Marriage and Sex With Robots – An Ethical Perspective

Author David Levy has worked in the field of Artificial Intelligence since graduating from St. Andrews University, Scotland, in 1967, and is the author of the book, Robots Unlimited: Life in a Virtual Age. Read his recent interview here .

The trend of robotics research and development, from industrial robots to service robots to companion and carer robots for the elderly, has as its logical continuation the design and construction of partner robots, sufficiently human-like and sufficiently appealing in various ways to take on the role of a partner in a relationship with a human being. This trend immediately raises many questions relating to humans loving and being loved by robots, treating robots as life partners and being similarly treated by them, marrying robots and having sex with robots.

Love and Marriage with Robots

Nowadays scientists, psychologists and philosophers are asking, more and more often, questions such as “Can robots fall in love?” Such questions might seem to be unnecessary because love is an experience peculiar to warm-blooded mammals—anything in a programmed entity is merely a simulation. But if a robot exhibits all the same signs and behaviours of a human in love, and if it professes its love for a human, then surely, so far as the recipient is concerned, their robot is indeed in love with them, whatever being in love means to a robot.

An even more challenging psychological, sociological and ethical question for the future is “Will people fall in love with robots?” I believe the answer to be an unqualified “Yes”, even though this idea will be abhorrent to many people. But consider the millions of people who develop extremely strong emotional attachments to their pet animals, and the rapidity with which millions developed emotional attachments to their Tamgotchis, both of which are indications as to the strength of emotion that can be felt by humans for non-humans, and as to the feelings that millions will develop for robots when they are more emotionally sophisticated than both animals and Tamagotchis.

Let us now consider this notion, that people can fall in love with a robot. In the past, before the Internet was invented, many people had pen friends with whom they exchanged letters. Through this type of correspondence some people developed long-term friendships, occasionally falling in love with their pen friends and agreeing to marriage even without having met them. Moderately unusual, yes, but only moderately. In recent years this phenomenon has been replicated countless times in cyberspace romances. It is easy to understand how two people can fall in love on the basis of their communications with each other, even without physical contact. Much of the emotional basis for love is based on your feelings about your partner’s character, their personality, their interests, ideas, how your partner talks (or writes) to you, . . . so many things that can be communicated verbally.

In the future significant numbers of people will fall in love with artificial partners they meet on the Internet, including robots. Of course many people will find the idea distasteful at first, just as many people have found the idea of same-sex love, same-sex physical relationships and same-sex marriage distasteful, but times change. Homosexual behaviour that was, as recently as the nineteenth century, punishable in the USA, Britain and some other countries by the death penalty, and later by imprisonment, has since become widely regarded as a normal expression of one human’s love for another. And just as sexual mores relating to homosexuality have changed so much with time, so attitudes and laws relating to human-robot relationships will develop with time. The difference will be in the rapidity of these changes in attitude towards robots. Social change nowadays happens at a very much faster rate than it did in the nineteenth and most of the twentieth century, as a result of which I expect marriage with robots to be legalized in some countries by mid-century, with the state of Massachussets most likely being the first jurisdiction where a legally accepted human-robot marriage takes place. The prediction regarding Massachusetts is not only because of its comparatively liberal views to same-sex marriage but also because it is the home of countless high-tech companies and academic institutions working in the field of Artificial Intelligence.

Arthur Harkins, an anthropology professor at the University of Minnesota, caused astonishment in the mid-1970s when he predicted that, before the year 2000, the first test case of a human-robot marriage would be in the courts. At that time, the media bombarded Harkins with requests for interviews, many of which were on TV talk-shows with a phone-in audience, and “. . . as people called in, once they got over their initial shock, their next question was invariably consistent: ‘Where do I get one?’ ” Clearly Harkness’ estimated timescale was wrong, but today his ideas are very much on the robotics road map.

One argument that has been levelled against the idea of loving relationships between humans and robots is an ethical one – what does it mean to be human and is it ethical for humans to marry and have sex with this particular category of non-humans? To this I would ask: “Who has the right to legislate against what consenting adults do with consenting robots in private?” I believe that the Massachusetts legislature, and then others, will come to share this opinion.

Human love for robots will be seen as a perfectly normal, understandable and positive extension of the affection and love most humans feel for other humans and for their pets. True, it will be different, because robots are different, but in my view there will be very few ways in which the robots of the mid- and late twenty-first century will be inferior to humans, and research into human-computer relationships has shown that already many humans are happier interacting with their computers than with most other humans, while 34 percent of the adults in a UK survey conducted in 2003 believed that by the year 2020 computers will be as important to them as their family and friends. These findings suggest that even more humans will be even happier interacting with humanoid robots than with most other humans.

