Neural Networks and Deep Learning

How it reads

Neural Networks and Deep Learning is narrative nonfiction (technical computer science textbook, specifically), and its dominant mood is hopeful. Whatever it puts its people through, it believes things can get better. The pace is moderate — neither a page-turner nor a slow burn — so it holds up whether you read it in long sittings or short ones.

It's set in several locations, in the present day, told in the third person, staying close to one character. Its central concerns are identity, power and politics and class and society. In weight it sits in the middle: substantial without being heavy going.

It stands alone, so there's nothing else you need to read first or after. The ending resolves happily. At 224 pages it's a quick read. It's well known, the kind of book most keen readers have heard of.