"there's obviously immense value in books and long-form reads"
Is there? Or are you just saying that because it's expected and you'd feel embarrassed if you said otherwise?
Humans are pattern matchers, what if we see patterns more easily from many examples, rather than one? What if we extract patterns more easily seeing them from many points of view instead of just one author?
Is a neural network better trained on one high detail photo, or a dataset of many photos?
Someone's got to have a better comeback than a downvote. You know where there's "obviously immense value"? Oil fields. People literally kill to control them.
Nobody kills to take control of a library.
At best you could say people get into massive debt for education. But at the same time, education is clamouring for online courses, videos, conferencing, teachers, interaction, and textbooks are widely considered a problem - low priority, low quality, a racket, going back years since Richard Feynman's famous story about reviewing them at least.
Books, especially academic books, are increasingly given away for free online - when people will pay for entertainment.
How many people learn from a teacher, a course, or learn by doing, vs how many actually learn from books?
People don't treat books the way they treat things they value. There may be immense value in books, it's not "obvious".
Is there? Or are you just saying that because it's expected and you'd feel embarrassed if you said otherwise?
Humans are pattern matchers, what if we see patterns more easily from many examples, rather than one? What if we extract patterns more easily seeing them from many points of view instead of just one author?
Is a neural network better trained on one high detail photo, or a dataset of many photos?