oblivion. LLMs are the anti-humanity, and the sooner we rid the planet of them the better off we'll be.
Downvote me all you want to show you have nothing to say. Return to my comment in 10 years to find I'm right.
Earlier quoted context omitted.
> it is just as likely it saw any number of things about it Is this based on inside information, or just the law of averages? Doesn't the fact that they openly admitted to having been trained on pirated books affect your priors?
They didn't, more conjecture
See post https://news.ycombinator.com/item?id=36659041 for a summary.
Earlier quoted context omitted.
Perhaps not, I thought one of the claims is interesting though, that they illegally acquired some of the dataset. What would be the damages from that, the retail price of the hardcopy?
Wouldn't they first need to prove that OpenAI didn't ingest summaries of the book, and not the book itself?
Earlier quoted context omitted.
> IMO doesn't constitute fair use Yes, the question of whether the way LLMs use the content they use qualifies as fair use is a separate question. My point was simply that that question can't even be reached if the maker of the LLMs doesn't have a legal right to fair use in the first place (because they don't legally own their copy).
> My point was simply that that question can't even be reached if the maker of the LLMs doesn't have a legal right to fair use in the first place (because they don't legally own their copy). I agree, and I expect that eventually we will start seeing injunctions against creators requiring them to remove content that they don't have legal access to from their training data sets. And this will probably end up at the Sup…
Self edit: I meant against LLM creators.