For example, we build computers to be efficient but brains evolved to be reliable. Most modern computers use the Von Neumann design shown in Figure 6.2, with a central processing unit (CPU) controlling everything, but if it fails, so does the whole system. However this design doesn’t work for evolution because biological parts fail regularly, so a brain that stops when one part does is too fragile to survive. Evolution needed a better way, so our brain doesn’t have a central processing unit.
We build computers at leisure then switch them on but evolution couldn’t do that because life never stops. It is like building a computer where the first part has to work or you don’t get to add the second! The bee’s neuron sliver then has to work for it just as our brain works for us. We don’t work on live computers because it is too dangerous but evolution had to, because brains are always plugged into life. The answer was a hierarchy of layers, where one layer runs things while another is evolving, and it was also a backup if the new one failed.
Finally, to survive, brains must add value, say by moving a creature towards light, as a sense like light detection is useless if it doesn’t cause action. Hence evolution let the brain direct the feedback loop between sensory input and muscle output in a way that computers don’t.
Our computers are centralized, single-layered (on a motherboard), and input-driven, but brains are decentralized, multi-layered, and direct feedback. Hence, decentralization, hierarchies, and feedback are evolutionary principles that explain the brain better than any computer design.