Love and Sex with Robots_The Evolution of Human-Robot Relationships

by David Levy

  Another lesson from chapter 1 on the subject of getting someone to fall in love with you was that self-disclosure of intimate details can be a powerful influence in this direction. Robots designed to form friendships and stronger relationships with their users will therefore be programmed to disclose virtual personal and intimate facts about their virtual selves and to elicit similar self-disclosure from humans.

  Now to the ten reasons for falling in love. Which of them might have parallels in human-robot relationships, parallels strong enough to lead humans to develop feelings of love for robots?

  Similarity

  Of the most important similarities referred to in chapter 1, only one of them—coming from a similar family background—is not easy for a robot to imitate convincingly, given that its human will know that the robot was made on an assembly line. But as to the other key similarities, I forsee no problem in replicating them, including the most important of all, similarity of personality. It will be recalled from one of Clifford Nass’s experiments, described earlier,* that not only do humans prefer to interact with other humans of similar personality, but they also prefer to interact with computers that have similar personalities to their own. That finding is of great significance when considering the importance of similarity of personality in the process of falling in love. Attitudes, religious beliefs, personality traits, and social habits—information on all of these can be the subject of a questionnaire to be filled out when a human orders a robot, or it could be acquired by the robot during the course of conversation. Once the robot’s memory has acquired all necessary information about its human, the robot will be able to emulate sufficient of the human’s stated personality characteristics to create a meaningful level of similarity. And as the robot gets to know its human better, the human’s characteristics will be observable by the robot, who can then adjust its own characteristics, molding them to conform to the “design” of its human.

  One example of a similarity that will be particularly easy to replicate in robots is a similarity of education, since just about all of the world’s knowledge will be available for incorporation into any robot’s encyclopedic memory. If a robot discovers through conversation that its human possesses knowledge on a given subject at a given level, its own knowledge of that subject can be adjusted accordingly—it can download more knowledge if necessary, or it can deliberately “forget” certain areas or levels of knowledge in order that its human will not feel intimidated by talking to a veritable brain box. This self-modifying capability will also allow robots to develop an instant interest in whatever are its human’s own interests. If the human is an avid train buff, then the robot can instantly become a mine of information about trains; if its human loves Beethoven, the robot can instantly learn to hum some of the composer’s melodies; and if the human is a mathematician, the robot will have the reasoning powers necessary to prove the popular mathematical theorems of that time. Not only will robots have extensive knowledge, they will also have the power of reasoning with that knowledge.

  Desirable Characteristics of the Other

  The key “desirable” characteristics revealed by the research literature are personality and appearance. Just as a robot’s personality can be set to bear a measure of similarity to that of its human, so it can be adjusted to conform to whatever personality types its human finds appealing. For a robot, as for a human, having a winning (albeit programmed) personality will be arousing in many respects, including sexually arousing. Again, the choice of a robot’s personality could be determined partly prior to purchase by asking appropriate questions in the customer questionnaire, and then, after purchase, the robot’s learning skills will soon pick up vibes from its human, vibes that indicate which of its own personality traits are appreciated and which need to be reformed. And when its human, in a fit of pique, shouts at the robot, “I wish you weren’t always so goddamn calm,” the robot would reprogram itself to be slightly less emotionally stable.

  A desirable appearance is even easier to achieve in a robot. The purchase form will ask questions about dimensions and basic physical features, such as height, weight, color of eyes and hair, whether muscular or not, whether circumcised (if appropriate), size of feet, length of legs (and length of penis, in the case of malebots)…. Then the customer will be led effortlessly through an electronic photo album of faces, with intelligent software being employed to home in quickly on what type of face the purchaser is looking for. The refinement of this process can continue for as long as the purchaser wishes, until the malebot or fembot of his or her desire is shown on the order screen. If it’s a pert nose that turns you on, your robot can come with a pert nose. If it’s green eyes, they’re yours for the asking. By being able to choose all these physical design characteristics, you will be assuring yourself of not only an attractive robot partner but also the anticipation of great sex to come.

  THE REPLIEE Q1 ROBOT WITH HER DESIGNER, HIROSHI ISHIGURO.

  Personality and appearance are far from being the most difficult characteristics to design into robots. Synthesizing emotion and personality are active research topics at several universities in the United States and elsewhere,* as well as in some of the robotics laboratories in Japanese consumer-electronics corporations. Creating a physical entity in a humanlike form that is pleasing to the eye is relatively straightforward, and the Repliee Q1 robot demonstrated in Japan in 2006 is perhaps the first example. By 2010, I would expect attractive-looking female robots and handsome-looking males to be the norm rather than the exception, all with interesting and pleasant (though somewhat unsophisticated) personalities.

  Reciprocal Liking

  Reciprocity of love is an important factor in engendering love—it is more likely for Peter to fall in love with Mary if Peter already knows that Mary loves him. So the robot who simulates demonstrations of love for its human will further encourage the human to develop feelings of love for the robot.

