635
AI labs buy, scan, shred millions of rare books (www.news.com.au)
submitted 2 months ago* by to c/technology@lemmy.world
179 comments fedilink hide all child comments
 

Artificial intelligence labs are in a new arms race to buy up millions of rare books, slicing them open, scanning the pages and pulping the remains — sparking concerns that the last remaining copies of out-of-print texts are being destroyed on an industrial scale.

ISBNdb notes that “print books from the pre-LLM era are structurally guaranteed to be free of this contamination”.

“Millions of the most valuable books have never been digitised. They exist only in physical form, scattered across library shelves, used bookstores, and out-of-print catalogues. We get them to you at scale.”

you are viewing a single comment's thread
view the rest of the comments
[–] 64 points 2 months ago (17 children)

Does anyone remember circa 2014 that radical jihadists in the middle east were destroying ancient and medieval artifacts, including stone heads? And I remember here in the liberal west we thought that was awful.

This is not the only thing that makes me think of that. Also Trump's recent calls to reduce the size of US national parks and open them up to oil and uranium mining exploitation.

  • permalink
  • fedilink
  • source
  • hideshow 17 child comments
    •
  • [–] 8 points 2 months ago (6 children)

    Yeah we remember. Because that’s a literal fucking warcrime. You’re not allowed to destroy historical sites.

    While I agree that destroying books is bad - even common ones - it’s also not technically illegal. If I want to burn a first edition of whatever book worth thousands of dollars… I’m legally allowed to do that.

    I can also see the benefit of having a digital version of a rare book that can be enjoyed by thousands as opposed to sitting on a shelf. But yeah, definitely hold companies accountable for destroying them. There has to be a better way to do that work and still have a useable book afterwards.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 6 child comments
    •
  • [–] 14 points 2 months ago

    I don't trust AI companies to not alter the texts of those books if there's something in them that does not align with their worldview.
    So any book added to the online archive(s) by these companies is suspect.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 12 points 2 months ago

    The point of bookburning is that it would only be accessible to AI, bringing forth it's monopoly on said first edition. No one would upload it to IA in the process. It is deleting one source of knowledge out of existence.

  • permalink
  • fedilink
  • source
  • parent
  • [–] 2 points 2 months ago (1 child)

    If they were archiving digital versions so they could be read by thousands in the future, this would be a totally different conversation.

    They don't even need to be as selfless as the internet archive - just follow Googles lead and allow people to search and view selections the archive, with automatic opening of the archive when you are confident copyright is expired. Even a closed archive owned by a third party with a dead mans switch to open it in the future when the business model is done would be something.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [+] -13 points 2 months ago (9 children)

    Random bulk rate books aren't artifacts. The vast majority of them probably have a copy in Anna's Archive.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 9 child comments
    •
  • [–] 29 points 2 months ago (2 children)

    If they did have a copy in Anna's Archive, why the fuck would AI companies find physical copies, buy them, rip them apart, scan them, and then digitize them before destroying them?

    Remember, you don't get the digital copy. It's only used to train their models.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 5 points 2 months ago (1 child)

    Absolutely. Every AI company has Anna's Archive in their training data already. They want to differentiate their models by having more training data than their competitors. There are way more books out there that have never been scanned than those that have.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 6 points 2 months ago (4 children)

    The use of the word "rare" implies that they are not bought in bulk....

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 4 child comments
    •
  • [+] -6 points 2 months ago (1 child)

    Rare doesn’t necessarily mean culturally valuable. How many of these rare books are actually just old instruction manuals, travel guides, training materials, phone books, and dictionaries?

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 8 points 2 months ago

    The issue with that is that we don't yet know what is culturally valuable or not.

    Just because we don't see a need for the specific book now, doesn't mean it won't change tomorrow.

  • permalink
  • fedilink
  • source
  • parent