There's something about that that simply sounds dangerous to me. I can't put my finger on it, but there's a certain resiliency in keeping printed copies of reference materials: they cannot be changed, disappeared (other than unloading them into the bin), or made impossible to access (unless the library starts putting books behind lock and key). If I want to learn about gardening (for example), I'd much rather get a reference text at the library than search for stuff online... which half the time is clickbaity or AI-generated trash.

It's not like the librarians have unilateral choice here. Old books on the shelves get vandalized and stolen; new books are not easy to come by, due to reduced print runs and supply-chain issues. How many times have we heard complaints about Amazon orders being "print-on-demand", and the quality is horrible? And if a published book is typeset in original PDF format anyway, why not distribute it that way to begin with?

Librarians have the demand side to cope with too. Personally, I don't enjoy checking-out books from the library. They're heavy; they require a backpack to carry them; they're not ubiquitously available to me wherever I am; they need to be physically lugged back to the same place where I found them. So yeah, I'd rather have an eBook.

But I contend (not in front of librarians) that a book such as a "Lakota Language Dictionary" is irreproducible in electronic form, because scholars have striven to compile those in print form; they developed new orthographies and documented the existing ones; and any new electronic-format dictionary must be recompiled, retypeset, and re-edited to satisfaction for a new publisher. So we won't have the same materials.

I used to derive great joy from finding really old copies of the Vedas, or a Navajo dictionary, but mostly Hindu texts in the original scripts. And yeah, they were painstakingly compiled by British colonisers and oppressors. But that history is preserved because of those colonists having a scholarly interest in "Hindooism". And those Vedic texts, and Panini's grammar, will not be directly transcribed to eBooks. They may take photographic images of them and shove them into a PDF, but those volumes will be given short shrift, because they're all Public Domain anyway.

The money's in stuff that you can copyright and IP that you can defend. And that's where libraries and librarians are going to follow.

Well, you don't need to think too hard about this when sites like archive.org are in legal danger, and the dream of Google Books is dead. I had not considered the "everything on the Internet is AI/SEO slop now" - that's a good point too: even if the stuff exists online, it's often almost impossible to find.

A few months ago I half-remembered a quote from a famous philosopher. Google and Bing returned only the vaguest, most useless search results - basically assuming I didn't actually want the quote, but general information about the philosopher. So then I turned to ChatGPT, which asserted that no such quote existed, but here were ones "like it" (they weren't.) Finally I skimmed through all the books I had until I located it.