Chatbot Humor in HMC: When Confidence Backfires

Read full text Summarise with VIRAT
Authors: Feya Hammerström, Katrin Etzrodt
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

One of the enduring utopian visions of human–machine communication (HMC) is the idea of a socially fluent artificial agent—a machine that not only informs but connects, amuses, and builds rapport with its human interlocutors (Breazeal, 2002; Picard, 1997; Reeves & Nass, 1996). Humor, as a mechanism of social bonding, features centrally in this imaginary (Brause et al., 2025): research shows it can enhance social presence, likeability, and relational closeness in HMC (Johanson et al., 2020; Morkes et al., 1999; Tay et al., 2016). At the same time, humor is inherently ambiguous and context-dependent, relying on interpretation (Lynch, 2002; McGraw & Warren, 2010). However, research has predominantly focused on affiliative humor, while other styles, especially aggressive (beyond mild forms), self-defeating, and self-enhancing humor, remain largely unexplored (Shin et al., 2023; Yang et al., 2022; Zhang et al., 2021). This study asks to what extent the utopian promise of socially humorous AI holds up when users’ interpretations diverge from designers' intentions.

Fields of Study & Subjects

Social Robot Interaction and HRI Humor Studies and Applications AI in Service Interactions

Publication details

Publication date
1 January 2026
Language
EN
Document type
Preprint