A broad, light and short read: Only valuable in the sense it gives a good overview of different usages of machine learning methods, for example coupling with physical models (that is different from the typical cases of machine learning done at Google, Facebook, Uber and telecom companies, because you have physical models to guide you, and don't have to rely on purely black-box methods).
Nothing more, and nothing less. Could be cut even shorter if some pages were not sent on personal backgrounds of enterpreneur-engineers.
Another free O'Reilly book which does a broad but ultimately half-baked overview of applications of machine learning across various disciplines such as agriculture, microbiology, materials science, education and genomics. It is organized as a series of interviews with experts in the industry.
While the breadth is decent, the book fails quite miserably in depth and the questions (are not more contextual) and the answers are not as insightful as could have been. the book is short but ultimately quite unsatisfying. some of the interviews sound like ads for the company in question. Opportunity wasted for a more in-depth informative book.