Her joints are "a bit tight and creaky." Her head circumference is smaller than expected, and there's "a slight chemical smell again." But "Mr-Smith" is mostly proud to introduce Page to other members of the message boards at DollForum.com. And they are happy to meet her, too: "Glad to see such an awesome lady of mystery!" one responds. "Have a fantastic honeymoon," types another. "Congratulations, she is a beauty," posts a third. "When you get around to completely introducing yourself to her, you will find that her softness will blow your mind."
Page is what's known as a "love doll" or "sex doll." She is "anatomically correct"—that is, built so people can penetrate her—but she doesn't move on her own or speak. There are at least a dozen high-end doll makers globally, and many more making cheaper models. "Even China is getting into it…in a year's time China has gone from being non-existent in the doll market to having like 15 different manufacturers," artist Stacy Leigh, who styles and photographs these dolls, told Acclaim magazine in 2013. "The world better be prepared, because love dolls are coming."
Katie Aquino, a futurist and self-proclaimed techno-optimist who goes by the name "Miss Metaverse" online, agrees that sex dolls and sex robots are poised to go big. But Aquino doesn't think improved industrial tech will be the main force driving the growth. Instead, she thinks hobbyists are the future: "I believe that the first truly lifelike sex dolls won't be made in factories, they'll be made in people's garages. Sex robots will be made by makers," she says, using a catchall term for the growing do-it-yourself subculture in everything from 3D printing to mead brewing. And she's mostly on board with this: "New sexual technologies will liberate us, allowing us humans to freely express our desires and fantasies while remaining safe and healthy from the comfort of our homes."
But Aquino also worries about possibilities like "a population decline because more people will choose synthetic relationships over 'organic' human relationships" and human women "comparing themselves to synthetics and therefore choosing to modify themselves, just as we see how Photoshopped models and celebrities affect women today." Some men are already predicting this day with glee, crowing on blogs and Reddit boards that human women will have to lower their expectations, step up their beauty rituals, or face the fact that many men will find sex robots a "better option."
On the other end of the spectrum, you have people like Sinziana Gutiu, whose presentation at the 2012 We Robot conference focused on how artificially intelligent sexbots could "foster antisocial behavior in users and promote the idea that women are ever-consenting beings, leading to diminished consent in male-female sexual interaction." In other words, she thinks sex robots may lead to more rape.
By promoting "lies about women's humanity," sexbots present "a danger that builds on and surpasses the harms attributed to pornography," Gutiu wrote in her conference paper. In this she joined the laments of social conservatives. "Sodom and Gomorrah never dreamed of sexual immorality like this," Jennifer LeClaire wrote last year in the Christian magazine Charisma. Dave Swindle, an associate editor at the conservative/libertarian site PJ Media, asked, "What happens when a bunch of teenage boys pool their money to buy a robot prostitute they can gang rape?…What will our world be when people lose their virginity to a machine?"
Is that last option even possible? Virginity is more a social construct than a physical state; we don't say someone whose hymen breaks using a Tampax or whose penis enters a Fleshlight have "lost their virginity" to tampons and sex toys. But it's this rather outlandish hypothetical that gets us to the crux of the issue: Will sex robots be more like vibrators, pets, partners, or slaves?
That question—and how technologists, potential customers, ethicists, and legislators will answer it—is mostly the concern of a few academics at this point. But in the not-too-distant future it will become much less hypothetical for billions of people. We are drawing ever closer to the era of realistic, affordable, emotionally intelligent robots, including sex robots. These have the potential to change not just how we relate to technology but how we relate to one another. The challenge: How can we make robots part of our social/sexual fabric without letting them remake us?
Meet the Sexbots
Contemporary commercial sex dolls can appear quite lifelike, but they're mostly non-robotic. The dolls, produced by companies such as California-based RealDoll and Japan's Orient Industry, tend to be made from silicone and a metal skeleton and weigh as much as 120 pounds. Depending on the company, dolls can be customized in a variety of ways, from hair and eye color to pubic hair style, plus the addition of features like artificial milk glands. Some offer simulated breathing, pulse, and heartbeat.
