Why are humans scared of robots?
The Psychology Behind Our Feelings About Robots

An illustration of people scared of robots
© Goethe Institut London. Illustration: Carlos D'Agaro.

Some people love them. Others are interested but cautious. Still, others are fearful. Why do we have such different reactions to robots? The answer, both complicated and fascinating, depends on a number of factors. Some of these factors are inherent in the robot observed, while others are concerned with the observer.

By Linda Blair

Let’s start by clarifying what we mean by a robot. According to Anca Dragan at Berkeley, a robot is "a physically embodied artificially intelligent agent that can take actions that have effects on the physical world". This sounds rather complex, so I asked Alan Winfield at the University of the West of England and Kurt Gray at the University of North Carolina, both experts in robotics, to offer their definitions. Both described a robot much more simply, as ‘an embodied AI’. 

AI, or Artificial Intelligence, occurs when algorithms - rigid sets of instructions responsive to particular predetermined triggers - become able to change by modifying their own sets of instructions in response to learned inputs and data. When AI assumes a physical form, we refer to that physical form as a robot.

However, Takuya Nomura at Ryukuku University in Japan, who studies cross-cultural differences in attitudes to robots, takes a broader view. He defines a robot as an autonomous machine that behaves according to its own rules - so far, just as do other experts. But according to Namura, an embodiment isn’t necessary: for example, he includes computer graphics in his definition of a robot. This wider definition is typical of the more inclusive outlook that characterises the Japanese worldview generally. More on this later - it will help us understand the differences between Eastern and Western attitudes to robots.

Let’s turn now to the different forms robots can take. In addition to Nomura’s inclusion of non-embodied AI, robots may resemble machines, animals or humans - the latter also referred to as androids, or sometimes as humanoids. Their appearance has an important part to play in influencing our reaction to them. Robots that look somewhat human may intrigue and interest us, but if the robot seems too human-like it becomes unnerving. Masahiro Mori at the Tokyo Institute of Technology introduced this sliding range of reactions in a 1970 paper when he coined the term ‘the uncanny valley’, our descent into eeriness when we encounter a robot that’s ‘too human’ in appearance.

What is it about that human-like appearance that’s so unnerving? Kurt Gray believes it’s when we think the robot can sense and feel - when it has what psychologists call ‘attribution of mind’. His work with Daniel Wegner at Harvard confirms this assertion. The researchers presented participants with a machine called Kaspar. Kaspar appeared machine-like to some participants, with wiring and electrical components visible, while for others Kaspar had a human-like face. Participants rated the more human-looking Kaspar as more likely to have feelings, and they were more unnerved by the android than they were by the more machine-like Kaspar. Turning this concept on its head, Gray and Wegner also found participants were similarly disturbed when viewing a human who was described as ‘unable to feel pain, pleasure or fear or otherwise experience what a normal person can experience’, thus, neatly explaining why ‘humans’ devoid of all emotion are often the subject of horror films.

It seems, then, that it’s the extent to which a robot appears to be human that determines how we react to it. The more truly human we think it is - particularly if we consider it capable of sensing emotion - the more disquiet we feel. 

However, this is not so much the case with robots that assume an animal appearance, particularly if they resemble a household pet, a baby animal or any cuddly creature. Such objects are widely welcomed. One of the most successful examples - at least in terms of the help and comfort it’s brought to many - is Paro, a therapeutic robot modelled on a baby harp seal. It was created by Takanori Shibata at Japan’s Institute of Advanced Industrial Science and Technology, and has been shown to reduce stress, induce relaxation, and encourage interactions between various groups of patients and their caregivers. A cross-cultural look at attitudes to Paro found it’s loved by the British and many Europeans, as well as the Japanese. Perhaps this broad agreement isn’t surprising: pets are important to many in the West, and Professor Nomura told me that in Japan, pets are considered "as family". 

Thus, when we’re presented with a robot, it’s the form that predicts our reaction. Robots that appear to be pure machines, those that perform a specific function such as cleaning or helping make things, are generally accepted and rarely give rise to an emotional reaction - as long as they don’t resemble humans. Robots that look like cuddly animals are generally welcomed and enjoyed. However, robots that assume a human or near-human form are likely to unnerve us, although not everyone equally. 

We now need to consider the observer. What are the factors that help predict different reactions to the same robot? The first is previous experience. Professor Nomura offers a lovely example: no one these days is afraid of telephones; but 50 years ago, a telephone was approached with trepidation. Computers were also initially met with distrust by many, but as they’ve become more common, we’ve grown to accept them as useful tools. 

Nomura’s observation chimes with my own experience as a clinical psychologist. We know that for children, it’s the totally unexpected that’s most likely to trigger a fearful reaction or even a phobia. The best way to help them (as well as adults) overcome an irrational fear is, therefore, desensitisation - gradually familiarise them with the novel object, person or situation.

The second factor that determines our reaction to robots; particularly androids, is the way our own society views the man-made and natural entities around them. Frederic Kaplan at the Sony Computer Science Laboratory in Paris explains elegantly how this works in his paper, ‘Who Is Afraid of the Humanoid?’ Kaplan chooses to compare Western attitudes with those of the Japanese. This is wise: the variety of attitudes across the Eastern world is so great that an effective ‘East-West’ comparison wouldn’t be feasible.

The unique Japanese outlook is heavily influenced by the prevalence of Shintoism and Buddhism, both of which encourage a blurring between the realisations in nature and the productions of man. Professor Nomura put it well: "whereas in Western thinking there’s a clear distinction between nature and artefact, this is not so for the Japanese."

In the West, machines are created both to assist us and help us better understand ourselves. However, we - unlike the Japanese - consider humans to be superior to all other creatures, which in turn means we’re likely to fear any machine that might become ‘better’ or more competent than we are. The Japanese, without this sense of superiority over machines or nature, do not experience this fear and are therefore unafraid of newer, more capable robots. 

Kaplan sums up the difference: "In the Western world, machines are very important for understanding what we are. We think of ourselves by analogy with the way machines work. But at the same time, technological progress challenges our specificity. That is why we can at the same time be fascinated and afraid when confronted with new machines. In Japan, in contrast, machines do not seem to affect human specificity. The difference between the natural and the artificial is not so crucial and building machines is a positive activity in the search of the natural laws that govern the world."*    

Robot technology is advancing at an unprecedented rate. It has the potential to help humanity in so many areas - industry, health, education, and service. We in the West could avoid anxious feelings by refusing to create robots in human form - but in so doing, we will limit their potential. On the other hand, we could choose to lose our (unsubstantiated) sense of human superiority and work with robots as equal partners in whatever form is most appropriate for the task at hand.  

*Kaplan, Frederic, ‘Who is Afraid of the Humanoid? Investigating Cultural Differences in the Acceptance of Robots’. International Journal of Humanoid Robotics, Vol. 1, No. 3. World Scientific Publishing Company, 2004, pp. 14-15.