From Eerie to Aww: Embracing the Imperfect Side of Humans Can Improve the Consumer-Robot Interaction: An Abstract

Interactions between consumers and humanoid service robots will soon be part of routine marketplace experiences. It is still unclear, however, whether consumers might react favorably to these humanoid robots. Previous findings have been inconsistent. Some studies showed that people experienced high sensory stimulation from interacting with service humanoids and subsequently evaluated their service experiences more favorably (Esch et al., 2020; Yam et al., 2020), while other studies showed that people felt stressed and timid by interacting with service humanoids, resulting in overall discomfort (Kim et al., 2019; Mende et al., 2019). Such discomfort is also referred to as the “uncanny valley” effect (Mori, 1970). We wish to explore a novel route to counteract the uncanny valley effect. One solution to reduce the eerie response is to make the humanoids even more ‘like us,’ ironically (Hansen et al., 2005). Specifically, we seek to show that embracing the imperfect side of humans can improve the consumer-robot interaction, thereby dodging potential “uncanny valley” effects. In the field of social psychology, presenting ‘clumsiness’ tends to humanize social agents who are viewed as distant and superior and subsequently improve their likability (Aronson et al., 1966). Humans are not perfect, and humanoids do not have to be either. We seek to show that imperfectness of machines as a function of appearing clumsy can rather improve human-robot interaction. We predict that clumsiness of robots serves as low-power signals in human-robot interactions and helps reduce eerie responses since clumsiness might make the humanoids less threatening. Previous research suggests that people tend to trust service robots more when they feel more comfortable interacting with the robots (Pixteren et al., 2019). As we predict clumsy robots will make people feel more at ease and more in control (power) in human-robot interaction, it is expected that clumsy robots will also make people trust the robots more and subsequently be more favorable of the robots. Thus, paradoxically, people might find robots that make clumsy mistakes more likable if the ‘clumsiness’ of robots makes the robot appear less ‘eerie.’