Lee (a fellow Cyberpsychology student) and I visited the Science Gallery during Science Week to attend a lecture by Dr. Cynthia Breazeal, the founder and director of the Personal Robots Group at MIT. She introduced the attendees to the topics of Social Robotics and Human Robot Interaction, followed by a brief Q&A session.
Social Robotics
Social learning theory proposes that people learn new information and behaviors by watching other people. This is especially prominent in children, as they are social learners. The Personal Robots Group at MIT are applying learning theories and developmental psychology theories towards building child like robots that will have natural learning capacities. They are developing socially intelligent robots that will interact with humans in human-centric terms, work as peers, and learn from people. Instead of having people learn how to communicate with robots or how to teach them, instead they are developing robots that will be able to learn from humans through natural social interactions in real world environments.
The first step towards social learning in robotics is social mimicking or mirroring, as highlighted very recently by the research of US robiticist David Hanson. The MIT robot Leonardo demonstrated this ability back in 2005, Leonardo is an advanced robot capable of social learning, self-motivated learning and logical problem solving. Another robot MIT are working on is the Huggable – a multifunctional teddy bear like robot designed to be an educator, a communication device and a therapeutic companion. Through Huggable the MIT group has explored the concept of “the educator as a peer” going back to the theory of social learning and children.
Human Robot Interaction – “Are these robots naked?”
One of the most fascinating studies Dr. Breazeal highlighted during the lecture was a long term study where participants interacted with either a computer or with an anthropomorphised robotic device (has human like features e.g. a face). Both devices were weight management aids, acting as a social support, with a form of patient-therapist dialogue and input options for calories consumed, exercise etc. They found that participants interacted more often and for longer periods of time with the robot, despite the computer running the same software. They found that people felt a stronger alliance with and trusted the robot more. People named and clothed the robot (they did not do this with the computer). This research, and many other studies, has indicated that screen agents (e.g. the paperclip in Microsoft Word) are never as effective, and that the physical properties of the device do matter.
The Value of Robotics Today
Dr. Breazeal discussed how Robots don’t have to be anthropomorphic. Cars are becoming more like robots every day, and she suggested that they are no longer a tool but are a partner. She described the value of robotics – exploration, capturing data, and closer to her own research: using robotics to study the theoretical models of social and cognitive abilities in humans. Social robots can be applied to the study of how we communicate and solve problems and can offer insights into psychology, neuroscience and interface design. We can learn a great deal about our own behavior when we try to apply these theoretical models to robots.
“Robot’s aren’t going to replace us, they’re going to help and empower us, with strengths and abilities to compliment people.”
Questions relating back to science fiction concepts of robots replacing humans or overthrowing humanity were of course brought up during the Q&A session. Dr. Breazeal argued that these were fears that were culturally based and weren’t apparent in Eastern culture, in countries such as Japan robotics have always been associated with good and advances in robotics have always been embraced and encouraged. She suggested that because robots will learn from humans, their behavior will depend on what we teach them.
The “Un-canny-valley” was also brought up, this is a hypothesis that says that when robots look and act almost like real humans it cause a response of revulsion amongst actual humans. Dr. Breazeal argued that this was based on conjecture and was unreliable, that personal values, preferences and culture had a bigger impact upon response to anthropomorphisim.
Overall I was very impressed with the lecture by Dr. Breazeal – she demonstrated a serious passion for her work. Finally, to sum up what most of the lecture was like… when questioned about whether robots will ever having feelings she replied “Robots will not have human emotions, but maybe they can have robot emotions”.
