Home » GSCT밖의CT이야기

[사색]Robots Are A Cultural Technology

October.2010 No Comment

Andrew Zoz Brooks, Visiting Professor at GSCT, KAIST

 “Where’s my flying car and my robot butler?” For decades, robots have represented a vision of the promise of technology for an easier, more pleasant future. But this pinnacle of industrialized bliss is, not surprisingly, limited to industrialized countries; and not just because robots are expensive and difficult to make. Our ideas of what robots are and what they might contribute to our societies is heavily dependent on the cultures of those societies.

Even the word “robot” is itself a term of art, not science. It comes from the Czech word ‘robota’, meaning work, or labor; but work in an oppressive sense, as in serfdom. It was popularized in Karel Capek’s 1920 play Rossum’s Universal Robots, and the robots in the stage production were synthetic humans who ultimately rebelled. In other words, they were a science fiction analogue to the Jewish Golem myth, in which a creature is animated from a lifeless material such as clay, leading to complex questions of unintended consequences and responsibility for creation — a truly cultural, rather than purely scientific, origin.

Though from the moment they were named they existed as a kind of culturally acceptable — though cautionary — mechanical slaves, the engineers responsible for building them have always had plenty of good reasons that we should do so. Robotics scientists used to say that robots were needed for the “3-D” tasks that humans couldn’t, or shouldn’t, be performing: Dangerous, Dirty, and Dull. True to these words, the majority of early uptake of robotics was on the industrial assembly line, replacing humans in the performance of repetitive, relatively unskilled tasks. That’s where things stood for a few decades, until cheap computing and sensing power caught up with the raw mechanical strength of electric motors and servomechanisms.

Today, though, if one looks at the directions of robotics research and attempts at commercialization, one finds strong cultural dependencies. The clearest case in point is Japan, which currently has the highest density of robots per capita of industrial workers — at 295 per 10,000 workers; nearly double that of the next highest country. Although this mainly refers to industrial robots, a look at Japanese robotics research is instructive. Japan also leads the world in research into robots for replacing people in human interaction-heavy fields such as elder care and child care. This fact is inseparable from two major influences, one demographic and one cultural. First, Japan has a rapidly aging population; there is soon simply not going to be enough of a domestic human workforce to look after the very old and the very young. And the blunt truth about the second factor is that Japanese culture is more accepting of a robot taking care of family members than a foreign human.

Compare this with the situation in the United States, which ranks relatively low in industrial robot density (less than a third of Japan), but is at the forefront of research robotics. One can predict where the trends in US robotics research are focused, based on a knowledge of US corporate culture: military, space and oil exploration, and consumer products such as robotic vacuum cleaners, which at a shopping-mall price point are more a novelty than a serious labor saving device. In a culture which is happy to keep wages for menial jobs low, a “killer app” for personal robotics has remained frustratingly elusive.

In analyzing what we should work on in robotics, then, we should first start by thinking about what we as a culture expect from them. For example, what should they be able to do? Public expectations of robot competence in the West derive heavily from a cultural staple: the science fiction movie. The most successful film robots are human-like in terms of capabilities, but even more so in terms of character traits. Far from the obedient, impersonal serfs of early dystopic visions, they exhibit idiosyncracies, eccentricities, and other potentially negative human tendencies; they talk back, are emotional, willful and headstrong, and make mistakes. Should this be what we want from robots in real life?

While pondering this question, consider that personalities, and levels of interactivity and anthropomorphism of objects, are also heavily culturally dependent. Countries with animist histories, such as Japan, usually represent objects graphically with lifelike traits, such as faces and expressions; countries with theist histories, like the United States, do not. So too, the notion of a “perfect employee” differs. Should such a thing, if it could exist, always show deference to authority, or should it take the initiative? Should it do what it’s told, or refuse orders that are illegal or inefficient?

Even Asimov’s famous three laws of robotics, themselves also a product of art rather than science, contain enough inherent contradiction to fill the plots of multiple novels. Take just the one law that prohibits harming humans through action or inaction: what if a smaller active harm, like punishment, can help prevent larger forms of self-harm that might occur through failure to act? How ought a robot to make that decision, whether programmed with cultural norms or not? We must also keep in mind that slavery both reflects and exacerbates a corrosion of culture. Antebellum plantation culture in the United States, or the rapacious culture of conquest pervading German and Japanese industrialism in the Second World War, are hardly examples to be emulated, whether the underclass is truly subhuman or not.

What does this mean for Korea? Korea is a country that is struggling with the simultaneous blending of a traditional, hierarchical culture with breathtakingly rapid advances in distributed, information-based topologies. This is reflected in the state of robotics: the engineering is spectacular, but in many ways the research march is following a drum beat set overseas. Yet Korea is in many ways uniquely placed to set the lead as a result of both its geography and its postwar cultural shifts. Koreans have a respect for communal infrastructure that should be the envy of public works engineers in the US, yet they have become used to radical changes in fundamental methods of doing things, from information access to transportation, that cause sleepless nights in demographically anxious Japan or centrally controlled China. And the barrier to entry for sophisticated robotics projects is dropping fast. Korea’s small size and cultural makeup makes it ideal for a grand experiment in the integration of robots with ordinary society. How will they best fit in? That’s a question for cultural technologists, as well as robotics engineers, to answer.