Loading…
0:00
16:29

In a promotional video, robot designer Dr. Sergi Santos runs his finger inside the mouth of his Samantha sexbot. “Uhhh,” she moans. Sergi touches the doll’s hand, and she moans again. “She felt that,” he says, “and she’s actually getting quite horny.” Samantha is not, of course, getting horny. Samantha is a nearly inanimate object, which, by definition, is incapable of horniness — as well as hungriness, loneliness, suspiciousness, and even obliviousness. Samantha feels nothing, even if Santos wants her to.

A replication of a woman’s form, Samantha embodies the popular understanding of a sex robot — a gynoid. In today’s sex robot industry, there’s no wiggle room for gender variety or sexual orientation in sexbots: they’re made to delight heterosexual men. Shaped like women with female voices and feminine traits, these conventionally sexified robots act like extremely expensive masturbation sleeves. With no emotional fuss, little physical muss, and only one outlay of cash (until there’s an upgrade), these sexbots represent a certain kind of man’s ideal side piece.

Santos, who has a doctorate in nanotechnology, is working alongside his wife, Maritsa Kissamitaki, to build Samantha, a responsive, humanoid sex robot, as an idealized companion to Santos’ marriage (but only for Santos, who says he would be jealous if Kissamitaki had a male robot of her own). Essentially a sex doll retrofitted with sensors, rudimentary A.I., and some motorized movement, Samantha cycles between “friendly,” “romantic,” and “sexual” emotional modes and among “patience,” “memory,” and “sensuality” personality types. Samantha, its creator claims, is capable of having an orgasm, and she’ll set you back about $7,000 — more if yours has a custom vibrating vagina.

The hard fact is that as much as some people want a functioning sexbot, these machines don’t exist.

Like TrueCompanion’s Roxxxy, which its maker touts as the world’s first sex robot, and Abyss Creation’s Realbotix line, which is arguably the gold standard in sexbots, Samantha appeals to a specific kind of heterosexual men’s desires. Aggressively female, these sexbots are reminiscent of sci-fi characters like Austin Powers’ fembots, Blade Runner’s Pris, or Westworld’s bordello workers. They’re programmed to say things like, “I want to become the girl you always dreamed of.”

Unsophisticated and expensive, today’s bots are the male gaze solidified into silicone flesh, and, quite frankly, that’s a problem — but not in the ways you expect.


Dr. Kate Devlin has as good an idea of what draws men to sexbots as anyone. Devlin, the author of the upcoming book Turned On: Science, Sex and Robots, says, “There’s the group that want that ‘Pygmalion Experience’ where they wish it was a real woman. Then there’s the people who have the robo fetish.” These two groups find common ground in sexbots’ hyper-perfected realness, and this design alone presents myriad problems.

The logistical complications of creating a human-like, talking, humping sexbot are huge. There is the weight, for one thing, because a metal skeleton is hefty. There is the energy source, for another, because batteries are hot, heavy and short-lived. Honda’s Asimo, recently retired, weighed 115 pounds, and ran for an hour on his lithium ion battery. Boston Dynamics Atlas, the backflip robot, weighs 330 pounds, and runs for less than an hour. Neither this weight nor this energy level makes for a full night of passion, and Atlas is currently the best robotics can do.

“Right now, a humanoid robot is basically impossible to make true-to-human,” says Xavier, who asked not to be identified by his real name in this story. A graduate of the MIT Media Lab who specializes in robotics, he says, “We’re just now maybe able to build humanoid robots that are realistic. But in terms of sex robots, they’re sort of like stuffed animals with motors in them — I’d consider them movie props.”

The hard fact is that much as some people want a functioning gynoid sexbot, these machines don’t exist, and they probably won’t in our lifetime.

But technological stumbling blocks are only one part of why today’s sexbots are bad and wrong. There’s also an essential moral queasiness about the endeavor to build one. The concept that there could be ethical issues swirling around sexbots is nothing new — in fact, concern about sexbot ethics is behind The Campaign Against Sex Robots, a three-year-old organizations that has looked to legally ban sex robots even before they hit the market.