  Reciprocal liking is another attribute that will be easy to replicate in robots. The robot will exhibit enthusiasm for being in its owner’s presence and for its owner’s appearance and personality. After an appropriate getting-to-know-you period, it will whisper, “I love you, my darling.” It will caress its human and act in other ways consistent with human loving. These behavior patterns will convince its human that the robot loves them.

  Any discussion of reciprocal liking with respect to robots will inevitably suggest questions such as “Does my robot really like me?” This is an important question, but a difficult one to answer from a philosophical perspective. What does “really” mean in general, and particularly in this context? I believe that Alan Turing answered all such questions with his attitude toward intelligence in machines—if it appears to be intelligent, then we should assume that it is intelligent. So it is with emotional feelings. If a robot appears to like you, if it behaves in every way as though it does like you, then you can safely assume that it does indeed like you, partly because there is no evidence to the contrary! The idea that a robot could like you might at first seem a little creepy, but if that robot’s behavior is completely consistent with it liking you, then why should you doubt it?

  Social Influences

  With time, social influences undergo huge change. What was considered a social aberration fifty years ago or less might now be very much the norm. One important example of this is the tendency in certain cultures for young people to be strongly encouraged to marry within their own culture. Not only are there fewer influences on marital choice nowadays, from parents, peers, and society in general, but there is more resistance from young people to be molded into marital relationships dictated by their cultural and social backgrounds. Attitudes to robots will also change with time—now they are our toys and items of some curiosity; before long the curiosity will start to diminish and robots will make the transition from being our playthings to being our companions, and then our friends, and then our loved ones. The more accepted robots become as our partners, the less prejudice the
re will be from society against the notion of human-robot relationships, leading more people to find it acceptable to take robots as their friends, lovers, and partners.

  Filling Needs

  If a robot appreciates the needs of its human, it will be able to adapt its behavior accordingly, satisfying those needs. This includes those relationships in which the human’s needs relate to intimacy, even to sex, as explained in part two of this book. One can reasonably argue that a robot will be better equipped than a human partner to satisfy the needs of its human, simply because a robot will be better at recognizing those needs, more knowledgeable about how to deal with them, and lacking any selfishness or inhibitions that might, in another human being, militate against a caring, loving approach to whatever gives rise to those needs.

  Arousal/Unusualness

  This factor depends for its existence on the situation in which a human and the potential love object initially find themselves together, and not on the love object itself. The arousal stimulus is external to the couple. As a result there would appear to be no difference between the effect of a particular arousal stimulus on someone in the presence of another human and the effect of that same arousal stimulus on that same someone in the presence of a robot. In both cases the stimulated human will find the situation arousing, possibly even to the extent that it might make the human feel more attracted to the robot than to another human under the same circumstances. After all, in a situation that appears dangerous, would not a robot be more likely than a human to be able to eliminate or mitigate the danger?

  Specific Cues

  Absolutely no problem! After a trial-and-error session at the robot shop, you will be able to identify exactly what type of voice you would like in your robot, which bodily fragrances turn you on, and all the other physical characteristics that could act as cues to engender love for your robot at first sight.

  Readiness for Entering a Relationship

  As in the case of arousal, with this feature it is one’s situation that gives rise to the affectionate feelings. If you’ve just been dumped by your partner and are looking for a flirtation or a fling to redeem your self-esteem, your robot can be right there ready for all eventualities, with no need for speed-dating sessions or for placing an ad in the lonely hearts columns.

  Isolation from Others

  This is yet another factor where the circumstance dictates what happens. If you have a robot at home, you will be likely to spend considerable time in isolation with it—as much time as you wish.

  Mystery

  Robots are already something of a mystery to most people. Imagine how much more of a mystery they will become as their mental facilities and emotional capacities are expanded as a result of artificial-intelligence research. This is not to say that robots should be “perfect.” By having different levels of performance that can be set or can self-adapt to suit those with whom a robot interacts, the behavior and performance of the robot can be endowed with humanlike imperfections, giving the user a sense of superiority when that is needed to benefit the relationship. The element of mystery, like variety, will be the spice of life in human-robot relationships.

  What Does This Comparison Prove?

  I submit that each and every one of the main factors that psychologists have found to cause humans to fall in love with humans can almost equally apply to cause humans to fall in love with robots. The logical conclusion, therefore, is that unless one has a prejudice against robots, and unless one fears social embarrassment as a result of choosing a robot partner, the concept that humans will fall in love with robots is a perfectly reasonable one to entertain. It is possible that at first it might only be the twenty-first-century equivalents of Sherry Turkle’s 1980s computer hackers* who fall in love with robots, the latter-day versions of the young man who’d “tried out” having girlfriends but preferred to relate to computers. Yet robots in a human guise will be far more tempting as companions and as someone to love than were computers to Turkle’s generation of hackers. And even if the computer geeks are the first to explore love with robots, I believe that curiosity, if nothing else, will prompt just about every sector of society to explore these new relationship possibilities as soon as they are available. What we cannot really imagine at the present time is what loving a robot will mean to us or how it might feel. Some humans might feel that a certain fragility is missing in their robot relationship, relative to a human-human relationship, but that fragility, that transient aspect of human-human relationships, as with so much else in robotics, will be capable of simulation. I do not expect this to be one of the easier tasks facing AI researchers during the next few decades, but I am convinced that they will solve it.

