The world’s biggest AI companies are buying antiquarian books en masse so they can be scanned to train their large language models (LLMs) before being destroyed.
I hate when comments are 50% right and 50% infuriatingly incorrect.
…
It’s one copy of every book.
Per entity that wants to train their AI. So probably never just one copy of any given book.
…
This is the direct result of people complaining about piracy
People complain about piracy? What alternate timeline am I in right now? Or am I still in the one where it’s primarily Corporations that complain about piracy.
If you mean people complaining about AI’s stealing art/etc. AI’s don’t pirate, they perjure other people’s work as their own. Massive difference. I didn’t download the car and sell it like I built it.
…
They don’t need to do it for anything older than Steamboat Willie. Their motivation is quantity, so they’re not bidding millions toward illuminated manuscripts, they’re taking obscure nonsense by the shovel-full.
Lack of need isn’t a restriction. Indiscriminately seeking quantity implies they’re not checking if it was made before steamboat willy.
…
If it’s rare it’s because nobody gave a shit.
Most incorrect thing you said. Not how physical objects work.
…
And it’s probably preserved in torrents anyway.
Bro I literally collect books made after steam boat willy that never made it to the internet.
And now twice a week we have headlines acting like they’ve uncovered a library firebombing campaign.
Maybe you don’t read books so you don’t care as much as people who do. But in the books I’ve read mass book destruction is called book burning and it’s never contributed to positive outcomes for anyone working class.
How else do you think mass-produced things become rare? Either they printed a zillion copies and nearly all of them were thrown away, or the print run was tiny for lack of demand. Action Comics #1 is worth a fortune and Puget Sound: Proceedings Of A Seminar Held January 21, 1987 is not.
Per entity that wants to train their AI.
So like twelve copies? There’s no explosion of mom & pop AI startups that would justify the comparisons to Nazi book burning that are in every thread about this. If any work’s last physical copy is threatened, it was probably headed for the trash. Is it better-off being lost entirely, with nobody having a scan?
And yes, people complained when these companies did the obvious thing and grabbed a torrent. Just as loudly as they’re complaining about them doing this on the up-and-up.
I hate when comments are 50% right and 50% infuriatingly incorrect.
…
Per entity that wants to train their AI. So probably never just one copy of any given book.
…
People complain about piracy? What alternate timeline am I in right now? Or am I still in the one where it’s primarily Corporations that complain about piracy.
If you mean people complaining about AI’s stealing art/etc. AI’s don’t pirate, they perjure other people’s work as their own. Massive difference. I didn’t download the car and sell it like I built it.
…
Lack of need isn’t a restriction. Indiscriminately seeking quantity implies they’re not checking if it was made before steamboat willy.
…
Most incorrect thing you said. Not how physical objects work.
…
Bro I literally collect books made after steam boat willy that never made it to the internet.
Maybe you don’t read books so you don’t care as much as people who do. But in the books I’ve read mass book destruction is called book burning and it’s never contributed to positive outcomes for anyone working class.
How else do you think mass-produced things become rare? Either they printed a zillion copies and nearly all of them were thrown away, or the print run was tiny for lack of demand. Action Comics #1 is worth a fortune and Puget Sound: Proceedings Of A Seminar Held January 21, 1987 is not.
So like twelve copies? There’s no explosion of mom & pop AI startups that would justify the comparisons to Nazi book burning that are in every thread about this. If any work’s last physical copy is threatened, it was probably headed for the trash. Is it better-off being lost entirely, with nobody having a scan?
And yes, people complained when these companies did the obvious thing and grabbed a torrent. Just as loudly as they’re complaining about them doing this on the up-and-up.
The fact that you see no problem with that is why I no longer respect you intellectually enough to consider you improving my lemmy experience.
have a nice life bro
Things get thrown away and it’s my fault somehow.