How might we work with digital evolution to produce useful products? Although we have a well established practice of plant and animal breeding, our relationship to digital evolution is quite different. Our ancestors were able to go out into nature and observe many highly evolved and complex organisms. They found uses for some of these, such as the ancestors of, rice, corn, wheat, chickens, pigs, dogs, etc. They then bred them to produce the much improved domesticated plants and animals that we know today.
However, in the case of digital evolution, we are starting with very simple organisms that have not yet achieved the complexity to be useful, so our first objective is to evolve complexity. Yet we have no prior experience with managing the evolution of complexity.
Probably any attempt to guide the evolution of algae to become corn, through artificial selection in the context of a breeding program, would prevent such a transition from occurring. I believe that we will never be able to guide the evolution of complexity by the use of artificial selection. To facilitate complexity increase we need a new and different approach.
Our inability to guide the evolution of complexity increase is based on two limitations: our inability to imagine where we are going, and our inability to provide the appropriate selective forces.