ABSTRACT
In this chapter, the author focuses upon certain analogies between the prebiotic evolution of ordered polymers from a disordered soup and neural network pattern recognition. Prebiotic organic soups and neural networks appear to have strongly analogous traits. Both may be mathematically modeled by spin glasses. An established procedure for modeling the behavior of genetic polymer ensembles begins the simulation with a random initial "soup base" of monomers, dimers, and trimers. One polymer is chosen successively among the members of the polymer ensemble to serve as a template. One of the remaining polymers in the soup is compared to this polymer to see if it is sufficiently complementary to the template to bind and form a double-stranded complex. In computer simulations that implement this Hopfield prescription, approximately half of the polymers evolved into the selected niches after about 200 generations. Every polymer may act upon the energy landscape simultaneously to stabilize or destabilize other polymers.