One of the few existing robotic sex dolls appears to be Roxxxy, from New Jersey-based TrueCompanion. With an appearance akin to an especially lifelike (yet not especially attractive) store mannequin, Roxxxy is in no danger of being mistaken for human. But she has three "inputs" (mouth, vagina, and anus), according to TrueCompanion's website, and the deluxe model boasts five programmable personalities, including Young Yoko, described on the company's website as "oh so young and waiting for you to teach her," and S&M Susan, "ready to provide your pain/pleasure fantasies." Roxxxy and her male counterpart, Rocky, are billed as responsive companions able to "listen, talk, carry on a conversation, and feel your touch." Owners can purportedly program them with likes, dislikes, and foreign languages, as well as upload their "personalities" to the cloud.
Roxxxy's renown has been wide since her debut at a 2010 adult-entertainment expo, garnering mentions everywhere from tech blogs to the BBC. But many in the love-doll community are skeptical that TrueCompanion has ever sold any robots (the company did not respond to requests for comment).
Davecat, 41, is one such person. A "Synthetik advocate," Davecat is part of a group known as the iDollators, who say they prefer sex dolls and robots to intimacy with "Organiks," a.k.a. human beings. Davecat lives with three dolls, whom he has named Sidore, Elena, and Muriel. He has made up personalities and created Twitter accounts for each of them.
Davecat was there for Roxxxy's debut, and he was not impressed. The product "fell far short of everyone's expectations," he says. "Robots by definition are capable of movement, which Roxxxy was incapable of." He and fellow iDollators found Roxxxy too heavy and visually unappealing, with "the guts of a laptop." And though advertised as the "first" sex doll responsive to stimulus, the Japanese doll company Axis Japan was already using the same sort of technology—sensors that trigger various MP3s to play when a doll is touched in different places.*
"Today's sex robot industry is underwhelming," says Aquino. "A new techno-sexual revolution is upon us," she explains, but it's currently focused on technologies like teledildonics and virtual reality, which are "converging to bring sexual fantasies to life while allowing users to participate in sexual activities safely and without risk of STDs."
The main thrust of "teledildonics" has been to combine things we conventionally think of as sex toys with haptic interfaces that allow users to "touch" and be touched remotely. Long-distance lovers, for instance, could use teledildonics to have robot-mediated sex, in combination with such technologies as shared virtual reality, webcams, or even old-fashioned phone calls. Users hooked up to virtual-reality headsets such as Oculus Rift could "participate" in porn or virtual erotic worlds. Simple teledildonics include things like Mojowijo, a set of paired vibrator attachments for the Wii, and OhMiBod, a vibrator that can be controlled remotely via an iPhone app. A website called Kiiroo allows teledildonics users to hook up with other users (known or unknown) from around the world, the ultimate fulfillment of the ancient promise of the AOL chatroom.
Meanwhile, those whose tastes are more technologically advanced must make do. Aquino says "a significant number of robosexuals, those who are attracted to robots, choose to partner with love dolls like RealDolls because they are limited by today's embryonic sex robot industry."
Count Davecat among that cohort. "All told, I'd rather have a Gynoid than a Doll," he says in an email, using the technical term for a female humanoid robot. "Dolls are fantastic, but realistically speaking, they can only do so much, and with a completely Synthetik lover, I'd have all the opportunities that are afforded in relationships with Organiks, but without all the drama."
The sex doll company Orient Industry announced in 2014 that it has developed skin "not distinguishable from the real thing." Sex robots could eventually be imbued with an almost real-time capability to "respond" to touch. Gerhard Fettweis, a professor of communications technology at Dresden University, believes that within 20 years wireless technology will match the speed of the human neural system. Some have proposed the idea of sexbots that mimic humans' biochemical signaling system, releasing pheromones corresponding to arousal and love at the appropriate times.
At the start of 2015, however, roboticists are still struggling with problems like making autonomous humanoid robots that can walk and move their faces realistically. Last summer, the National Museum of Emerging Science and Innovation in Tokyo debuted a girl and woman android, Kodomoroid and Otonaroid, to much fanfare. The robots are used to greet and read news to museum visitors and hold press conferences announcing new robots. They can make facial expressions and move their upper bodies, but they can't walk and can only lip-sync recorded speech. Convincingly human, emotionally intelligent androids of the kind seen in sci-fi are, for now, far more fantasy than reality.
