Comment by jandrese

Imagine a little known work from 1920 written by an author that died in 1955 featuring a boy wizard in a magic school who's estate sues J. K. Rowling in 1998 for copyright infringement. We might never have gotten any further books.

This probably seems unlikely, but it's the flipside of exceptionally long copyrights, especially ones held by corporate interests who hire lawyers specifically to enforce copyright. The growth of AI is only going to make this more of a problem in the future. Imagine a ContentID like system but on the concepts and themes of works.