In an empirical study of personality recognition, social presence and social response in human-robot interaction, researchers acknowledge that robots are assimilating into our daily lives as designed to socially interact with humans, in caretaker or assistant roles, for example. Social characteristics of robots have to be believable and trusted by humans, relating to prior readings for this course about trust in technology and AI services, in order to work and be effective in our evolving society. The study looked at personality (traits “relating to thinking, feeling and behaving”) of AI in order to increase amount of research done on social aspects of robots (emotion, gestures, dialogue, etc.), highlight that robots have the capabilities of genuine social interaction rather than merely tools or objects, and lastly, robots with more advanced abilities and design focused on improved social components will improve human-robot interactions in the future. Intimacy and believability in social interactions with robots are crucial for future success. The study focused on assessing how people react to robots’ extroversion versus introversion through verbal and nonverbal cues.
Extroverted and introverted participants recruited from a large, private university in the study were randomly assigned to one of four potential conditions in a 2 by 2 balanced, between-subject design (social robot designed by Sony, AIBO, either extroverted or introverted personality) x (participant personality: extroverted or introverted). The social robot AIBO, that looks like a pet dog, could communicate with humans by recognition of 53 common phrases, and it could show emotion through expression with eyes, tails, ears, and light features. It also had learning ability and memory-processing capability. Lights of the extroverted AIBO robot were more expressively colorful with more frequent changes to response than the introverted AIBO condition. Motion was more exaggerated in the extroverted AIBO, as well. Specifically, the study looked to see how people recognize a robot’s personality, how people are affected by a robot’s social presence, and whether individual differences in response to social robots exist.
Results showed that people could successfully recognize the personality characteristics of a social robot, and people enjoyed the robot more when it had a “complimentary personality” to theirs. If the personality of the social robot complimented a person’s own personality, the subjects of the study reported the robot as being more “intelligent, more attractive, and more socially present.” This, to me, was surprising, as I expected individuals to approve more of the robot with a similar personality, instead of the complementary personality. Why did opposites attract, in this case? Is this always the case in human-to-human connection, as well?
However, studies in the future of human-robot interaction (HRI) will further compare human-to-human interaction to human-robot interaction in sociopsychological ways. HRI patterns can be predicted by spontaneous trait inferences, automatic goal inferences and social politeness/appropriateness, according to the study.
I think the findings of this study were super interesting, especially that robots that complimented the participant’s personality were found to be more “intelligent, attractive, and socially present”. This says a lot about how people interact with their Roomba, Alexa, Siri, etc. People who are more talkative and outward enjoy Siri or Alexa to converse with and prompt for responses. For introverted people, these personality feature’s of the AI/robots are found to be invasive and irritating. I think that customizing robots to fit a person’s interaction style could go a long way in their perception by society. It is ironic that the personality and interaction style of a robot is decided/written by a person. Even if we wanted robots to be more efficient or better than humans, at the end of the day they were still written by another human. Will we reach a day where robots can create AI/robots better than themselves?
It is definitely interesting that a lot of concepts in social psychology that are reflected by human interaction can also be seen in human-robot interaction. I certainly don’t doubt that humans can feel powerful emotions towards robots, especially when they are programed to exude personality traits compatible with our own. I fear that this connection can be taken too far, to the point where it could lead to human distress. In reality, humans can feel all sorts of ways towards a robot, but that emotion is never really reciprocated– that reciprocity is beyond a robot’s range of abilities. If a relationship cannot be fostered by both sides, what is really the point? Couldn’t an emotional dependence on a machine incapable of truly reciprocating these feelings be problematic? Also, while we do want to trust these machines, why do we need to feel connected to them on a personality level? I think those connections should remain between humans.