alt Hacker News>From a historical precedent standpoint, this is akin to the burning of the library of Alexandria, where centuries of knowledge was destroyed and leaving a limited version of the history, the surviving one, and depriving successive generations of significant amount of latent knowledge.
How tho?
Ever sunday at the flea market, I see thousands of books that are rotting, hoping for someone to buy them or at least take them home, so the seller doesn't have to pack them for the trip back. Just the other day, there was a whole bin of books in front of a shop, offering them for 50 cents a piece. They will be destroyed anyways.
Unless they are buying and destroying really old, rare books or important small-print books, it is not much damage. It is not like they will buy "all copies of all of the books", just one. And its just that the data in physical print most likely hasn't been used for training, so this can help you find more unmined quality data. Nobody is stealing your books, preventing you from buying more or destroying all copies of a single book.
And some of these books would rot out of circulation or be destroyed anyways. Some people throw away 80-100 year old books on the regular, as they might just be unimportant to them or the world in general. And once the last copy is thrown or rots, that book will die forever. This way, it will live forever instead, scanned and trained on, conjoined with the rest of our knowledge in a magic machine.
Replies
>it will live forever instead, scanned and trained on, conjoined with the rest of our knowledge
This sentence reminded me of “The Things” by Peter Watts. The Thing in that short story believes it’s actually the good guy and decides to commit “violent integration” for the sake of humanity.
I’m not sure I buy the links premise that Anthropic is doing this with malice, but if it were, would be an eerie parallel.
Good to see someone making this point. I'm confused by the panic, because they are making it out like AI companies are destroying every copy of the book. They only need one, and they destroy it after scanning it only because they don't want to store them all. And storing or archiving all these books is not a trivial task.
> This way, it will live forever instead, scanned and trained on, conjoined with the rest of our knowledge in a magic machine.
But that's just it, it won't live on forever because, from the perspective of preservation, training is a lossy, noninvertible transformation. The LLM cannot legally produce the book verbatim, it will only spit out a regurgitation of the information, chopped and mingled into a broad information space.
Furthermore, these "magic machines" are not the property of the public. They are owned by a handful of corporations who want to charge you continuously for every token output by the machine. So, not only is the original text locked away forever behind company walls, you now need to pay for access to an approximation of the original contents which you can no longer even verify as being correct because the source is no longer accessible.
If you are cool with this, from a cost perspective you are cool with a deal whereby I trade you access to a definite resource for a one time fee of $N for, instead, a perpetual cost of $M to you every month/day/hour for access to an amalgam in which you cannot even determine what proportion of the resource you are actually getting. You're basically saying you're cool with me selling you some unknown portion of wine for a monthly subscription price instead of selling you a definitive amount of wine for a one time fee. lol.
> This way, it will live forever instead, scanned
But unreadable by humans, right?
Those books at the flea market aren't "rotting", they are available for sale.
You seem perfectly fine living in a world where your flea market is devoid of books.
Oho you think they're buying common books? Books they can find online already?
You’re equating something being ubiquitous and affordable with it not being valuable.
Yes, there are plenty of books, many were printed, many have lasted a very long time (plenty over 100 years!).
That says more about the success and utility of the technology than it does about whether individual books should be shredded.
Yep, this feels pretty much it. Looking at the "rare books", it was books that nobody would care about or would just rot away anyway. (The example book of Old books of agriculture is probably not that important today)
Liberians have to accept that much of their job is sending books to be burned. A lot of them try their best to get people to be interested in older books, but they have to make way for "newer" books instead.
The article is written in the same way as how dogs are getting murdered in the dog shelter, even if "everyone" agrees it is wrong, yet nobody adopts them.