Can certain expressions and gestures help you decide who to trust? No matter how you answered this question, the real answer is probably yes. And surprisingly, these feelings of trust and distrust can stretch as far as to an expressive robot—specifically, MIT's Nexi.
In an interesting experiment run by a team from Northeastern University, Cornell University, and MIT, researchers have managed to identify specific facial expressions and hand gestures that, when displayed by a person (or humanoid robot), can bring about feelings of distrust and wariness.
Imagine the following scenario: You're playing a short game of prisoner's dilemma for real money with someone you don't know. How can you tell if you can trust this person or if he'll stab you in the back and take all the money? According to this research, a set of non-verbal cues such as touching the face, leaning back, or crossing one's arms can help us foresee someone's bad intentions.
To make things even clearer, Nexi, the expressive and eerily human-like robot, was brought into the game, and unbeknown to the participants, was controlled by researches to either perform these silent cues or not. When Nexi touched its face, leaned back, or crossed its arms during the initial conversations, participants were more likely to keep their money to themselves and not trust the robot.
Aside from the exciting implications for general human behavior, this research is also intriguing in a whole different aspect. If humans can build robots that are human enough and expressive enough to trigger actual feelings of trust and distrust, how far can these robots go before we realize they're only robots?
According to Robert H. Frank, one of the study's authors, “It makes no sense to ascribe intentions to a robot.“ However, he goes one to explain that "we have certain postures and gestures that we interpret in certain ways. When we see them, whether it’s a robot or a human, we’re affected by it, because of the pattern it evokes in our brain responses”.
This story, "Expressive robot helps researchers understand trust" was originally published by TechHive.