However, it is currently unknown what variation mechanisms can give rise to protein circuits of the complexity found in biology, within realistic population sizes and realistic numbers of generations.
We suggest that computational learning theory offers the framework for investigating this question, of how circuits can come into being via a Darwinian process without
a designer.
(oops ) We formulate
evolution as a form of learning from examples. The
targets of the learning process are the
protein expression functions that come closest to best behavior in the specific environment. The learning process is constrained so that the feedback from the experiences is Darwinian. We formulate
a notion of evolvability that distinguishes function classes that are evolvable with polynomially bounded resources from those that are not. The dilemma is that if the function class that describes the expression levels of proteins in terms of each other, is too restrictive, then it will not support biology, while if it is too expressive then no evolution algorithm will exist to navigate it. We shall review current work in this area.