Driven cellular automata, adaptation and the binding problem

The binding problem in neurobiology and the synchronization problem in distributed systems address a fundamental question, namely how can a collection of computational units interact so as to produce a stable response to environmental stimuli. It is proposed that the synchronization of responses of a complex system to a transient stimulus, socalled transient synchronization, provides a solution to this problem. Evidence is presented from studies involving inhomogeneous, asynchronous, adaptive cellular automata in support of this contention. In the transient synchronization approach, information is encoded as a distribution of response patterns in a global pattern space. The activation of such information requires a dynamic interaction between the system and its environment. Information is implicit rather than explicit. This permits a several orders of magnitude improvement in storage capacity over current models.