by Dr Dylan Evans

When the Czech playwright Karel Capek sat down in 1920 to write a play about humanoid machines that turn against their creators, he decided to call his imaginary creations 'robots', from the Czech word for 'slave labour'. Ever since then, our thinking about robots, whether fictional or real, has been dominated by the two key ideas in Capek's play. Firstly, robots are supposed to do the boring and difficult jobs that humans can't do or don't want to do. Secondly, robots are potentially dangerous.

These two ideas remain influential, but not everyone accepts them. The first dissenting voice was that of the great Russian-American science-fiction writer, Isaac Asimov, who was born the same year that Capek wrote his notorious play. In 1940, barely two decades later, while others were still slavishly reworking Capek's narrative about nasty robots taking over the world, Asimov was already asking what practical steps humanity might take to avoid this fate. And instead of assuming that robots would be confined to boring and dangerous jobs, Asimov imaged a future in which robots care for our children, and strike up friendships with us.

The most impressive entertainment robot currently on the market is undoubtedly the Aibo, a robotic dog produced by Sony. According to Onrobo.com, a website devoted to home and entertainment robotics, Aibo is the standard by which all other entertainment robots are measured. Special software allows each Aibo to learn and develop its own unique personality as it interacts with its owner. But at over a thousand pounds a shot, they aren't cheap.

Commercial products like the Aibo still have some way to go before they have the quasi-human capacities of 'Robbie', the child-caring robot envisaged by Asimov in one of his earliest short-stories, but the technology is moving fast. Scientists around the world are already beginning to develop the components for more advanced sociable robots, such as emotional recognition systems and emotional expression systems.

Emotions are vital to human interaction, so any robot that has to interact naturally with a human will need to be able to recognise human expressions of emotion and to express its own emotions in ways that humans can recognise. One of the pioneers in this area of research (which is known as 'affective computing') is Cynthia Breazeal, a roboticist at the Massachusetts Institute of Technology who has built an emotionally-expressive humanoid head called Kismet. Kismet has moveable eyelids, eyes and lips which allow him to make a variety of emotional expressions. When left alone, Kismet looks sad, but when he detects a human face he smiles, inviting attention. If the carer moves too fast, a look of fear warns that something is wrong. Human parents who play with Kismet cannot help but respond sympathetically to these simple forms of emotional behaviour.

Another emotionally-expressive robot called WE-4R has been built by Atsuo Takanishi and colleagues at Waseda University in Japan. Whereas Kismet is limited to facial expressions and head movements, WE-4R can also move its torso and wave its arms around to express its emotions.

The gap between science fiction and science fact is closing, and closing fast. In fact, the technology is advancing so quickly that some people are already worried about what will happen when robots become as emotional as we are. Will they turn against their creators, as Capek predicted? In the new Hollywood blockbuster, I, Robot (which is loosely based on an eponymous collection of Asimov's short stories), Will Smith plays a detective investigating the murder of a famous scientist. Despite the fail-safe mechanism built into the robots, which prevents them from harming humans, the detective suspects that the scientist was killed by a robot. His investigation leads him to discover an even more serious threat to the human race.

I, Robot is set in the year 2035, thirty one years in the future. To get an idea of how advanced robots will be by then, think about how far videogames have come in the last thirty one years. Back in 1973, the most advanced videogame was Pong, in which a white dot representing a tennis ball was batted back and forth across a black screen. The players moved the bats up and down by turning the knobs on the game console. By today's standards, the game was incredibly primitive. That's how today's robots will look to people in the year 2035.

Will those future people look back at the primitive robots of 2004 and wish they hadn't advanced any further? If we want to avoid the nightmare scenario of a battle between humans and robots, we should start thinking about how to ensure that robots remain safe even when they are more intelligent. Isaac Asimov suggested that we could make sure robots don't become dangerous by programming them to follow the following 'Three Robot Laws':

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey orders given it by human beings except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

At first blush, these three laws might seem like a good way to keep robots in their place. But to a roboticist they pose more problems than they solve. Asimov was well aware of this, and many of his short stories revolve around the contradictions and dilemmas implicit in the three laws.

The sobering conclusion that emerges from these stories is that preventing intelligent robots from harming humans will require something much more complex than simply programming them to follow the three laws.

Note on the Author: Dr Dylan Evans teaches robotics at the University of the West of England, Bristol.