Consider what most people want from a life-partner, a spouse. All of the following qualities and many more are likely to be achievable in software within a few decades – your robot will be: patient, kind, protective, loving, trusting, truthful, persevering, respectful, uncomplaining, complimentary, pleasant to talk to, and sharing your sense of humour. And the robots of the future will not be jealous, boastful, arrogant, rude, self-seeking or easily angered, unless of course you want them to be. In short, your robot spouse will be everything you want of him (or her), submissive or domineering, exactly as intelligent as you are, or more so, or less. He or she can be made in whatever likeness you wish, a Gwyneth Paltrow, Julia Roberts, Brad Pitt or Michael Douglas look-alike, or a custom design created specially for you. (S)he can be as tall or as short as you wish, as fat or as thin, as dark or as fair.

Some of the implications of love and marriage with robots raise interesting ethical issues. Although humanoid robots are artificial people, will the humans who fall in love with robots somehow reduce the degree of artificiality, by endowing their humanoids with a measure of moral standing? Will it still be so clear that the status of humanoids is firmly “Not alive”? At MIT, Sherry Turkle has found that, in discussing virtual pets such as Furby, children have a notion that is part way between “alive” and “not alive”, a twilight status that children refer to as “kind of alive”, “almost alive”, or something similar. As humanoid robots become increasingly sophisticated and increasingly humanlike in their appearance and behaviour, the notion “kind of alive” will become an increasingly appropriate epithet for adults to apply to humanoids, until, eventually, it becomes almost irresistible to think of them and treat them as being “almost alive”. And with the inevitably changing view of what constitutes “almost alive”, robots will become regarded more and more as our peers, worthy of our affection, of our love. As a result of this change in perception of the “aliveness” of humanoids, one of the ethical conundrums that will face our children and grandchildren relates to what sort of rights humanoids will deserve. The debate on roboethics is very much focused on issues that we regard as the unethical use of robots. But what about the unethical treatment of robots? Should we not, in this debate, be speaking also on behalf of the robots of the future?

As the idea of humans marrying robots gains currency, we should consider this prospect not only in terms of what it will mean for society, but also what it will mean for the humanoids. Today most of us disapprove of cultures where a man can buy a bride without taking into account her wishes. Will our children and their children similarly disapprove of marrying a purchased robot? Or will the fact that the robot can be set to fall in virtual love with its owner make this practice universally acceptable?

Sex With Robots

One of the first objections raised by sceptics when discussing the possibility of sex with robots is “Why would anyone want to?” One answer to this question has been given by Joel Snell:

Sexbots would almost surely be programmed to be highly intuitive, keeping track of what worked and what didn’t. They would become better sexual partners as they learned more about their human counterparts, storing everything in their memory banks from gasps of pleasure to frequency of orgasm. Every time they had sex with a human, it might get better.

In order to be somewhat systematic in our own attempt to answer the same question, “Why should anyone want to?”, we should first consider the various reasons why humans engage in sex, and the extent to which our motivations for sex (with humans) will still be valid when considered with respect to sex with robots.

Sigmund Freud asserted that the goal of the sexual instinct is to obtain pleasure, and this is certainly one of the prime reasons for wanting sex. But there are several other reasons that have been enumerated by research psychologists, including boredom, adventure, recreation, and simply having an opportunity for a sexual encounter. For most of the reasons listed by research psychologists there is no justification for assuming any reduction in their importance when we consider sex with robots: the physical delight of orgasm will be the same, assuming that the robot’s construction provides for an accurate shaping and texture of its genitalia; while having an opportunity for a sexual encounter with a robot, being bored, and wanting adventure or recreation, are all reasons that will be equally applicable in human-robot relationships. Given the breadth of this overlap in motivations, it seems clear that there will be enormous numbers of people for whom these motivations for having sex with robots will prove sufficient to light their sexual fire.

