The Story of Beau: Exploring the Potential of Generative Diaries in Shaping Social Perceptions of Robots

Hyungjun Cho, Tek-Jin Nam

The increasing prevalence of robots in our daily lives underscores the growing importance of how humans socially perceive them. While these social perceptions of robots are crucial to the success of human-robot interactions, more than traditional design attributes such as appearance and behavior may be required to draw intended perceptions. We explore a new approach to shaping social perceptions of robots using generative diaries. Generative diaries are automatically generated text entries that enable robots to communicate their experiences and perspectives to humans. We present and assess a case in which this was used to influence social perceptions of a public service robot named Beau. We designed a set of research prototypes, including Beau, a curation interface called HeyBeau, and a social media-like mobile application named IamBeau. Over 16 days, we deployed Beau in a university building and conducted an in-situ field study with 12 participants. Our findings indicate that generative diaries can enhance social perceptions of robots, particularly regarding warmth, by increasing awareness of robots’ intelligence, consciousness, emotion, identity, and desire for social communication. These findings demonstrate the potential of generative diaries in shaping social perceptions of robots and open up new possibilities for designing future human-robot interactions.