Human Agents and Intelligent Agents: An Ex Internet
Jerome Yen · 1997
To understand the strengths and limitations of software agents is veiy important. Many issues and limitations have been discussed in the literature, for example, sofhvare agents have limited learning capabilip. However, $we could see such limitations in a controlled environment and measure how Such limitations afect the agents' performance. Such research can be extremely helpfit1 to design better software agents. We developed an artificial market with networked computers to simulate the environment of Internet. This artijkial market allows players, such as, humans or software agents, to compete in a simulated dynamic market. We conducted two sets of experiments to study the limitations of software agents. The first set of five experiments were used to identrjj the knowledge or strategies that human subjects applied in competition. We classified the knowledge or strategies into four levels: perception, uses of simple model and heuristics, application of long-term strategies and psychological tactics, and prediction or forecasting. The second set of five experiments studied how much can humans transfer their knowledge to the software agents.