Written by IEEE | October 13, 2016 | Updated: March 30, 2017
With many recent advancements in the world of robotics, if robots evolve to understand human emotion, the concern is that they’ll be in a position of control. As emotional beings, it’s no surprise that interactions with robots can become complicated. In fact, research has shown that humans are sensitive to robots’ feelings, going so far as to lie to spare them any pain.
Researchers at the University College London recently conducted a study that examined interactions between humans and robots. The robots used were humanoid in appearance — equipped with eyes and a mouth — and were programmed to react in three different ways. There was a silent robot that did not make mistakes, a second silent one that made one mistake and a third robot that spoke and could interact with humans through yes or no commands. Humans preferred interactions with the expressive robot that could speak and apologize for making a mistake even though it was the slowest robot.
At the end of the interaction with the expressive robot who made mistakes, the robot asked the human for a job. The participants in the experiment felt bad saying no to the robot. They didn’t want to upset the machine. The emotional side of AI ultimately goes both ways — people fear being under robot control, while also harnessing human emotion to show empathy for a fleet of expressive machines.