Criticism From Artificial Agents: Prior Interaction Reduces Negative Effects

Abstract

The present study explored interaction between humans and artificial agents from a relational perspective. According to theories on interpersonal communication and relationships, criticism by known agents, as compared to unknown agents, could be less negative (the buffering effect of relationship) or more negative (the expectancy violation effect) to human-agent relationships. Using online survey tools, participants who had interacted with an agent were led to one of the same or different agent conditions, in which the same or a new agent gave negative comments (criticism) to participants. Participants evaluated the agent and the interaction with it before and after receiving the criticism. Results showed that a brief acquaintanceship with agents lessened the negative effects of criticism. Participants’ evaluations decreased to a lesser degree when the negative comments were given by the acquaintance agent rather than the stranger agent. Thus, the buffering effect of relationship was supported in human-agent relationships.

Keywords:

• Artificial Agent
• Chatbot
• Computers Are Social Actors
• Human-Computer Interaction
• Relationships

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Funding

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea [NRF-2018S1A5A8029299].