Trust, (Dis)Comfort, and Voice Quality: Anthropomorphism in Verbal Interactions with NAO

This chapter investigates the anthropomorphism of voice-based conversational AI interfaces, particularly the role of robotic voices in human-machine interactions. It discusses the psycholinguistic relevance of voices and the attribution of human properties to non-human entities. The study highlights the uncanny valley phenomenon, where high levels of human-likeness in robots can lead to rejection due to feelings of eeriness. The chapter also explores how contextual factors, such as the robot's physical appearance and the topic of conversation, influence anthropomorphization and user attitudes. Through a lab experiment with the NAO robot, the study examines the effects of synthetic and human voices on user perceptions of trust, sympathy, and likability. The results indicate that while human-like voices can enhance anthropomorphization, they may also cause cognitive conflicts when not aligned with the robot's embodiment or the conversation topic. This chapter provides valuable insights into the complex interplay between voice-based anthropomorphization and user acceptance, offering a nuanced understanding of the factors that shape human-robot interactions.