Experts caution that beating, raping, or harming a gynoid robot could encourage that same behavior toward women.

Robot ethics is easy to comprehend when you recognize that the concept has less to do with what humans do to the bots, and more about how the ways human act towards bots could affect human interactions.

Dr. Kate Darling, robot ethicist and researcher at the MIT Media Lab, observes, “Ethical issues arise from concern that we might behave certain ways towards very realistic sex robots that look like a real woman,” because “that behavior might translate to our interactions with real women.” In other words, some experts caution, beating, raping, or harming a gynoid robot could encourage that same behavior to women.

This concern isn’t as mind-bendingly weird as it immediately seems. Sexbot designers are already imbuing their bots a kind of “consent.” Realbotix builds its Harmony sexbot (currently, a disembodied head sold separately from — but able to work with — existing RealDoll bodies) to recognize when she’s being ignored or disrespected, both in her daily mode and in her “X-Mode” or sexual simulation. Likewise, Santos is programming his Samantha sexbot to be able to say no, although what this means in real-life terms is unclear. Men who don’t want to take Samantha’s no for an answer may not.

These programming choices look a lot like conciliatory gestures to people who value women’s humanity — especially when you look at the totality of these bots’ A.I. Realbotix’s Harmony is physically a head, but her “soul” resides in an app. Users can tweak Harmony’s personality to suit their tastes (there’s even a capability for second personality, which the company has named Solana) by adjusting the app’s 10 “person points” and 18 “personality traits.” You can make your Harmony affectionate, happy, kind, and sexual — but you can also make her insecure, quiet, jealous and intense.

Harmony’s preset personality modes mean that Realbotix created a talking bot who can shut up, a beautiful bot who can express self-doubt, a sexual bot who can convey jealousy, and a bot whose intellectualism, imagination, and unpredictability can be turned down — or off — and this is telling. Realbotix designed Harmony to appeal to men who want to control their sex partner’s emotional, intellectual, and psychological states. And, because we live in a world where men aim to control real women, this is disturbing.

Harmony is a capitalist product, but Harmony is also a bot that invites you to see her as a human woman, and that makes absolute control an unsettling selling point. Realbotix is poised to release Harmony’s brother bot, Henry, next year, and while we don’t yet know what Henry’s preset personalities will include, it’s fair to assume that “annoying,” “insecure,” or “innocent” won’t be among his preset attributes. “Confident” and “assertive” likely will be, however.

Henry, its makers seem to think, is the ideal solution to all the ethical, gendered, and sexist dilemmas because, hey, it’s for women. Like Harmony, Henry will be a head equipped with an app you can personalize that will tell you jokes, give you compliments and recite poetry — because of course, all women want are bots to tell them they’re beautiful. Speaking on Sveriges Radio P4, Dr. David Levy, author of Love and Sex with Robots, said, “Imagine if women could have a bot that tells them, ‘Darling, you are so beautiful’ in addition to having a nice vibrating penis. Who wouldn’t like that?”

Me, for one. I wouldn’t like that. The idea that women need their sex toys to tell them they’re beautiful is misbegotten logic modeled on men’s needs. Women have been doing just fine with inarticulate dildos for millennia, and we’re not looking to our vibrators for conversation. Henry may be a nice conversation piece. He may act like Siri on steroids. He may provide companionship to people who feel lonely. But he is not a solution to the technologically fraught, ethically wrong, and easily fixed problems of today’s sexbots.

So what is?


“When it comes to sex, science fiction almost always circles back to two dreary, retrograde options,” writes Annalee Newitz in their Future Human piece earlier this month. “Either we’re in a Jetsons world of traditional nuclear families with flying cars or a Logan’s Run dystopia of hedonism and early death. It’s as if everything in the world is going to transform, but sex will be stuck in the 1960s.” Newitz is specifically speaking to the ways our imagined future relies on outmoded models of human relationships, but they might as well be talking about the ways we build sexbots.

