Yes the author refers to a couple of archives.
Libgen, the one that we are most familiar with, is particularly useful for students/academics in the social sciences/humanities, as it has a large repository of books (old and new) in those disciplines.
Sci-hub, in contrast, is more useful for peer reviewed articles.
Interestingly the author is not that worried about books on math and physics since, as he argues, "simply paying close attention to reality" will recreate those ideas.