The question wasn't "which ML use cases are potentially good", it was "which ML use cases are seen as predatory by users". So it doesn't matter that face detection has the possibility to be good, it's currently bad
I've read both of these, and don't recommend them.
House of Leaves has a very uninteresting plot, and raises more questions than it answers. It's physically painful to read, because most pages, you have to rotate the entire book every which way since the words go in spirals.
Infinite Jest is good, but way too long. A good 70% of the book could have been cut out or condensed. The writing is also intentionally bad, which makes it harder to read. There are definitely good lines, but it's kind of like DFW used a random sentence generator and some of the lines just happened to be amazing, but 90% of it is garbage.