Although this book provides a new perspective on how to build more intelligent machines that centers around the neocortex, it reads more like a bold manifesto than a legitimate theory with supporting evidence. That key difference was a bit frustrating. However, the converse strength is that Hawkins uses many anecdotal analogies that visually illustrate the main model that the book proposes, the memory-prediction framework. For example, a key component is the ability for the neocortex to store “invariant representations” of a phenomenon that adapt to individual circumstances, like how you know how to catch a ball coming at you even though the trajectory of the ball will be different every time.