Given our current understanding of how mirror neurons and human emotion works, a new study from the University of Duisburg Essen should come as no surprise. Astrid Rosenthal-von der Pütten and other researchers used fMRI to measure the reaction in human test subjects who watched video of a Pleo robot being either tortured or treated with affection. They also watched video of a fellow human treated either badly or affectionately. The test subjects exhibited similar reactions to the treatment of both the robot and the human, with only the intensity of their emotion varying. The bigger question, it seems to me, is whether we understand emotion and empathy well enough to add those qualities to our robots, so they'll feel empathy toward us as well. The original news release (in German) is available on the University website. An English language version is available on Eurekalert. IEEE Spectrum posted a more in-depth description of the research. And, if you can bare to watch, read on for the Pleo torture video.