The ecology of computation
Bernardo A. Huberman · 2003
Summary form only given. The author has developed a perspective on computational ecologies as well as a theory of their behavior which explicitly takes into account incomplete knowledge and delayed information on the part of its agents. When processes can choose among many possible strategies while collaborating in the solution of computational tasks, the dynamics lead to asymptotic regimes characterized by fixed points, oscillations and chaos. Spawn in an ongoing project which utilizes idle computational resources in a distributed network of high-performance workstations. The possible existence of a universal law regulating the way in which the benefit of cooperation is manifested in the system has been considered.>