Agent Transparency

摘要

Humans are increasingly working with intelligent agents—from robots to complex networked systems—as collaborators rather than tool users, due to the tremendous advancements in artificial intelligence in recent years. In order for human–agent teaming and communications to work effectively, agent transparency is a critical component that needs to be implemented in the joint human–agent systems. This chapter reviews theoretical frameworks and recent findings on agent transparency, including operator trust calibration, situation awareness, workload, and individual and cultural differences. The implications of these findings for user interface designs are discussed. Examples of transparent interface designs include human interaction with a small robot, multiagent systems, and robotic swarms. The final section of the chapter discusses current challenges and provides suggestions for future research.