A recently published study in the PLOS One Journal seems to suggest that people have a difficult time shutting down robots that express an objection to being turned off. The study, titled ‘Do a robot’s social skills and its objection discourage interactants from switching the robot off?’ was conducted by a group of social psychology researchers out of the University of Duisburg-Essen – as well as one technology researcher from RWTH Aachen University. A total of 89 volunteers took part under the impression that the research centered around social or functional human-robot interactions. As those designations imply, participants were told that they needed complete either a practical or casual interactive set of tasks based on which group they were assigned to. However, although the existence of a power button and its operation were explained, no indication was given that it needed to be shut down following the completion of assigned tasks. Despite that, most participants expressed an intention to shut the technology down and, for 43 of the sessions, the robot effectively begged not to be shut off.
In some cases, that included expressions of fear about the process and questions about whether or not the user was going to complete the action. Participants who witnessed the robot objecting to being turned off were much more hesitant to turn the robot off, taking around twice the time to reach a decision. Thirteen individuals chose to forego pressing the power button entirely. Responses queries about the reasons behind decisions not to turn the robot off, users fell back on compassion and free will. Namely, the robot didn’t want to be shut down. Other respondents stated that they wanted further interaction with the robot or that they didn’t want to “do something wrong.” Surprise at the robot’s response was yet another reason given for not turning it off.
Those results are not necessarily surprising since the study was intended to address a theory centered on the fact that humans tend to treat inanimate objects as living things. In particular, the group was looking at whether the perception of autonomy would impact that interaction. However, those who had a functional interaction with the robot were also more hesitant than those who interacted socially, despite that those interactions made the robot “less likable” and reduced the stress of having to shut it down. The researchers theorize that response from the volunteers is the result of witnessing a sudden “emotional” response as compared to participants having time to become accustomed to that. With that said, they also indicate that because most participants were students, further research is still needed before any conclusion can be reached with regard to human-robot interactions.