How Much Is That Robot in the Window?
In a 2014 paper, the Brown University psychologist Bertram Malle and Matthias Scheutz, director of the Human-Robot Interaction Laboratory at Tufts University, defined social robots as "any robots that collaborate with, look after, or help humans." Kate Darling, a robot ethics researcher with the Massachusetts Institute of Technology (MIT), prefers the wordier "a physically embodied, autonomous agent that communicates and interacts with humans on an emotional level." Social robots, according to Darling, can also "follow social behavior patterns, have various 'states of mind,' and adapt to what they learn through their interactions." Sexbots, of course, would fall squarely in this category. So would robots designed to interact with nursing home patients and robot pets.
Early examples of social robo-pets include Furbies and Tamagotchi, which lived on tiny screens on key rings and alerted owners when they needed food or bathing. The Roomba, an autonomous robot vacuum cleaner that has sold millions since 2002, is considered a primitive social robot. Robotic puppies, seals, and other animals are now being tested to interact with nursing home residents and autistic children, with promising anecdotal results.
Human beings love their pets, in large part, because of our deep tendency toward anthropomorphism: the imputation of human-like qualities onto animals and nonliving things. Anthropomorphizing a pet doesn't require believing the pet is fundamentally human, it just means its personality and behavior inspires humans to treat it like a person with complex desires, motivations, or memories. It is a near certainty that we will do the same with social robots as they become increasingly commonplace.
The human inclination to anthropomorphize animals "translates remarkably well to autonomous robots," Darling noted in her 2012 paper, "Extending Legal Rights to Social Robots." A robot that can mimic human behavior, social gestures, and facial expressions "targets our involuntary biological responses."
In 2013 Julie Carpenter, a psychology researcher at the University of Washington, interviewed 23 U.S. soldiers working with bomb-disarming robots. While the troops defined the robots as technological tools, they were still given to naming them, gendering them, and talking about them with empathy. "They would say they were angry when a robot became disabled because it is an important tool, but then they would add 'poor little guy,' or they'd say they had a funeral for it," Carpenter explained in a statement about her work.
In a 2007 study from the University of California, San Diego, toddlers introduced to the humanoid robot QRIO quickly lost interest when the robot merely danced continually. But when dancing and giggling were triggered by their touch—when the robot was responsive in a human-like way—"that completely changed everything," study leader Javier Movellan said in a press release.
It is this illusion of agency that helps endear social bots to human beings. Social robots are designed to elicit anthropomorphic reactions. "There are many of us in the robotics community that study not just robots but human psychology," says Ron Arkin, an American roboticist and roboethicist who teaches at the Georgia Institute of Technology. To Arkin, the central question is: "Can we effectively design robots to interact with people in the way that people want to be interacted with? And that involves understanding the human mind as well as the robotic mind."
People bond with pets in part because we like things that seem to need us. This trait transcends flesh and blood. "Nurturing a machine that presents itself as dependent creates significant social attachments," wrote the MIT scholar Sherry Turkle in her 2006 paper "A Nascent Robotics Culture." Turkle found people are prone both to nurturing feelings toward autonomous robots and to believing, at least on some level, that robots reciprocate these feelings.
So is this something we should worry about? Projection onto traditional objects can be ignored and revived at will, noted Darling. But an artificially intelligent robot "that demands attention by playing off of our natural responses may cause a subconscious engagement that is less voluntary." Scientists and ethicists alike are exploring where to draw lines. Is it wrong, for instance, to "trick" dementia patients into caring for robo-pets?
Right now, social robots' potential benefits for everything from elder care to education seem to outweigh ethical concerns. But right now, intelligent and autonomous robots don't exist. In "The Inherent Dangers of Unidirectional Emotional Bonds Between Humans and Robots," Matthias Scheutz raises concerns that robot companions will have the ability to "exploit human innate emotional mechanisms that have evolved in the context of mutual reciprocity…which robots will (not have to) meet."
What are the potential repercussions of this? "Unfortunately, there is currently very little work aimed at trying to minimize the natural human tendency to anthropomorphize," Scheutz tells me. "The key question is how to walk the fine line between making robots useful to people without having them fall for robots."
