The Evolution of Genetic Representations and Modular Neural Adaptation

In nature we find beautifully complex organisms. The only way a stochastic search scheme like evolution can find such systems is by adapting the genetic representation. This work develops a theory on the implicit adaptation of genetic representations during evolution. Non-trivial genotype-phenotype mappings, neutrality, and the adaptation of phenotypic variability form the core of this theory. Complex, highly structured phenotypic adaptability, and rich innovatability triggered by neutral reorganizations of the genetic code can be explained in these terms. Some ideas are also transferred to neural adaptation.