This is a bit of a strange book, because it's a book about innovation—literally capturing all "greatness"—written by Machine Learning researchers. It's short and easy to read, but doesn't entirely shy away from their ML research. Overall, it's one of my favorite patterns: someone sees an abstraction that deeply applies to their work, and over time, realizes the same abstraction expands to cover many different disciplines. How do we invent? How do major jumps scientific understanding happen?
This book is primarily about how these don't happen: they don't happen by planning. You cannot build an impactful, new invention directly. Instead, the path towards greatness has no clear path — only in retrospect is it clear how great things were created.
To this point, the authors convince me. It's real, and should impact how you approach creativity. However, the authors largely cannot answer, "well, how do we create greatness?" And, I suppose, that's fair. But, I would love to have a followup: "How Greatness Can be Stumbled Upon".