ExplainabilityTransparencyTrustworthinesscs.HC
Why Did the Robot Cross the Road? A User Study of Explanation in Human-Robot Interaction
Zachary Taschdjian
November 30, 2020
Interacción
Key Takeaways
-
Different explanation styles matter for robot understanding: The study tested three types of explanations (contrastive, causal, and example-based) and found they have varying effectiveness in helping people understand why robots make decisions, suggesting that a one-size-fits-all approach to AI explanations is insufficient.
-
Social science insights improve robot explanations: The research demonstrates that explanations for robots are more effective when informed by social science principles rather than purely technical approaches, indicating that how we communicate AI decisions should draw from psychology and communication research.
-
Context is critical for meaningful explanations: By studying explanations in a realistic human-robot interaction scenario (the robot crossing a road), the work shows that explanations must be tailored to specific situations and user needs rather than being generic technical descriptions.
-
Explainable AI requires interdisciplinary thinking: The paper positions XAI as a social challenge, not just a technical one, suggesting that responsible AI development requires collaboration between computer scientists, psychologists, and other social scientists to create explanations that actually help people understand and trust AI systems.
Abstract
This work documents a pilot user study evaluating the effectiveness of contrastive, causal and example explanations in supporting human understanding of AI in a hypothetical commonplace human robot interaction HRI scenario. In doing so, this work situates explainable AI XAI in the context of the social sciences and suggests that HRI explanations are improved when informed by the social sciences.