ABSTRACT
This chapter considers combinatorial optimization on "rugged fitness landscapes" as a crude model of technological evolution. It discusses optimization for "neighboring" uses, an issue which arises in some forms of product differentiation. The chapter describes the question of what kinds of products can be both complex and "perfectible". This issue is important in its own right in biological evolution, and may harbor hints about economic evolution. The chapter also considers an image borrowed from ecology, that of the coevolution of entities, where the "fitness" of each entity is governed in part by the other entities which form its niche. The actual web is a subset of the possible; and the actual evolution of economic webs must be a path through the ever expanding set of next possible economic webs. The statistical connectivity features of the resulting random graph, such as cycles, input-output matrices of intermediate products, graph radius can then be analyzed.
