Replying to a post on babka.social
I'm pretty sure no one is trying to destroy all physical books.
Replying to @serge@babka.social
@serge nah its an epidemic rn :(( you should google it. it's really crazy. they're going for rare books first
Replying to @batzy@babka.social
FuturismAnthropic Knew the Public Would Be Disgusted by How It Was Destroying Physical Books, Secret Documents RevealNewly unsealed documents suggest that Anthropic was well aware that destroying books to train its AI would look bad.Replying to @batzy@babka.social
Yeah this is what I thought. It's not attempting to destroy all books. It's that scanning books is hard when they still have their spines. Google and even Archive.org have done similar. Not doing so is *extremely* challenging.
And then once the books are digitized, storing them is very expensive. Destroying physical media is easier.
I'm not "defending" it but I know this space. It's less sinister than it sounds, honestly.
Replying to @serge@babka.social
@serge @batzy seconding this, and I'll add an anecdote. When I was in college at the end of the 90's and the early 2000's, the way we got our textbooks in an accessible format, (not braille, just text we could read on a computer using a screen reader), was to buy two copies of a book, take apart its binding, feed the pages into a scanner, and then, once done, destroy/get rid of the destroyed books. Scanning in any other way meant a ton of time spent manually correcting the text, which added significant delays onto an already time-consuming process. That part of digital scanning of books hasn't really changed.
Replying to @acarson@babka.social
@acarson @serge @batzy the problem is that I had to scan French textbooks written in a mixture of Russian and French (because at that time I was studying French from Russian), also with IPA and sometimes with remnants of some idiot students' notes done by pencil or even pen. Really, as we say in russian, as I remember, I shiver.