Integrating economics and genetics models in information ecosystems

Abstract

Multi-agent software systems can be modeled as complex, dynamic systems in which agent adaptation and interaction occur continuously and concurrently. A genetics inspired view has agent adaptation occurring via the exchange of encoded agent characteristics (genes). An economics inspired view has agent adaptation driven by changes in prices and supply and demand. Interactions of economics and genetics models have a long history in evolutionary computation. This paper describes new work towards a synergistic combination of these views, in agent-based settings. The paper includes a re-statement of Holland's (1992) basic theories in an agent-based context, a discussion of different self-organization principles taken from economics, presentation of results from two agent systems that synergize economics and genetics models, and a discussion of future directions.