Nadine is a move in robotics which raises a lot of questions. “Social” robots? Many years ago, a friend told me about new Japanese robots which were so lifelike it was almost impossible to tell they were robots. The question was, and remains, what actual use these robots are.
They are definitely a step up in artificial intelligence, a very big, and very necessary, step. AI in its classic form isn’t famous for its thoroughness, or much else. Quite the opposite; at consumer level, particularly in gaming, it’s notorious for a very low level of performance. It’s been a sort of DIY Holy Grail for robotics engineers, software, and the whole idea of a robot-inclusive society.
Social robotics is defined as “a field of research in to the use of robots in social spaces.” Helpful definition? Not very. This idea goes way back to Asimov’s unavoidable and very thought-provoking I, Robot series, and has been very much part of the general thinking about robots for decades. Now that it’s coming true, the social and functional issues are just as complex as the theory.
Robots don’t actually need to look or act human. Nobody would be too put out if C3PO was working as a receptionist. The mental associations with characters of that type, in fact, would pay for the robot, and the overall response would probably be positive. You’d attract a lot of curious business, at least.
Novelty, however, doesn’t last too long. The core values of a machine with a social side are highly debatable, particularly in a business context. High receptivity, and appropriate responses, particularly from a receptionist, are the required values in Nadine’s role. Apparently she’s good at her job, which is a factor that will outlast any degree of novelty.
Nor is Nadine just some sort of “talking phone queue.” Receptionists have to deal with a range of moving targets on a daily basis. Professor Nadine Thalmann of NTU Singapore, after whom Nadine is designed, makes the point that social robots can fill many niches in roles like companions, child care, health care, and more.
She calls them “media robots,” which is an interesting take and newer angle on a major point — as a medium of communication, a robot equipped with basic communications technologies has a lot going for it as a network operator, source of information, etc.
Telepresence/proxy social robots — When you the robot and you the person are in different places
Another version of social robots, the idea of “telepresence” involves using a robot as your proxy self, in which the robot imitates your gestures, and allows you to operate it from anywhere in the world. All you need is a webcam and a telepresence robot. A lot of people would like that, particularly those whose media schedule is clocking up a lot of frequent flyer points, personal security issues, and jet lag.
Nadine’s conceptual sibling and telepresence exponent EDGAR is also able to act on his own, respond to the people he meets, and quite probably deal with a range of situations independently.
The theory of social robots, however, raises a lot of questions about human society:
*Why robots instead of real people at all?
*Is a social robot less confronting, or more confronting?
*Do robot social interactions change the emotional nature of those interactions? Are you going to be more or less aggravated?
*What if the robot is more plausible than the people you talk to?
*Does communicating with a robot tell you “We’re just going through the motions here. You can’t argue with a robot.” This image has some possible major downsides. Most companies aren’t too good at client relations management as it is, without including a robot to emphasize the point.
*Are social robots going to be “discounted people”, fitting in to roles which interact with messy, sometimes nasty, and unpredictable humans on a “If someone hits it with an axe, we can just replace it” basis? If so, does that mean that the social robot culture is likely to be pro forma, unresponsive to human issues?
*Does information you receive from a robot get the same level of acceptance? Wouldn’t people ask to talk to a human, if they don’t like what they hear from the robot?
*Maybe robots can be trustworthy in ways people can’t? An imponderable, but a natural development of the dynamics of robot interactions.
Frankly, and while not wishing to detract in any way from Thalmann’s important achievements, I think the problem is likely to be people, not robots. Robot social skills may be admirable, even excruciatingly appropriate, but most people’s social skills could use some work. Talking to a highly civilized robot in a personal crisis may not be the ideal outcome for most people.
On the positive side, I think Thalmann may have found a way out of the classic image of a bland, thematically barren, gizmo-based robot-served future in to a more practical environment. Social robots may be the default realists of the future, the reassuringly blunt but neutral and useful elements in human issues.
Maybe emphasizing the differences will make people more human-aware? I can see a scenario where some kid tells the family robot “You were more fun before your upgrade”. I can also see a movie — “Nadine, come home”. Ah well…
