Love and Sex with Robots_The Evolution of Human-Robot Relationships
Page 14
Inexpensive and effective technologies that enable computers to measure the physiological indicators of emotion also allow them to make judgments about a user’s emotional state. Thanks largely to Picard, detecting and measuring human emotion has become a hot research topic in recent years. By measuring certain components of the human autonomic nervous system,* it is already possible for computers to distinguish a few basic emotions. A simple example of such measurements is galvanic skin response—the electrical conductivity of the skin. This has long been known as an indicator of stress and has therefore been employed in some lie detectors, but more recently it has also been used as a metric for helping to recognize certain emotional states other than stress. Heart rate is another easy-to-measure example—it is known to increase most during fear but less when a person is experiencing anger, sadness, happiness, surprise, and disgust, the last of these eliciting only the barest minimum of a heart-rate change. Yet another example is blood pressure, which increases during stress and decreases during relaxation, the biggest increase again being associated with anger.
It is a relatively simple matter to measure human blood pressure, respiration, temperature, heart rate, skin conductivity, and muscle tension using what are currently regarded as sophisticated items of electronic equipment. Research into “affective wearables,” usually items of clothing and other attachments that may be worn unobtrusively and come with electronic sensors for taking such measurements, will inevitably lead to the development of technologies that can monitor all these vital signs without our even noticing that we’re wearing them. By transmitting the measured data, affective wearables will thus enable robots to recognize and quantify at least some of our emotions, allowing them to judge our moods, based on our displays of emotion as they appear to the electronic monitors. For example, by combining the data from only four different measures—respiration, blood pressure volume, skin conductance, and facial-muscle tension—Rosalind Picard, Elias Vyzas, and Jennifer Healey developed an emotion-recognition system capable of 81-percent accuracy when distinguishing among eight emotions: anger, hate, grief, platonic love, romantic love, joy, reverence, and the neutral state (no emotion).
Additional help in detecting human emotion can come from auditory and visual cues. Facial-recognition technology is making dramatic advances, spurred on by the impetus of a fear of terrorism—the technology that today successfully identifies faces seen on a closed-circuit TV camera will tomorrow be identifying not only the person behind the face but also that person’s mood. Similarly with voices. Voice recognition has taken on increased import as a means of identification for security purposes, turning the sound characteristics of the human voice into measurable quantities that can act as an additional aid to identification. Iain Murray and John Arnott have investigated the vocal effects associated with several basic emotions, establishing links between voice characteristics and emotion that make possible the design of a voice-based emotion recognizer. This particular slant on the technology comes from the measurement of the pitch of a voice, the speed with which words are uttered, the frequency range of the voice, and changes in volume. Someone who is sad or bored will typically exhibit slower, lower-pitched speech, while a person who is afraid, angry, or joyous will speak louder and faster, with more words spoken at higher frequencies.
In summary, the creation of natural and efficient communication between human and robot requires that each display emotions in ways that the other is able to recognize and assess. But the emotionally intelligent robot must not only be able to recognize emotions in humans and to assess the strength of those emotions, it should also demonstrate that it recognizes the emotions displayed by its human. As the development of emotion-recognition and emotion-simulation technologies advances, so will the development of emotional intelligence in robots, and their relationships with humans will come to mirror a healthy human-human relationship.
Three Routes to Falling in Love with Robots
There are three distinct progressions that I believe will lead enormous numbers of humans to develop affection for and fall in love with robots. One route will develop in a humanlike loving way, as robots become more and more human in appearance and personality, encouraging us to like and to love them. This is a natural extension of normal human loving and is the easiest of the three routes to comprehend. And, just as with the Tamagotchi, the human tendency to nurture will help to engender in us feelings of love for robots.
Another route is via a love for machines and technology per se, sometimes called “technophilia.” People who “love” computers and machines do so in different ways. There are those who rush out and buy every new technological gizmo the moment it is put on sale—theirs is a love for all new technology. There are those for whom the technology converts into some other form of emotional or even erotic stimulation, such as pornography on the Internet or on a DVD. There are the technophiles, usually programmers but also those who love pressing buttons to make their gizmos do weird and wonderful things; theirs is a love of control, whether it is control by writing the programs that instruct their computers what to do or the much simpler form of control achieved by pressing the buttons on devices that have already been programmed. And the act of programming has itself been compared to sex, in that programming is a form of control, of bending the computer or the gadget to the will of the programmer, forcing the computer to behave as one wishes—domination.
