Designing social robots
This blog series is a documentation of my work in Human-Robot Interaction course at the University of Kansas.
After the first week, we have read multiple research articles to examine the general concepts, context and different design challenges in building a truly sociable robots. By definition, social robots are able to interact and communicate with human beings in a natural way. They are often physically embodied in various levels and forms. In this context, the designers strive to create robots that can exhibit peer-to-peer interaction skills, human-oriented perceptions and social intelligence. “Life-like” robots can be used in a variety of applications: education, home assistance, etc.
There are many design issues when it comes to building a robot that can act like a living creature. Throughout the evolution of A.I. and robotics technology, many have tackled the problem in different ways. ASIMO, created by Honda in 2000, is a humanoid robot that can walk, recognize moving objects, gestures and sound, which enable it to interact with human physically. Meanwhile, Kismet only has a head with moving eyebrows, lips, ears and eyes that allows it to perceive and react with facial expressions and emotions. Both examples address the importance of embodiment, emotion and human-oriented perceptions in building social robots. However, do we see these robots as peers or living beings? Perhaps not. Personality and intentionality distinguish human beings from robots.
Designing Personality
This is the next step in building social robots. AI assistance such as Siri and Alexa are simply able to response upon requests. We do not think they have a character. When we look at other AI from science fiction, such as J.A.R.V.I.S. from Iron man, we see a bit of a personality. Apart from Siri, J.A.R.V.I.S. is a highly intelligent home computer which can mock its own creator. In short, it shows a sarcastic personality, making it more humane than other AI.
However, J.A.R.V.I.S. is still far from a life-like being. Not only it does not have a physical body, but also there is no motivations rather than answering to Stark’s requests. Let’s look at a different example from Avenger: Age of Ultron. Ultron itself “wakes” up as a computer program acting with internal emotion stages and confusion. Unfortunately, the disturbed Ultron believes that it must eradicate humanity to save Earth. Ultron is just as socially intelligent as J.A.R.V.I.S.. However, Ultron has a purpose to destroy human race. Later on in the movie, with the superpower from the god’s gem, Ultron is capable of creating its own body from metal. To this point, Ultron is alive. Comparing to
J.A.R.V.I.S., we see Ultron as a real living being because it has a purpose of living and a humanoid body. Ultron’s physical body is able to feel pain when Scarlet Witch took its heart out. Ultron feels anger, hatred and fear.
Personality and motivation differentiate living creatures from robots. These factors inform how robots act and “live”. In the future, social robots will eventually become our peers, assistance, and even facilitators. With the growth of technology, robots will hold more essence of humanity with character and intentionality. Perhaps, soon we will pick our robots based on the look and personality, just as how we pick our pets.