The Atlantic ran an article or two about the Books3 database, into which some 180,000 books were fed to form a corpus for Meta’s large language model LLaMa, among other purposes. There’s an app that lets you search for an author’s name to see if it’s in there. And, well:
I haven’t fully thought through (nor, I am heartened to see, have legal scholars) the implications of one’s books being thus used. But I was not best pleased that Meta had apparently acquired the text of my books from a site containing pirated material. After all, surely Mark Zuckerberg can afford the possibly dozens of dollars it would cost to acquire these three texts. Why must the heralds of the future be simultaneously grandiose and ungenerous?