This nice overview of the differences between biological brains and modern computers highlights important reasons why our digital electronics and even neural network simulations running on them fall short of the abilities we take for granted in living systems. The list of points include content-addressable memory, variable clock rates, pointers as short term memory, lack of separation between software and hardware, unification of memory and processing, the underestimated complexity of synapses, self-organization, and the overall size of the systems.