Positing a Sense of Agency-Aware Persuasive AI: Its Theoretical and Computational Frameworks

Abstract

AbstractThe notion of a persuasive technology (PT) that is autonomous and intelligent, and more importantly, cognizant of and sensitive to human sense of agency (SoA), i.e., the subjective feeling or judgement that oneself is in control of situations, remains to be theorized, conceptualized and elucidated. Three important questions have emerged from our investigations: (1) why does SoA matter in the design of PT, (2) what computational principles in artificial intelligence (AI) underlie an adaptive PT, and (3) how can this intelligent PT sense, make sense of, and respond sensibly to dynamic changes in SoA under complex settings? We elucidate in this paper our theoretical and computational frameworks to answer our research queries. For the theoretical aspect, we propose an integration of pertinent theories in the cognitive, social and neurosciences that explain the emergence and disruption of SoA. Using this integration as theory of mind, we propose a computational framework for SoA-aware persuasive AI that integrates methods in cooperative inverse reinforcement learning, causal inferencing, explainable AI planning and generative actor-critic learning.KeywordsSense of agencyPersuasive technologyArtificial intelligence