The idea behind cellular automata is to appropriate concepts from cognitive neuroscience and behavioral science into things that were previously considered to be inanimate materials. The aim of this is to control low-level rules, like protein interactions, to achieve an error minimization function over a new space.