Love and Sex with Robots_The Evolution of Human-Robot Relationships

by David Levy

  There are of course many reasons an owner could develop a sentimental attachment to a particular object, but these reasons normally derive from something connected with the source of the object—perhaps it was a gift from a loved one, a memento of an emotionally important event in the owner’s life, or a personal possession the owner has used caringly for several years. What is different about the nature of the possession attachment felt for a computer is the element of control—the computer is at its owner’s every beck and call. Russel Belk’s 1988 paper “Possessions and the Extended Self” discusses the notion that we are “extended” by our possessions, they become part of us, extending us, whether they be material possessions or human “possessions” such as “my” friend, “my” partner, “my” spouse; and Beck cites David McClelland’s suggestion that the greater the control we exercise over an object, the more closely allied with that object we become.

  Thus, through the great measure of control we exercise over computers, we have the potential to become close to them. Because of the high level of use we make of them and the interactive nature of that use, computers have the potential to hold a special meaning for us, to strengthen the attachment we feel for them. Combine these with the potential to extend ourselves by virtue of our possessions and it is not difficult to imagine that the computer—controlled, interactive, used, and possessed—could create in us the level of attachment necessary to engender a kind of love. And if, as suggested by Frayley’s thinking, one’s capacity to experience romantic love depends on one’s attachment history, an attachment history that involved computers or electronic pets could provide a basis for the capacity to fall in love with robots.*

  How Proximity and Repeated Exposure Affect Falling in Love

  There have been a number of studies on the effect of proximity on attraction. In one of the earliest studies, conducted during the 1930s in Philadelphia, the addresses of marriage partners were recorded for some five thousand marriage licenses. It was found that 12 percent of the couples lived in the same building at the time they applied for a marriage license* while a further 33 percent lived within five blocks of each other. For a similar study, this one in Columbus, Ohio, during the 1950s, the investigators interviewed 431 couples and found that 54 percent of them lived sixteen blocks or less apart when they first dated, and for 37 percent of these couples the distance was five blocks or less. Surveys at MIT and the University of Michigan found similar results for couples living in student dormitories. The MIT study showed that the most important factor in creating emotionally close couples was the distance between their apartments—the closer they lived, the more likely they were to become friends, while the University of Michigan study indicated that roommates were much more likely to become close friends than were students living in different rooms several doors away from each other.

  The overwhelming conclusion to be drawn from these and many similar studies is that seeing someone frequently, referred to by psychologists as “repeated exposure,” creates a much more fertile atmosphere for friendship and love than seeing someone less often, and the proximity of their living quarters clearly has a significant effect on how frequently two people meet. If two people live close to each other, they are more likely to develop a familiarity than if they live farther apart—familiarity in terms of seeing each other more, spending time with each other, thinking about each other, and anticipating interaction with each other.

  It has also been shown that even without any personal contact with the other individual, repeated exposure to them generally creates a feeling of liking for them. The reason that repeated exposure appears to create such a positive effect on human attraction has been suggested by Ayala Pines to “arise out of an inborn discomfort that we all feel around strange and unfamiliar things.”5

  In an experiment conducted by Richard Moreland and Scott Beach at the University of Pittsburgh, four women pretended to be students attending classes. The women avoided all contact with the other students in the class, and they attended different numbers of lectures: One of them attended once, another ten times, one fifteen times, and the fourth one not at all. At the end of the course, the students in the class were shown photographs of all four women and asked about their feelings and attitudes to each of them. Even though none of the students had had any personal contact with any of the four women, their reported liking of each of the women was strongly related to how often that woman had attended the class—the one who never attended was liked the least, with the level of liking rising as the number of attendances in the class rose. The study also found that the more often a woman attended the class, the more likely she was to have been described by the students as attractive, interesting, intelligent, and similar to themselves.

  The common factor in the studies described above is that in each case the repeated exposure was to another person, but Robert Zajonc has shown that repeated exposure to almost anything increases our tendency to like it and that a direct correlation exists between the frequency of exposure and the level of liking. In one of his experiments, he pretended to be conducting a test of visual memory and asked his test subjects to look at photographs of different people, with each viewing lasting for thirty-five seconds. He varied the number of times each photograph was shown—some were shown once while others were shown two, five, ten, or as many as twenty-five times. Zajonc found that his subjects tended to feel more positively toward the person in a photograph if they were shown their photograph more often, indicating that the physical presence of the object of one’s affection is not a prerequisite for developing that affection. This result concurs with the phenomenon of pen pals falling in love without meeting, and also with its more recent and more prolific parallel—falling in love on the Internet.*

  Why People Fall in Love

  Let us now retrace our steps a little, from falling in love as a form of attachment to a discussion of this question: Why does Human X fall in love with Human Y rather than with Human Z? What is it about our partners that causes us to fall in love with them?

