Paper Abstract
"Human-centred artificial intelligence is a fast-growing research stream within the artificial intelligence (AI) and human–computer interaction (HCI) communities. One key focus of this stream is the enablement of trust between end users and the intelligent solution. Although, the current body of literature discusses and proposes a range of best practices for the design of user interfaces for intelligent solutions, there is a dearth of research how such interfaces are perceived by users and especially focusing on trust in these interfaces. In this paper, we investigate how the Big Five personality traits affect trust in AI-enabled user interfaces. We then experimentally verify which design best practices and guidelines proposed by Google enable trust in AI-enabled user interfaces for the different personality types. Initial results (n = 211) reveal that three of the Big Five personality traits – Extraversion, Agreeableness and Open-Mindedness – show a significant correlation between the degree of the personality trait and trust in the proposed storyboards. In addition, we identified significant positive relationships between the perception of trust by users and four out of the twelve design principles: review implicit feedback; connect the feedback to UX changes; create opportunities for feedback; fail gracefully and highlight failure. This paper is of a highly explorative character and provides first experimental results on designing for trust to the HCI/AI community and also highlights future research directions in the form of a research agenda."

The full paper is available through SpringerLink, provided that you have proper access rights.