Comment by xienze
7 days ago
I think the term "rare" might be a bit loaded. DOES it mean "only a handful in existence" or more like "1000 copies?" And at any rate, at least by scanning the book they're theoretically making its contents available to the public. What are the other people who own these books doing besides having them sit on a shelf?
"Only a handful" is a precious or extremely rare book. 1000 copies of a book total would certainly qualify it as rare. That's not a lot of copies.
> What are the other people who own these books doing besides having them sit on a shelf?
Not destroying them. By continuing to exist, it keeps the possibility that the books will eventually go to someone else. Maybe even a library that specializes in rare books.
> By continuing to exist, it keeps the possibility that the books will eventually go to someone else.
Good news, there's hundreds or thousands of copies that continue to exist. Anthropic, Google et al only need a single copy.
And shockingly, for all this talk of rare books, we don't see the other owners of said books stepping up and offering to have these books scanned non-destructively. I won't hold my breath for them to do so in the short or long term.
FURTHERMORE, the books aren't being scanned for no reason, they become part of new model training data. Meaning that whatever insights these books contain might be available to users in the future. Again, more valuable than all the copies rotting away on someone's bookshelf that might be scanned and made available some day.
It's hard to describe what happens to the book's contents, or determine whether it would be more valuable intact. Insights are the result of a person interacting with a book, they aren't like juice that can be squeezed out mechanically.
I care a lot about the high-illustration picture books of the 80s. Since they have ISBNs, they're presumably on the list. Does the LLM ingest the artwork? Even when it comes to text-only novels, it isn't making the experience of reading the book available.