Back in the day, text mining for science was mostly ignored, even though it was technically illegal. Authors didn't feel threatened by models for spam detection

Back in the day, text mining for science was mostly ignored, even though it was technically illegal. Authors didn't feel threatened by models for spam detection or sentiment classification. Demanding money from poor academics was pointless (they'd just move on to a different author) and bad PR. I mean, books3 contained hundreds of thousands of copyrighted books, and people released it under their own name.