  We like or dislike a person according to how we feel in that person’s presence—“like” is a feeling for someone in whose presence we feel good. The extent to which we are attracted to someone has been found to depend on the number of positive and negative feelings we have toward that person and was expressed by Donn Byrne in 1971 as “Byrne’s Law of Attraction,” derived from investigations into the attraction feelings of students who volunteered to take part in psychology experiments. The formal expression of Byrne’s law does not exactly make for romantic reading—it looks like this:

  What this means is that the strength of attachment one feels for another person is governed by the strength of one’s positive feelings for that person in relation to the strength of all of one’s feelings for them. Here are two very simple examples: Let us assume that you have feelings about ten different facets of a particular person—their character, their looks, their personality, their conversational style, and so on—and that the feelings you have about each of these ten facets are identical in strength. If you have good feelings about eight of those ten facets and bad feelings about two of them, then the fundamental measure of attraction that you experience toward them is

  8 / 10

  because you have eight positive feelings for them out of a total of ten feelings (positive and negative). But if you have good feelings about only three of the person’s facets and bad feelings about the other seven, then the fundamental measure of attraction that you experience will be only

  3 / 10

  This somewhat cold mathematical approach to the magic of human attraction might appear, to the more romantically inclined reader, to be utter nonsense and just about as far removed as one could imagine from reality. But in fact the accuracy and usefulness of Byrne’s law has been proved in many psychology experiments since he first stated it—experiments in which various positive and negative emotions were manipulated and where the calculation of the “attraction” measure produced
the values that the experimenters predicted. And although Byrne’s results stem from the experimental-psychology laboratory and are derived from first impressions of one person about another, Byrne discovered “that the same factors found to operate in the laboratory are also found to operate in determining real-life friendship, love, courtship and marriage.”6

  Interestingly, Byrne’s Law also shows that we are more inclined to like someone when we are experiencing positive feelings for reasons that might not be associated with that particular person but which are causing the same feelings in them, such as both hearing good news (“We’ve passed!”) or listening to music that they both enjoy. Conversely, it has been discovered that two people will tend to be less attracted to each other, or even to dislike each other, if they are sharing negative feelings, such as “We’ve failed” or listening to music that both hate.

  A simpler way of inducing people to fall in love was investigated by a team led by Arthur Aron at the University of California at Santa Cruz. In 1991, Aron experimented by taking pairs of students who had never met, putting them in a room together for ninety minutes, and asking them to exchange intimate information, such as their most embarrassing moment and how they would feel if they lost a parent. Immediately following this part of the experiment, they were asked to stare into each other’s eyes for two minutes without speaking. At the end of the experiment, the two subjects left the room through different doors, in order to remove any possible feelings of obligation to see each other in the future. (Despite this cautionary ploy, the very first couple that took part in Aron’s experiment were married six months later.) All the students in the experiment were asked to rate the closeness of the relationship formed within their pair at different stages of the ninety-minute period, and the ratings were compared with those of a group of similar students who were asked to rate the closest relationships in their lives. A key result from the experiment was that after only forty-five minutes of interaction the relationship between the paired students was rated as closer than the closest relationship in the lives of 30 percent of similar students. Although there might have been some bias among the paired students when giving their “closeness” ratings, due to the fact that they knew they were involved in an experiment, this 30-percent figure suggests that self-disclosure can be a powerful and fast-acting device in getting someone to feel attracted to you.

  Talking intimately about one’s most embarrassing moments and baring one’s emotional soul as means of engendering affection from another person could prove to be a double-edged strategy. If a robot tried this on someone who was not in the mood to reciprocate, the response from the human might be to suggest that the robot needed therapy or that its software or hardware needed fixing. But the strategy works well when the behavior is reciprocated, because the other person will understand the emotional risk that lies in emotional self-disclosure, and if he or she is willing to share that risk, then the mutuality of the risk will likely become a bonding agent. It is well established that couples who experience the risk of physical danger together—for example, being in the same vehicle in a traffic accident—tend to bond strongly and swiftly.

  Measuring Love

  Neurobiologists Andreas Bartels and Semir Zeki of University College London reported in 2000 on an analysis of fMRI scans* of the brain activity of love-struck students† while they were gazing at photographs of their loved ones. In these cases, the brain-activity pattern was very different from when the same students looked at photographs of close friends with whom they were not in love. Bartels and Zeki also compared these scans with those taken of people in different emotional states and found that the pattern corresponding to romantic love was unique.

