There's a fascinating article on socially intelligent robots over at the New Scientist Web site. In short, roboticists are wrestling with how to make robots socially aware. Did you know that just having a robot show up is enough to affect people's moods? It's true.
Giving the Tin Man a Heart
To this point, most robots use a sort of brute force approach to interacting with their world, combining simple reactions based on sensor inputs with some higher-level planning functions in their software. But, according to Shuji Hashimoto, director of the Humanoid Robotics Institute at Tokyo's Waseda University, robots also need kansei, a Japanese word reflecting emotional intelligence.
Note that this is about human-robot interactions. (If any of you are fans of Douglas Adams' Hitchiker's Guide to the Galaxy series, you may have flashbacks to the Sirius Cybernetics Corporation and its creation, Marvin, the clinically depressed but wildly intelligent android.) Humans are complicated creatures and since even we humans are poor at understanding each other (check out the vast selection of self help books available at your nearest bookstore if you don't believe me) robots don't have a chance. Well, not yet, anyway.
Elizabeth Croft's team at the University of British Columbia in Vancouver, Canada, is using skin conductance, heart-rate, and facial muscle sensors attached to human volunteers to measure their physiological and emotional responses to robots. The goal is to integrate the information into the robot's controller to enable the robot to react appropriately.
What If I Just Want Machinery to Act Like Machinery?
These are interesting philosophical problems and they engender some very real engineering challenges. But I'm not sure I want an emotionally aware robot. First, we, as a species, are very good at anthropomorphizing. We ascribe human emotions and motivations to our pets on a regular basis (which doesn't work too well, since it just confuses the living daylights out of the poor animals). We talk back to (and fluently curse) technology when we feel thwarted by it. Well-intentioned but flawed attempts to have inanimate objects interact with humans (talking cars, automated phone systems) just make us crazed and cranky.
If, however, we do get to the point of having robot helpers in the home, figuring out the most effective way for robots to interact with us is vital. I just hope we're more successful at it than the Sirius Cybernetics Corp. and its Genuine People Personalities.
(Note that New Scientist is providing free access to the article "Antisocial robots go to finishing school" for one week only.)