Shop 'Til You Drop II: Collective Adaptive Behavior of Simple Autonomous Trading Agents in Simulated Retail Markets

In a companion paper [9] we argued that human economic interactions, particularly bargaining and trading in market environments, can be considered as adaptive behaviors, and that the tools and techniques of adaptive behavior research can be profitably employed in modeling naturally-occUlTing markets or constructing artificial market-based systems. H groups of simple artificial agents interact to exhibit market-level behaviors that are similar to those of human markets, explanations of how the behaviors arise in the artificial system may be viewed as candidate explanations for the same behaviors in human markets. In this paper, we illustrate these arguments by means of an example. We present results from experiments where an elementary machine learning technique endows simple autonomous software agents with the capability to adapt while interacting via price-bargaining in market environments. The environments are based on artificial retail markets used in experimental economics research. We demonstrate that groups of simple agents can exhibit human-like collective market behaviors. We note that, while it is often tempting to offer explanations of human market behavior in terms of the mental states of the agents in the market, our agents are sufficiently simple that mental states can have no useful role in explaining their activity. Thus, explanations of the human-like collective market behavior of our agents cannot be phrased in terms of mental states; thereby inviting comparisons with Braitenberg's influential "law of uphill analysis and downhill invention", with eliminative materialism in the philosophy of cognitive science, and with dynamical-systemsbased analyses of adaptive behavior.