Just as a thought exper­i­ment, let’s con­sider how Nat­ural Selec­tion might “learn” in a sim­i­lar way to how our brains learn.

How do humans learn? Humans can learn by trial and error, from expe­ri­ence, from habit, from con­di­tion­ing, or from play­ing or tin­ker­ing. The com­mon thread for all learn­ing is behav­ior cou­pled with feed­back. By behav­ing in a cer­tain way and gath­er­ing feed­back about that behav­ior, we learn. Pos­i­tive feed­back rein­forces and refines behav­ior, while neg­a­tive feed­back lets us know when behav­ior is wrong. Any sys­tem with behav­ior, feed­back and the abil­ity to apply that feed­back to behav­ior, will learn. (Notice the cycli­cal aspect of this kind of system.)

Nat­ural Selec­tion ini­tially deter­mines behav­ior through ran­dom muta­tion. New genes come about by slight copy­ing mis­takes (essen­tially typos) in the genetic code. Since the muta­tions are ran­dom, the behav­ior will be ran­dom — at first. Then, pos­i­tive feed­back is given for behav­ior that helps, and neg­a­tive feed­back for behav­ior that hurts. Nat­ural Selec­tion actu­ally col­lapses feed­back, and the appli­ca­tion of that feed­back to behav­ior, into a binary step: either the gene sur­vives and passes on the accu­mu­lated feed­back and behav­ior, or it dies and col­lected feed­back and behav­ior are lost. In the brain, when neg­a­tive feed­back is received, it is used to change the brain, where in Nat­ural Selec­tion neg­a­tive feed­back is death. Nature makes up for this inef­fi­ciency by dis­trib­ut­ing this win/lose game over a mas­sive net­work of ani­mals. Only a very select few sur­vive, but they have a 100% pos­i­tive feed­back rat­ing. Every sin­gle ances­tor of theirs was a sur­vivor. The large net­work of ani­mals acts like a dis­trib­uted brain and has the appear­ance of learning.