Upward Influence Tactics: Playful Virtual Reality Approach for Analysing Human Multi-robot Interaction

The interest, the potential, and also the technical development in artificial intelligence assistants shows us that these will play an essential role in the future of work. Exploring the interaction and communication between human and artificial intelligence (AI) assistants forms the basis for the development of trustworthy and meaningful AI-based systems. In this paper we focused on the question how humans react to AI - more precisely, AI gents as robots - that act to influence human behavior and emotions by using two upward influencing tactics: Ingratiating and Blocking. For this purpose, we developed a playful virtual reality approach that creates a leader-subordinate relationship between humans and the AI agents in a factory environment. We explore how humans react to those agents. Among other things, we found that behaviors that are seen as likable in humans are perceived as distracting in robots (e.g., compliments used by the ingratiating tactic). Further, robots were perceived as a group and not as individuals. Our findings showed us directions and open questions which need to be investigated in future work investigating human-multi-robot interaction at the workplace.