Before Roomba and Roxxxy
While coverage of Roxxxy and her sisters tends to focus on the unprecedented nature of "lifesized robot girlfriends," creating convincing facsimiles of human beings in order to masturbate into them is actually an ancient pursuit. A Japanese anthology published in the late 1600s refers to Koshoku Tabimakura, a "traveling pillow," with an azumagata ("woman substitute") made from thin layers of tortoiseshell lined with velvet, silk, or leather. The dolls were also known as tahi-joro, or "traveling whores." In the 1904 book Les Detraques de Paris (which loosely translates as The Paris Crazies), Rene Schwaeble quotes "Dr. P," who sold "fornicatory dolls" (though he had to pretend to police he made balloon animals) for around 3,000 francs apiece in French catalogues. "Every one of them takes at least three months of my work!" said Dr. P. "There's the interior framework which is carefully articulated, there's the hair on the head, the body hair, the teeth, the nails! There's the skin, which has to be given a certain tint, certain contours, a particular pattern of veins….The only thing these haven't got is the power of speech!"
In 1908, the German doctor Iwan Bloch wrote of "hommes or dames de voyage," the "artificial imitations of the human body, or of individual parts of the body" sold in France with "genital organs represented in a manner true to nature." Dames were equipped with oil-filled pneumatic tubes, the hommes an apparatus by which "the ejaculation of the semen is imitated." By the 1920s, customizable sex dolls were advertised "fitted with a phonographic attachment, recording and speaking at will."
Though perhaps some were attracted to the dolls, these were largely considered masturbatory devices, or in some cases a tribute to a dead loved one. Will sex robots be similarly functional, or will they provoke desire in their own right?
"Right now, we're at an inflection point on the meaning of sexbot," Kyle Machulis, a systems engineer with Mozilla, told Aeon magazine last summer. "Tracing the history of the term will lead you to a fork: robots for sex (idealized version: Jude Law in the movie AI), and people that fetishize being robots (clockworks, etc.). There was a crossover in the days of alt.sex.fetish.robots, but I see less and less people fetishizing the media/aesthetics, and more talking about actually having sex with robots."
In a survey of 61 DollForum.com members—75 percent men who own dolls, 10 percent women who own dolls, and 15 percent men thinking about purchasing a doll—the psychology researcher Sarah Valverde asked owners what motivated their purchase. Not-mutually-exclusive answers included sexual stimulation (70 percent), companionship (30 percent), and using the doll in sex with a human partner (17 percent). About a third of male owners reported some issues with sexual functioning. Most rated their sex with dolls as "above average" to "excellent."
None of the respondents were in therapy related to their relationship with the sex dolls. Most were employed, educated, and reported similar anxiety and depression levels as the general population. While the use of sex dolls is often seen as pathological, Valverde makes the case that "a diagnosis of paraphilia would be unwarranted, without significant distress or impairment in functioning. Provided a doll-owner doesn't need the sex doll in order to achieve sexual satisfaction, a diagnosis of a fetish would not be appropriate" either. "Anecdotal evidence suggests these dolls have brought relief, security, and happiness to their owners," she concluded.
In his 2007 book Love + Sex with Robots, the artificial intelligence specialist David Levy—a former professional chess player and now president of the International Computer Games Association—pinpoints 11 major triggers that inspire emotions humans recognize as love. Many of these factors could presumably be inspired by social robots, including proximity, reciprocal liking (liking things that like us), need-fulfillment, a sense of mystery, and the presence of certain desired characteristics (like red hair or a deep voice). To Levy, it's not a stretch to imagine some humans falling in love with and even marrying robots within a few decades.
First, Do No Harm
One result of this influx of robots into our bedrooms is that it may "trigger a broader role for the concept of moral harm in law," suggests University of Washington law professor Ryan Calo in a 2014 paper, "Robotics and the Lessons of Cyberlaw." Certain uses of robots may be deemed undesirable because they compromise the actor, rather than a specific victim or society. The fact that a robot itself can't feel pain or be exploited may not stop pushes to prohibit particular uses of or behavior toward social robots.