  Helen Fisher, an anthropologist at Rutgers University, carried out a similar experiment in collaboration with Arthur Aron, in an attempt to find results that supported the work of Bartels and Zeki. Her team analyzed the brain scans taken of seventeen recently smitten college students, ten women and seven men, whose ages ranged from eighteen to twenty-six and who had been in love, on average, for some seven months and whose feelings of love were at a more intense level than in the participants in the Bartels-Zeki experiment. The scans for each student were taken over a forty-five-minute period, during which the subjects were shown photographs of their loved one alternating with those of a familiar acquaintance of the same age and sex as their beloved but in whom they had no romantic interest. The scans showed that the experience of romantic attraction activated those pockets of the brain with a high concentration of receptors for dopamine, a chemical closely associated with states of euphoria, craving, and addiction.

  The uniqueness of the “in-love” brain scans could serve as the basis for robots to determine whether or not a particular human was falling in love with them. A robot who wants to engender feelings of love from its human might try all sorts of different strategies in an attempt to achieve this goal, such as suggesting a visit to the ballet, cooking the human’s favorite food, or making flattering comments about the human’s new haircut, then measuring the effect of each strategy by conducting an fMRI scan of the human’s brain. When the scan shows a higher measure of love from the human, the robot would know that it had hit upon a successful strategy. When the scan corresponds to a low level of love, the robot would change strategies.

  Ten Causes of Falling in Love

  The first systematic study of why someone falls in love with a particular person was published in the Journal of Social and Personal Relationships in 1989. Arthur Aron, Donald Dutton, Elaine Aron, and Adrienne Iverson modeled their study mainly on three earlier accounts of falling in love, obtained from surveys conducted by other psychology researchers.* One of these surveys was based on detailed written accounts of falling in love by students who had done so during the preceding eight months. A second study compared the experiences of two hundred attendees of a seminar titled “Love and Consciousness,” who wrote accounts of their experiences of falling in love or falling “in friendship.” For the third study, a questionnaire was compiled to investigate the subjects’ most recent experiences of falling in love and, in particular, the moment when they first experienced a strong feeling of attraction.

  A review of the reports on these surveys reveals eleven factors that appear to be major contributors to the process of falling in love. One of these factors, proximity, is an explanation of why people come to be in a situation that engenders love rather than a factor that causes love to develop when they are in that situation, and for this reason I have not included a discussion of proximity in this section.† We therefore have ten factors to consider, and in chapter 4 we shall see that most of these factors are equally applicable for engendering love, by humans, for robots.

  1. Similarity

  There is strong empirical evidence that people tend to like other people who are similar to themselves in one or more important aspect. It might be a similar level of education, similar attitudes, a common interest, a similar family or religious background, similar personality traits, similar social habits, or similarity in any of a host of other characteristics.* Similarity is thus one of the dominant reasons for initial feelings of romantic attraction. The first study to examine this phenomenon was carried out by Sir Francis Galton during the 1880s. His findings, and those of later psychologists, concluded that couples tend to be similar in all sorts of different traits: psychological traits, physical traits, and personality.

  There is evidence from psychology research that we like people better when they change to become similar to us, as compared to when they are consistently like us. This can happen because those who change in order to make other people happy are often perceived as being “nicer” than those who always try to make other people happy—it is the gaining that promotes attraction here, the earning of esteem rather than experiencing it from the first encounter.

  2. Desirable Characteristics of the Other

  Most of the studies of romantic attraction have revealed, unsurprisingly, that person
ality and appearance are two of the most important factors in engendering a feeling of attraction. Ayala Pines found that more than 90 percent of the men and women she interviewed about the factors that caused them to fall in love mentioned a characteristic of their partner’s personality, with women mentioning personality traits as a crucial factor slightly more often than men. But when it came to appearance, 81 percent of men said that they were attracted to the physical appearance of their loved one, while only 44 percent of the women interviewed said that they were attracted by the appearance of their man. Given the importance of appearance in the attraction process, it is easy to understand why sex-doll manufacturers choose sexually alluring appearances for their dolls, as we shall see in chapter 7, a policy that sexbot designers will inevitably follow.

  3. Reciprocal Liking

  Knowing that one is liked by the other appears to be one of the dominant factors in falling in love. This factor is emphasized in Shaver’s adult-attachment theory, in which the loved one (read “the cared-for” one) perceives themself to be loved by the love giver (read “the primary carer”), as a result of which the loved one knows that they are likable, which makes them feel good. And when we feel good in the presence of a particular person we are more likely to develop feelings of attraction toward them. One test of this factor came from Arthur Aron’s experiment described earlier,* which had a secret ingredient added. He told both people within a couple that the other one would like them. “That expectation had a huge effect,” said Aron. “If you ask people about their experience of falling in love, over 90 percent will say that a major factor was discovering that the other person liked them.”