A love of technology and its benefits was at first largely the province of the technically more adept, the economically upward mobile, and, predominantly, of adolescents and those in their twenties and thirties. As the cost of electronics has come down, enabling consumer-electronics manufacturers to create electronic toys and other products especially for children, so the age range of technophiles has widened considerably. Nowadays, with primary-school children and even preschoolers finding themselves the owners of a plethora of electronic products, we are creating future generations of adults for most of whom the latest gizmos will seem perfectly normal rather than amazing. And so it will be with robotics. Those who are born surrounded by electronics will grow up eager for and receptive to whatever new electronic inventions become available during their lifetimes. The love that yesterday’s children and young adults demonstrated for their Furbies and Tamagotchis will be the basis for the adults of the future to find it perfectly normal first to love their interactions with robots and then to love the robots themselves.
The evolution of loving relationships between humans and robots will be yet another example of how technology changes the way we live in dramatic, even mind-boggling ways. One of the most glaring examples from the twentieth century is television. Who at the time of the First World War would imagine that one day they would be able to look at a box that showed something happening, at that very moment, on the other side of the world, or even on the moon? Who at the time of theSecond World War would have believed that by the end of the century telephone booths in the street would fast become redundant, because just about everyone would be walking around with their very own wireless telephone in their pocket? Who at the time of the Vietnam War would have expected handwritten letters to gradually go out of fashion in the United States and many other countries as more and more people would take to the computer as their primary or sole means of writing letters and even sending them, at virtually no cost, to their friends and relatives, in no more time than it takes to click a computer mouse? And which did you, dear reader, use more recently as a source of information, a reference book or an Internet search engine such as Google?
The entertainment industry has been reshaped more than most by the tools of technology. Animation, made so popular for generations of children (and adults) by Walt Disney and originally hand-drawn, painstakingly, by teams of artists, is nowadays created automatically by superfast computers, costing animators their jobs by the thousands. Music that in my youth came into our homes on gramophone records that rotated at 78 revolutions per minute, and later at 45 and then 33 r
pm, the slower speeds allowing more music to be stored on a single disc, now comes to our handheld boxes by “download” via the Internet, making available to us a colossal collection of pop, rock, jazz, classical, and all other types of music without our having any need to go to a store. And then there are video games, probably the biggest-ever product success in the entertainment industry—games that today offer the user the most amazing sights, sounds, and action, all in an easily portable package. Other video-based products such as DVDs and their precursors—videocassettes—have also created huge changes in the way we entertain ourselves, enabling us to have the movies of our choice in our homes, to watch and watch again as often as we wish. (And the genre that has achieved the biggest financial success in that particular technological field is pornographic movies, because sex seems always to find a way to reach the marketplace. Sex sells.)
But back to robots. A third route in the evolution of love for robots will arise out of emotions similar to those that have made Internet relationships so hugely popular. Let us recall Deb Levine’s words, quoted in chapter 1:
For some people, online attraction and relationships will become a valid substitute for more traditional relationships. Those who are housebound or rurally isolated and those who are ostracized from society for any number of different reasons may turn to online relationships as their sole source of companionship.6
The same could equally be said of human-robot relationships, and some will find this worrying. Most people who develop emotional attachments to robots, and to whom robots exhibit their own demonstrations of love, will have in their mind the knowledge that the robot is just that, a robot, and not a human being. This “you are only a robot” syndrome will be some kind of boundary across which a human must pass to feel love to its fullest extent for a robot, though in the case of certain groups within our society, crossing that boundary will seem perfectly natural. Those who prefer to relate to computers rather than to humans will doubtless find it no problem at all. Nor will many nerds, many social outcasts, and those who will be only too happy to find someone, almost anyone, who exhibits affection for them. But what about the more normal members of the population? What will it take for them to cross this boundary? One could argue that the first requirement will be incredibly good engineering, so that robots are as convincing in their appearance and actions as Stepford wives—almost indistinguishable from humans. But as we saw in chapter 3, the Tamagotchi experience and the reactions of the owners of AIBO pet dogs indicate that very strong emotional attachments can develop in humans even when the object of such affection is not humanlike in appearance.
This aside into the world of Internet romances and its implications has another important point to make in my line of argument on the subject of love with robots. One conclusion that can safely be drawn from the phenomenon of falling in love via the Internet, as with a pen pal, is that it is not a prerequisite for falling in love ever to be in the presence of the object of one’s love. The falling-in-love process can be conducted completely in the physical absence of the loved one. This is consistent with and a much stronger form of the phenomenon noted by Robert Zajonc.* Of course there are photographs and video images of the loved one that can be received via the Internet. And the loved one’s voice can be heard via the Internet or the telephone, but their physical presence is simply not necessary.