It is of course not especially difficult to believe that humans who love their robots might experience a desire for sex with them. And as for owners who do not develop feelings of love for their robots – they might well purchase their robot specifically for its function as a sex object, so it becomes important to ask whether they will enjoy the sexual experience. I believe that they will. Although one commonly held view is that sex without some (human) love, or at least (human) affection, is not especially enjoyable, there is plentiful evidence to refute this position. Just consider all the people, and they are not only men, who admit to using the services of prostitutes on a regular basis. Clearly their encounters with prostitutes provide the pleasure and satisfaction they seek from sex, otherwise these clients would not be “regulars”. There are also those whose visits to prostitutes are infrequent – possibly they go for a specific and very occasional reason, such as wanting to lose their virginity or spending a few last moments of “freedom” before marriage. In all these cases the clients, unless they were deluding themselves, knew full well that their chosen sex objects did not love them or bear them any affection other than for the size of their wallets. Yet prostitution continues to thrive, so we must conclude that, for a significant proportion of the population, neither affection nor love is an essential element in a sexual relationship. In fact the opposite is often true – “uncomplicated sex”, “sex without commitment” and “sex with no emotional ties” are key reasons, repeated often to researchers in this field, to explain visits to prostitutes. And the same motivations equally apply to non-commercial casual sex. Will not the same reasons, the same expectations of pleasure, motivate many people to seek sex with robots?

Those who regularly enjoy sex with prostitutes are, of course, a minority, albeit a significant minority. Let us now consider the majority – those who do not pay for sex. Given that sex with robots will inevitably become a popular form of human sexual satisfaction, it is important for the roboethics community to consider the ethical implications of such encounters: the implications for the human participant in a sexual encounter with a robot, the implications for those (humans) who are close to the human participant, the implications for society as a whole, and even the implications for the robot participant.

For the human participant, certain of the benefits of robot sex create their own ethical justification. The capability of robots to teach all known aspects of sexual technique will turn receptive students into virtuoso lovers. No longer will a partner in a human-human relationship need to suffer from lousy sex, mediocre sex or anything less than great sex. Marriages and partnerships that today are in trouble in the bedroom, will no longer be at risk, thanks to the practical instruction in sex that will be available to all. And, having taken as much instruction as is necessary, to enable a participant to satisfy their human partner, the student’s sexual confidence will be boosted, creating a more balanced human being.

Another benefit of robot sex will come from robots’ therapeutic capabilities, in helping those who suffer from psychosexual hang-ups. Most people cannot afford to seek professional help in resolving such problems, and amongst those who do have access to human therapists a significant proportion are too embarrassed to discuss their sex lives with others. But robots, in addition to being excellent sex teachers, will also be sympathetic counselors, curing far more cases of psychosexual inadequacy than could ever be achieved by human therapists.

Yet another benefit for the human participant is the possibility of being able to experiment with different sexualities. People who are uncertain of their own sexuality will be able to try out same-sex and opposite-sex robots in total anonymity – the robots in question having been programmed to keep confidential the identities of the human experimenters and then, when the period of experimentation is over, to forget about them.

For humans whose partners are sexually involved with robots, there can be benefits over and above the joy of sex with skilled lovers. The availability of regular sex with a robot will dramatically reduce the incidence of infidelity as we know it today, though some human spouses and lovers might consider robot sex to be just as unfaithful as sex with another human. Will sexual ethics come to regard encounters with robots as being as innocent as the use of a vibrator is regarded today?

Yet another benefit of choosing robots rather than human partners for casual sex, lies in the confidence that sexually transmitted diseases are most unlikely to result, and unwanted pregnancy does not even come into consideration. In the early days of sexbots, STDs will be easily avoidable by following the simple rules of sexbot hygiene, and eventually sexbots will be designed with automatic self-cleansing systems that are activated by the robot itself when appropriate.

For society as a whole there are clear social and ethical benefits in making sexbots available to those who cannot refrain from indulging in illegal and antisocial sexual practices. In some cases the sexbot will be able to provide not only the satisfaction necessary to assuage the craving but also whatever therapy is necessary to cure the underlying problem that causes the illegal or antisocial practice. And in any event, the human will no longer be breaking the bounds of social convention, or breaking the law.

Because of all these benefits, I believe that amongst and for the vast majority of the human population, sex with robots will come to be regarded as ethically “correct”, as a good thing. There is, however, one aspect of robot sex that might cause the practice to be regarded by some as ethically suspect. When robots are so highly developed that, without inspecting their innards, they are almost indistinguishable from humans, should we assume that, simply because they are not biological creatures, it is totally acceptable for us to have sex with these objects of our creation whenever we wish? If robots become, for all emotional and practical purposes, surrogate humans, will we not have ethical obligations towards them? This question is another challenge for the roboethics community.

Read a recent interview of David Levy at

[tags]robot, sex, sex robot, roboethics, love robot, robotics[/tags]