[3O1-GS-13] AI application: Dialogue system and mind

[3O1-GS-13-03] The effect of the difference between shared goals and means on trust in artificial intelligence

〇Katsuhiro Isokawa¹, Takuma Yonesugi¹, Tomoki Ishida¹, Misa Sano¹, Momoka Konishi¹, Ryo Ichiyama¹ (1. Doshisha University)

Keywords:AI, trust, Intrinsic motivate, goal setting theory

The purpose of this paper is to examine whether the difference between goal sharing and means sharing affects trust in AI.
[Yokoi & Nakayachi 19] indicated that sharing goals may promote trust rather than sharing means, based on experiments conducted in previous studies. We conducted a survey experiment 2 × 2 between subjects design for 218 participants, by considering difference of informants (AI or human) and value-sharing conditions (goals sharing or means sharing). As a result, we found the interaction between value-sharing conditions and information providers. Besides it was also found that sharing means with AI was more effective for trust, and on the contrary, sharing goals with people was more desirable. By sharing appropriate information between people and AI within an organization, it is possible to promote trust from people to people and from people to AI, and have a positive impact on improving corporate performance.