"The Kantian philosophical argument for preventing cruelty to animals is that our actions towards non-humans reflect our morality—if we treat animals in inhumane ways, we become inhumane persons," noted Darling in her paper "Extending Legal Rights to Social Robots." "This logically extends to robot companions. Granting them protection may encourage us and our children to behave in a way that we generally regard as morally correct."
In her We Robot conference paper, Gutiu suggests that, "if regulated," we may be able to use sex robots "to correct violent and demeaning attitudes toward women." But this sort of large-scale social-engineering-through-sexbot could quash the potential for their more individualized use in rehabilitation.
Levy imagines a role for sex robots similar to sex surrogates, therapists who use actual sexual intimacy to address clients' issues. "All of the most common sexual dysfunctions and their cases can be treated by surrogate-partner therapy, including premature ejaculation, nonconsummation of a relationship, erection difficulties, performance anxiety, and fear of intimacy," he explains in Sex + Love With Robots. The book cites the California sex therapist Barbara Roberts, who laments that "we have no traditional rite of passage nor meaningful ceremonies to initiate young people into informed adult sexuality"—a role Levy also envisions for sexbots.
And then there's the inevitable question of kiddie sexbots.
Last summer, at a Berkeley Law School panel on ethical and legal challenges in robotics, Arkin spawned a flurry of sensational headlines by suggesting that "childlike robots could be used for pedophiles the way methadone is used to treat drug addicts," potentially reducing recidivism rates for sex offenders. Many people find this idea immediately distasteful. Arkin empathizes with them, he tells me, but he thinks it's better to investigate the therapeutic potential of such robots "in a controlled way" rather than simply avoiding research because it makes us squeamish. While no U.S. companies are publicly selling them, childlike sex dolls are already available online from foreign makers.
In Canada, child sex robots are illegal, but there are no U.S. laws yet specifically criminalizing them. In fact, there is reason to think the U.S. courts might carve out some legal space for them, as unlikely as that might seem: In 2002, the U.S. Supreme Court struck down parts of a federal law criminalizing "virtual" child pornography, described as either digitally created images or those featuring young-looking adults pretending to be younger. "I could see that extending to embodied [robotic] children," said Calo, the law professor, at the panel,"but I can also see courts and regulators getting really upset about that."
Regulatory concerns notwithstanding, "it's coming to the time when we start talking about these things," Arkin argues. "Should the design of [sex robots] be informed by science? Yes. Is anyone doing true scientific study on intimate robots at this time? Not to my knowledge. I would encourage that line of research to be undertaken if we can get past our Victorian taboos."
One area where academics and journalists seem enthusiastic about the possibilities for sexbots concerns robot prostitution. Love doll brothels can already be found in Japan. In a 2012 paper, "Robots, Men and Sex Tourism," the New Zealand researchers Ian Yeoman and Michelle Mars enthusiastically predict that robot prostitutes will overtake human sex workers by 2050. Yeoman and Mars paint an elaborate portrait of a posh Amsterdam robot brothel catering to a high-end clientele and niche sexual preferences-a situation the writers see largely as a social good, capable of invalidating all the messy moral concerns that human sex workers present.
Prostitution is illegal in the U.S. and many other countries, and various nations have previously criminalized everything from vibrators to adultery, so lawmakers may well move to block robot brothels also. But should robot prostitution be legalized, would "the oldest profession" find itself, like so many others, vulnerable to technological disruption?
In his 2014 paper "Sex Work, Technological Unemployment, and the Basic Income Guarantee," John Danaher, a law lecturer with The National University of Ireland, Galway, rejected the idea that sex workers and clients will all go quietly into the good robot night. This is largely due to the fact that people like having sex with other people; even in the presence of a robust robot sex trade, those inclined to pay for sex will still sometimes want to do so with a human being. But we also shouldn't discount sex-worker resiliency—like the move from streetwalking to advertising on Backpage, those in the sex trade will adjust to suit the times. "Prostitution could well be one of the few forms of human labour that is likely to remain resilient in the face of technological unemployment," posits Danaher.