Now consider the following situation: At the other end of an Internet chat line, complete with a webcam to transmit its image, a microphone to carry the sound of its voice, and a smell-detection and-transmission system to convey its artificial bodily scent to you, there is a humanlike robot endowed with all of the artificially intelligent characteristics that will be known to researchers by the middle of this century. You sit at home looking at this robot, talking to it, and savoring its fragrance. Its looks, its voice, and its personality appeal to you, and you find its conversation simulating, entertaining, and loving. Might you fall in love with this robot? Of course you might. Why shouldn’t you? We have already established that people can fall in love without being able to see or hear the object of their love, so, clearly, being able to see it, finding its looks to your liking, being able to hear its sexy voice, and being physically attracted by its simulated body fragrance can only strengthen the love you might develop in the absence of sight, sound, and smell.
And if you do fall in love with a robot, what will be the nature of this love and how will it differ from the way you feel about the love of your life in the world as it is today?
As noted earlier, one important difference will be that robots are going to be replicable, even to the point of their personality, their memories, and their emotions. Those readers who are frequent computer users will know that it is good practice to back up your work on the computer just in case of a disaster that causes the loss of some or all of your data. Similarly, it will become common practice for the knowledge, personality, and emotion parameters—and all the other software aspects of a robot’s “brain”—to be backed up on a frequent basis. By midcentury this process will almost certainly be fully automatic, so that neither the robot nor its owner needs to do anything. At regular intervals the contents of the robot’s brain, its consciousness, its emotions, will all be transmitted to a secure memory bank. If, heaven forbid, a robot is damaged or destroyed and its owner wishes an exact copy, the physical characteristics can be replicated in the robot factory, and then the contents of the brain, predamage, can be downloaded into the new copy of the original robot. This capability creates one enormous difference between the love one feels for another human being and the love that will be felt for robots. If you love someone enough, you will willingly undertake any risk, or knowingly sacrifice your own life, in order to save theirs. This is only partly because of the strength of your love for them. It is also partly because they are irreplaceable. But in the case of love for a robot, it will be as though death simply does not exist as a concept that can be applied to the object of your love. And if it can never truly die, because it can always be brought back to life in an exact replica of its original body, there will never be any need for a human to sacrifice their own life for their robot or to take a major risk on its behalf.
Another important difference is that robots will be programmable never to fall out of love with their human, and they will be able to reduce the likelihood of their human falling out of love with them. Just as with the central heating thermostat that constantly monitors the temperature of your home, making it warmer or cooler as required, so your robot’s emotion-detection system will constantly monitor the level of your affection for it, and as that level drops, your robot will experiment with changes in behavior aimed at restoring its appeal to you to normal.
Robot Personalities and Their Influence on Relationships
Personality is one of the most important factors that drive the processes of falling in love and falling in lust, so before we examine the specific causes of falling in love with robots, we shall first consider some of the significant research on robot personality that has been conducted during the past decade or so.
Robot personality is a subject that some readers might regard with skepticism—how can a robot have a personality? In the mid-1990s, Clifford Nass and some of his colleagues in the Department of Communication at Stanford University showed it to be relatively straightforward to create humanlike characteristics in computers—computer personalities—using a set of cues drawn from the extensive literature on the subject of human personality. In psychological terms, personality is the set of distinctive qualities that distinguish individuals. Nass and his group have conducted more than thirty-five experiments to investigate some of these qualities, to determine how they can be simulated in computer programs and how such simulations compare with the corresponding trait in humans.
One of the experiments carried out by Nass’s group is related to the team element of a partnership relationship. Couples act as a team in myriad ways: She might wash the dishes while he dries, she mig
ht do the laundry while he does the gardening, he might be the principal breadwinner while she devotes more time to taking care of the children—or vice versa. It is not only the drudge tasks that are shared in a partnership relationship, it is also the more pleasant ones, and in both cases the sharing of responsibilities will often act as a bonding factor, helping to sustain the relationship. A study of computers as teammates is therefore of considerable interest in estimating how a computer-human dyad might also function as a team.
Nass and Byron Reeves based their study into computers as teammates on social-psychology experiments showing that there are two key factors in a team relationship—group identity and group interdependence. Group identity simply means that a team must have something to identify it by, often just a name such as “Mr. and Mrs. Bloggs,” or “the Smith family,” or “Christine and David.” The importance of group interdependence lies in the fact that the behavior of each member of a team can affect all the other members.*
The teams created for this study each consisted of a human and a computer, with the team identified by a color and the members of the team sporting a ribbon of that color and a notice saying “blue team,” for instance, on that team’s computer. Half of the people in the experiment were told they were on the blue team. They were also told that their performance would be graded and that its final evaluation would depend not only on their own efforts but also on those of the blue team’s computer. The other half of the people in the experiment were treated as though they were not on the same team as the computer with which they were collaborating. These subjects also wore a blue ribbon, but their computer was dressed in green and carried a notice affirming that it was a “green computer.” The experimenters made no mention to the humans in the second group of any collaboration between them and the computer, in order to avoid creating an association of teamwork in their minds. These subjects were told that their performance would be graded solely on the basis of their own work with the computer—that the computer was simply there to help.