AI companies are buying up and destroying tons of old books for training data
https://www.404media.co/ai-companies-are-buying-tons-of-old-books-because-theyre-free-of-ai-slop/
https://www.404media.co/ai-companies-are-buying-tons-of-old-books-because-theyre-free-of-ai-slop/
25 Comments
DickFiasco@sh.itjust.works · 55 pts · 13d
The judge's logic is baffling:
AI companies are using the book to create a commercial product. Just because that product isn't an exact copy of the book doesn't mean they're not abusing the copyright.
riskable@programming.dev · 18 pts · 13d
No, it does mean that. If a work doesn't contain enough of the original that a layman could identify it as a copy (plagiarism) or derivative, it's not violating or abusing copyright.
Mountains of popular books (and all media, really) are built upon previous works. If copyright worked the way you suggest, none of those stories would be legal.
The act of copying a book you own into a database is legal. That's all Alsup is reaffirming in that quote you shared.
Remember: Copyright is all about distribution. Until a copyrighted work is distributed, copyright law doesn't even really come into play.
Those of us who remember the MPAA/RIAA wars of the 90s and 2000s understand that it is your right to make as many copies of a thing as you damned well please. Fuck the MPAA and fuck you if you're taking their side.
We must never allow corporations to control what we do with our own shit. You know that, "you will own nothing and you'll love it" quote? If you're against what Alsup is saying in this ruling, you support that stance (of owning nothing).
Miller@lemmy.world · 0 pts · 13d
It is not baffling, humans have been inspired and influenced in their own work by the work of others since the very start. If you read a book and learn a few words and concepts and then use them in your own book it is called creativity but if AI does it then its called slop. I am human and I am on the side of the humans here, my point is the two things look identical in law and we need a way to distinguish them in the same way acknowledged derivation is differentiated from plagiarism.
Ilixtze@lemmy.ml · 10 pts · 13d
Ai is not human, it is a corporate product fuck this idea that the stochastic parrot is “inspired" fucking llm's are not inspired by shit.
Miller@lemmy.world · -10 pts · 13d
Maintaining an unsubtle and primitive view will not help ring-fence and protect human creativity. Your words are just noise and add only confusion and distraction to an argument that needs to be clear and precise. AI model output can be inspired and creative, even while in the strictest sense this is still just human creativity repackaged. The most vital thing is that it has to be reliably labelled as AI generated and its primary sources acknowledged, then it can be judged legitimately on its own merit.
Ilixtze@lemmy.ml · 10 pts · 13d
"Human creativity repackaged" the jokes write themselves, these joyless clowns don't want to do the necessary work for creative ideation and yet they yearn so much for their llm enabled intellectual cuckoldry to be recognized. It always comes off as pleading. This is the main reason why it's so difficult for AI content to be properly labeled as AI; Because while the AI users want to promote themselves as the "next step of human creativity" Once it is labeled as AI assisted people stop caring.
So sorry for not enabling the induced Dunning-kruger maybe you should talk to an llm that enables and supports your views unconditionally and calls your way of thinking "precise thought". Than come back and share with us some more of your ai psychosis.
Miller@lemmy.world · -6 pts · 13d
You are foaming at the mouth, I do not see how you think that can help things. Listening to a little counter argument might advance matters but that is not really your aim is it cuz. You want lynchings because that's what you like.
Ilixtze@lemmy.ml · 2 pts · 13d
That's good man! you are writing some fiction about my reaction! Keep doing that and you might stop needing the LLM's.
Miller@lemmy.world · -2 pts · 13d
In 1141 when Matilda took power to be the first ruling Queen of England in the full sense of female King the clerical chroniclers of the time falsely accused her of arrogance to demean her and undermine her legitimacy.
zarkanian@sh.itjust.works · 5 pts · 13d
No, it can't, because inspiration and creativity arise out of brains, and LLMs don't have brains or anything similar to brains. They just output tokens that are statistically likely to appear together.
Seppo@sopuli.xyz · 0 pts · 13d
You have no fucking clue how an LLM works. It's not "AI", that's just marketing.
Strider@lemmy.world · 17 pts · 13d
Ah yes, for the good of humanity.
kamenlady@lemmy.world · 13 pts · 13d
I've been reading similar headlines for a while now. This means they're apparently successful at it and not stopping until books are practically gone.
Grimy@lemmy.world · -14 pts · 13d
Please, books aren't in danger of disappearing.
zarkanian@sh.itjust.works · 18 pts · 13d
There is a danger for rare books, though. If AI companies buy up and destroy these rare books, they become much harder to find and might disappear entirely.
Grimy@lemmy.world · 0 pts · 13d
Fair point. Shredding rare books does have a bad ring to it.
Ilixtze@lemmy.ml · 7 pts · 13d
The tech degenerate billionaires want to centralize the sources of information in their wasteful propaganda machines, so books are in danger because they might provide dangerous alternative points of view. A lot of tech overlord are turning out to be nazis so it makes sense that a roundabout way of burning books is part of their programme.
Grimy@lemmy.world · 1 pts · 13d
This only works if we lived in medieval ages without a printing press. Mind you, what you describe is probably their fantasy knowing how sick and twisted billionaires are, but burning one copy of every book is just a piss poor way of doing it. The whole idea is a bit silly tbh.
Ilixtze@lemmy.ml · 1 pts · 13d
They don't need to really destroy the books; Literacy in the US is already very low according to studies and reports; The plan of the tech nazis goes more along the way of trying to fill the world with so much LLM generated junk and making that junk so accessible that the information contained in books is easy to undermine. A process we are already seeing in the Good ol' USA with the slow dismantling of their scientific institutions that Elon musk took into his own hands. It goes without saying that in that same year Musk pushed to create his own revised version of the wikipedia. It's erasure by burying in a land yard of llm slop, followed by de-skilling of any potential opposition.
Dryad@lemmy.world · 8 pts · 13d
Gotta love how ai generated content is considered “contamination.” I agree, it’s just weird hearing it out loud.
Ilixtze@lemmy.ml · 3 pts · 13d
Llm content is useless spam. This is not weird at all.
Dryad@lemmy.world · 2 pts · 13d
I know it’s not factually weird, it’s just weird to hear an insider admit it.
Luisp@lemmy.dbzer0.com · 6 pts · 13d
Literally Fahrenheit 451
heartSagan5@lemmy.zip · 1 pts · 13d
quick_snail@feddit.nl · 1 pts · 12d
Not sure if this is worse or better than them just pirating them