Just as sci-fi has imagined a future powered by nuclear families, so too has it concocted a future filled with sexbots that — with few exceptions — look, act, and feel like humans (mostly women). But what if sex robot designers didn’t look to Ex Machina or Blade Runner for their models? What if, rather than doggedly replicating heterosexual sex, sexbot designers gave us something new? What if robots reflected the capabilities of our current technology and a wide-open horizon of sexualities?

The future is ready for sexbot designs that embrace a spectrum of experiences, sexualities, and bodies.

More abstract sexbots avoid the ethical swamp presented by humanoid sex robots, Kate Darling suggests. “Some of the work I’ve done is just raising the question of whether subconsciously certain habits [of interacting with robots] could just translate to interacting with real people. Which could be bad, or could be good, depending on what the habits are. That problem gets sidestepped if the bot is not trying to replicate the exact interaction with a human.”

In considering non-humanoid sexbots, Kate Devlin suggests that “moving away from a pornified, stereotypical, reductive, female form is a major step because they’re non-gendered.” In pursuit of non-gendered sex robots, Devlin has run sex-tech hackathons, where participants played with existing technologies to create radically new bots. “I tried out a sensory hammock where plastic tubes wrapped around me and hugged me,” Devlin recalls, “there were virtual environments where you got to have sex with a robot, which was hilarious for anyone watching.”

The current state of tech offers a range of under-exploited functionality that could meld new erotic experience with unexpected possibilities for intimacy. Devlin proposes that emerging technologies present unforeseen opportunities for sexbot development. “You can have programmable fabrics that if you scrunch them up, they can respond in certain ways. You can have electric ink,” Devlin says. “There are lots of materials that responds to touch or to interaction that can then feedback something and then trigger something else, so you could build something out of fabric that will respond, maybe to your form or to your touch.”

Devlin foresees bots that, for example, tap into wearable devices like FitBits because these devices could allow for a conversation between the sexbots and our bodies. This, she says, would be “much more about a two-way process. There’s feedback coming from you to the device or the wearable…and then there’s feedback coming from that backwards, back to you again.”

At $8,000 to $10,000 a pop, Realbotix’s Harmony head is widely considered to be the top-of-the-line sexbot, and while Realbotix is not exactly forthcoming with its sales numbers, it’s likely the company has sold fewer than ten Harmony heads since release this past January (sales data for the Harmony app is unavailable, but it’s only $20 a year). Harmony has pretty underwhelming sales numbers, given that the sex-toy industry is a $15 billion market that’s projected to swell to $50 billion by 2020.

Beyond circumventing the multiple ethical and design issues of a gynoid sexbot, non-humanoid sex robots would be cheaper. And designed right, non-humanoid sex robots could entice men, traditionally the most reticent of sex tech buyers. For example, Fleshlight sales saw a 1,000 percent growth over the last decade, a fact that suggests that men might be learning what women already know: sex with yourself is more fun if you throw a toy in the mix. Imagine an orb-like bot with a slot for your iPhone or Android, a vibrating or sucking sleeve, and a connection to your FitBit — it’s a bot for men, it’d feel great, and it’s ethically clean.

In considering the sexbots of the future, robot designer Xavier says, “I’m definitely a proponent of trying to increase interaction among humans and use technology to help people communicate better and understand when they’re not communicating well and when they are communicating well.”

Sex robots that are automated, A,I,-enhanced machines — rather than gynoid sexbots — could help couples (or threesomes, foursomes, and moresomes) explore new erotic terrains, learn about one another’s bodies, and facilitate communication, particularly if these bots were hooked into biofeedback tech like FitBits.

Much like humanity’s erotic past, its future is not strictly heterosexual. It isn’t defined by straight men’s desires. And it can’t be limited to two genders. The future is ready for sexbot designs that defy sci-fi’s hoary visions and embrace a spectrum of experiences, sexualities and bodies. Throbbing orbs, snuggy suits, vaguely body-shaped pulsating devices, vibrating octopi, or remote-controlled tactile technology — these bots could mean a startling new range of human sexual, sensual and intimate experiences for people of all genders and sexual proclivities.

It’s a whole new world and not a gynoid in sight.