Research on why men pay for sex has found, more than any other common denominator (variety, convenience, etc.), a desire for mutuality. Clients want to feel, at minimum, like a sex worker somewhat enjoys her time with them. In a 1997 study of male prostitution clients ages 27 to 52—nearly half of whom were married—a desire for sex was frequently met with "social, courting behaviors that were often flavored with varying degrees of romance." Interviewing clients at a New Zealand massage parlor, researcher Elizabeth Plumridge found they "all wanted a responsive embodied woman to have sex with. This they secured by ascribing desires, response and sexuality to prostitute women. They did not know the true 'selves' of these women, but constructed them strategically in a way that forwarded their own pleasures."
Read one way, this research could support the future popularity of robot prostitutes, which could theoretically be programmed to portray care and lust sufficiently well that we fall for it. This, of course, depends in part on how effectively artificial emotional intelligence and sociability is developed. But even if we grant that realistically emo sexbots are possible, will they be "real" enough to afford mutuality? Whether we're talking orgasms or affection, convincing oneself that a human sex worker isn't faking it rests on the fact that, technically, she may not be. With robot companions, the fakery is inherent. It's a given. How much that actually matters remains to be seen.
Giving sex toys and sex dolls the illusion of agency will attract new users, Arkin suspects. "Not for everybody—it may go from one-tenth of a percent to 1 percent—but it would grow the demographic exponentially."
In a June 2014 YouGov poll, Americans were split on whether using a sexbot is moral. Forty-three percent of those surveyed said using sex robots is wrong, and 39 percent said it's acceptable. Only 10 percent said they would use a sex robot themselves.
Should sexbot use reach the mainstream, couples will have to wrestle with questions like how to handle jealousy over robot companions and whether robot sex counts as cheating. Is having sex with a robot more like using a vibrator or having a fling? Is it uncouth for friends to share a sexbot? What if someone creates a sexbot in your likeness?
Sex robots also present ethical issues for academics. "From a researcher's point of view, what is appropriate?" asks Arkin. "There are no guidelines for researchers in this particular space."
The goal of many roboticists is to get to a point where robots can successfully manipulate our emotions. To make robots more like sociopaths, able to recognize and use social cues, create an illusion of empathy, and gain trust and intimacy without reciprocity.
Osaka University's Hiroshi Ishiguro, who supervised last summer's "Android: What Is Human?" exhibition in Tokyo, has said that "the process of understanding (human) nature is the most interesting part of androids." And for some, a faith in a quintessential humanness—something even the most sophisticated and intelligent robots can't approximate—is one way to mitigate worry over the future of social robots. If human beings bond with robots not for what they are but what they inspire in us, perhaps our insurance lies in what they can't inspire: a sense of mutuality, reciprocity, and genuine agency. To paraphrase David Levy, people don't fall in love with an algorithm but a convincing simulation of a human being. Yet can any simulation really be convincing enough? Enough to have mass appeal? Enough to significantly change the social fabric?
Near the end of Sex + Love with Robots—a techno-utopian volume if there ever was one—Levy writes that he does not believe for one moment that sex between two people will become outmoded. "What I am convinced of," he declares, "is that robot sex will become the only sexual outlet for a few sectors of the population—the misfits, the very shy, the very sexually inadequate and uneducable—and that for some other sectors of the population robot sex will vary between something to be indulged in occasionally…to an activity that supplements one's regular sex life."
On the margins, sexbots could dissuade some individuals from pursuing human-to-human intimacy and relationships, just as pornography, sex toys, and everything from alcohol to work are also sometimes used to avoid attachments. But it has become clear through countless bouts of cultural and technological change that, for the most part, people see no substitute for knowing and loving another person. To predict sexbots as even moderately widespread stand-ins for sex and relationships reveals a not-insignificant misanthropism.
That isn't to say that individual use of sex robots is misanthropic. For many men and women, they will remain ancillary to interhuman relationships, more like sex toys than humanity surrogates. For a subset, social robots may provide opportunities for companionship and sexual satisfaction that otherwise wouldn't exist. When this occurs, we'd all do well to remember that having faith in human institutions and relationships means not panicking over new possibilities. Staying conscientious but open-mined toward the use of social robots, including sex robots, can only enhance our understanding of what it means to be—and to fall for—human beings.
* Editor's Note: A previous version of this article included an additional paragraph regarding TrueCompanion. It has since been deleted. Some minor word-changes have been made as a result.