Having a person read out a book, then using AI to convert that back into the original book feels like a very convoluted way of getting the text. Is this just a licensing thing? Is it to get the timing right of the captions?
The text for these books already exists, they're books that are being read out.