Abstract
This paper examines the use of evolutionary programming in agent‐based modelling to implement the theory of bounded rationality. Evolutionary programming, which draws on Darwinian analogues of computing to create software programs, is a readily accepted means for solving complex computational problems. Evolutionary programming is also increasingly used to develop problem‐solving strategies in accordance with bounded rationality, which addresses features of human decision‐making such as cognitive limits, learning, and innovation. There remain many unanswered methodological and conceptual questions about the linkages between bounded rationality and evolutionary programming. This paper reports on how changing parameters in one variant of evolutionary programming, genetic programming, affects the representation of bounded rationality in software agents. Of particular interest are: the ability of agents to solve problems; limits to the complexity of agent strategies; the computational resources with which agents create, maintain, or expand strategies; and the extent to which agents balance exploration of new strategies and exploitation of old strategies.