Particularly about human vision and how visual information gets processed, priority and timing, etc. Some cool experiments (upside down face recognition).
He does a good job explaining the crux of a much deeper science; his intent is mainly to provide background knowledge for aspiring doctors, but his own research is full-on neuroscience. It's a good example of topological sorting, building topics up from no-knowledge. And of course any knowledge of neuro-anatomy is humbling :)