  Robot Fidelity, Passion, and the Intensity of Robot Love

  For the benefit of most cultures, robots should be faithful to their owner/partner—what we might call robot fidelity.* Robots will be able to fall in love with other robots and with other humans apart from their owner, possibly giving rise to jealousy unless the owner is actually turned on by having an unfaithful partner. Problems of this type can, of course, be obviated, simply by programming your robot with a “completely faithful” persona or an “often unfaithful” one, according to your wishes. How different life would be for many couples if the possibility of infidelity simply did not exist. But, in contrast, while the infidelity of one’s robot might be something to be avoided by careful programming, the possibility equally exists for humans to have multiple robot partners, with different physical characteristics and even different personalities. The robots will simply have their “jealousy” parameters set to zero.

  Being able to set one’s robot to any required level of fidelity will be but one feature of robot design. It will also be appealing to be able to set the love-intensity level and the passion level of your robot to suit your desires. Your robot will arrive from the factory with these parameters set as you specified, but it will always be possible to ask for more ardor, more passion, or less, according to your mood and energy level. And at some point it will not even be necessary to ask, because your robot will, through its relationship with you, have learned to read your moods and desires and to act accordingly.

  Marrying a Robot

  For many of the readers of this book, any discussion on the history or current status of the institution of marriage will take place within the somewhat conservative confines of traditional Judeo-Christian thinking and attitudes and those of some of the other major world religions. Within these confines, marriage can only be the union of one man with one woman, a union intended to last for life, a union that usually has as one of its principal goals the creation of children. Yet this view of marriage is not the only view, because there are and long have been cultures within which marriage is viewed very differently. One of the most obvious examples of such differences is that between monogamy, one of the fundamental tenets of marriage in Western society, and polygamy, which is and has been the norm in many other cultures, including tribes in Africa, North and South America, and Asia, and a bedrock of religions such as Mormonism and Islam.* Surely if we are to enter a balanced debate on the history, the current state, or the future of marriage, our discussions should take into account all cultures, their customs, and how they regard marriage. Why should any of us assume that our own attitudes are inevitably the only correct ones and that cultures other than our own are in some way wrong?

  America is perhaps the best example in the world of a mixture of races, religions, and cultures that is, precisely because of its mix, fast becoming a society in which the tolerance and acceptance of nontraditional customs and ideas create the very basis of society as it evolves. In such a society, if it is to evolve and thrive harmoniously, such acceptance is an essential moral prerequisite. Sometimes we must accept that it is our own views that might be inappropriate, possibly because they are outmoded, and that the more radical, more modern views of others are more suitable for the times in which we live and for the future. This phenomenon, whereby changes in o
pinion lead to massive social change, has been seen in recent decades with attitudes to homosexual relationships.†

  The trend toward the toleration and acceptance of same-sex marriages is but one aspect of the changing face and meaning of marriage. The November-December 2004 issue of Harvard Magazine published a highly charged essay, “The Future of Marriage,” by Harbour Fraser Hodder,8 which, although primarily intending to examine how changes in demographics, economics, and laws have altered the meaning of marriage in America, actually makes a number of points that can also be used to support the prediction that marriage to robots will by midcentury raise no more eyebrows than same-sex marriages and civil unions do today. One such point is based on the observation by Nancy Cott, a Harvard professor of American history, that “marriage itself has therefore come in for a broad reassessment.”9

  The reassessment to which Cott refers is that due to the polarizing views of the advocates of same-sex marriage and their “family values”–oriented opponents. Cott explains that “as same-sex couples line up for marriage licenses at courthouses across Massachusetts, opponents predict the death of marriage itself. One side sees tragedy in the making, the other wants to rewrite the script entirely.”

  It is my belief that marriage to robots will be one of the by-products of the rewriting of the script, a belief rooted in the type of argument employed by those judges who have ruled in support of same-sex marriage. In 1998, for example, in a superior court ruling in Alaska, Judge Peter Michalski called the right to choose one’s life partner constitutionally “fundamental,”10 a privacy right that ought to receive protection whatever its outcome, even a partner of the same sex. “Government intrusion into the choice of a life partner encroaches on the intimate personal decisions of the individual…. The relevant question is not whether same-sex marriage is so rooted in our traditions that it is a fundamental right, but whether the freedom to choose one’s own life partner is so rooted in our traditions.” Michalski’s 1998 ruling and many since then have pointed the way not only to a liberalizing of the legislature’s attitude to same-sex marriage but also to a strengthening of the attitude toward the right to choose.