You can be forgiven for thinking that this is a book about artificial intelligence. In fact, it's a book about the future of life on Earth over the next billion or so years. It's just that, the way things are going, it's a good bet that a big chunk of that is going to involve artificial intelligence.
This is also an optimistic future-speculation book, and thus is likely to be far rosier than the future will turn out to be.
The author also has a tendency to name-drop. The last chapter, in particular, can be skipped almost entirely, unless you enjoy reading dashing tales of hobnobbing with movers and/or shakers.

But all of that aside, there's a good amount of substance: What is AI good at now, and what will humans continue to be better at in the next few decades? What does superintelligent AI look and act like? How can we keep it in check? How can we reevaluate the goals we give the AI, if our original goals turn out to be misguided or obsolete?