About 9% Of You Would Have Sex With A Robot

Statistically speaking, nearly a tenth of people would–if given the chance–do it with a sexbot. But the most interesting thing about those future erotic engineering marvels aren’t the physical acts they’ll be able to perform, but rather what they reveal about a human’s concept of sexual relationships.

About 9% Of You Would Have Sex With A Robot

Since I’ve started tracking the story of sexual computing I’ve received many emails and countless tweets stating that while developers and engineers may be working on sexbots and other sexual technologies, no “normal” person would ever use such tech in their sex life.


But now thanks to a few recent surveys we know that’s just not true. Here’s the first, as Alexis Kleinman writes for The Huffington Post:

Nearly 20 percent of young adult smartphone owners in the U.S. between the ages of 18 and 34 use their smartphones during sex, and nearly 1 in ten U.S. adults who own smartphones use them during sex.

While this survey was conducted with a fairly large sampling size of 1,100 people, it did not specifically ask respondents what they were using their smartphones for while having sex. Theoretically the smartphone use could have been for anything from checking texts, to taking pictures and recording videos, to referencing Kama Sutra guides. Regardless, it does demonstrate a willingness among people–especially youth–to allow technology into their most intimate moments.

And it’s these people who are the reason why “smart” vibrators like the Vibease, which sets its speed according to cues from spoken erotic e-books, are finding an increasing marketplace among a population that is growing ever more comfortable–even intimate–with technology.


And while checking your phone–or even using it to enhance your technique–in the sack is one thing, surely no one but a pervert would ever sleep with a real sexbot, right?

Wrong. A YouGov/Huffington Post survey of 1,000 US adults found that a whopping 9% of them admitted they would have sex with a robot if they could. That’s almost one out of every 10 people–and those are just the ones who would admit to it. And there’s good news out there for the 9% who would like to: It will happen. When I asked the adult app store MiKandi’s CEO Jesse Adams to postulate the not too distant future of sex and tech, he had this to say:

There will be sex toy like robots that can react, listen and give you the exact pleasure you want to enhance these virtual experiences. These devices will understand and measure your arousal levels, body fluids, your breathing patterns, your body movements. They will also learn and adapt and react to your feedback.

And it will be those sexbots that can quantify and understand our needs and desires that will raise the most interesting human quandary of sex with tech: not mechanics, but ethics. Specifically, does sex with a robot that can understand our physical needs constitute cheating?


This is where it gets interesting, because while most people would say the use of a smart vibrator (which is essentially a “dumb” robot) would not constitute cheating, the YouGov/Huffington Post survey revealed 42% of respondents said sex with a robot would be cheating (while 31% said it wouldn’t, and 26% were not sure).

To put that another way:

Sex with a vibrator = not cheating.


Sex with a vibrator that has legs and eyes and a face = cheating.

Ironically, it’s the possibility of sex with inhuman robots that reveals something very human about our concept of what sex is. To humans sex is more than mechanics and pleasure; it’s emotion and connection, which are primarily conveyed through human-only traits, like eye contact, empathy, and a partner’s careful observation. But one day machines will be able to convey those traits, and when that happens is when the real debate over sex and technology begins.

Previous Updates

Does The “Lolita Bot” Help Catch Online Predators Or Create More Of Them?

July 16, 2013


A group of researchers in Spain have successfully coded software that mimics the language and attitude of a 14-year-old girl. The software, officially called Negobot–but now colloquially referred to as a “Virtual Lolita”–appears to be the first AI that accurately mimics an adolescent. As Jillian Scharr writes for TechNewsDaily:

The bot — which can speak multiple languages thanks to translation technology — is also programmed to act in a manner that could be considered vulnerable, trusting and naïve. But what makes Negobot unique — aside from its crime-fighting mission — is its use of game theory to trap potential pedophiles. That means, essentially, that Negobot treats conversations as a game, with the objective of gathering as much potential evidence of pedophilic tendencies as possible.

And it’s the use of “game theory” that makes Negobot so impressive–and so much more than just an app. Negobot starts off in a “neutral” mood. It will chat to strangers in a chat room about ordinary subjects, but once a stranger starts speaking in innuendo or overt sexual overtones, Negobot goes into game mode, tagging the chatter as “possibly pedophile.” Here it will start “revealing” personal details any 14-year-old girl might bring up: problems with siblings or school, life at home, boys. If the chatter continues along sexual lines of conversation, Negobot tags that chatter as “allegedly pedophile” and goes into full-on gaming mode. Now Negobot will try to titillate that chatter to keep them talking as long as possible–and hopefully get them to reveal personal details about themselves and agree to a meetup.

Negobot is a clever, indeed ingenious, use of software. The skill at which the programmers have enabled it to accurately mimic a 14-year-old girl approaches a work of art. Negobot can be shy, coy, and insecure. It knows the latest adolescent slang and pop references. It even misspells words on purpose and uses text speak (“u” instead of “you”) while chatting. It drifts from enthusiasm to boredom and back.


The only problem is: It’s a terrible thing to have built. Before anyone accuses me of being soft on pedophiles, let me state that I’ve interviewed many sex trafficking victims and learned more of the horrors perpetrated on them than I care to remember. My research led me to write a book about it and only strengthened my view that sexual violence is perhaps more pernicious than any other type of abuse.

However, using brilliant AI software–especially one that uses game theory–to lure potential predators is wrong. It’s entrapment. Don’t agree? If Negobot tags a chatter as “possibly pedophile” and then that chatter tries to leave the conversation, Negobot then bumps its game mode to the next level and tries to “win” at any cost. Negobot’s 14-year-old personality becomes more suggestive, more like the fantasy that every pedophile has in their head–hence the Lolita reference in my headline.

At that point, Negobot becomes the predator.


And I’m not the only one who thinks Negobot shouldn’t be used for its designed purpose. As the BBC reports:

John Carr, a UK government adviser on child protection, welcomed any move to relieve the burden on real-world policing. But he warned the software risked enticing people to do things they otherwise would not.

“Undercover operations are extremely resource-intensive and delicate things to do. It’s absolutely vital that you don’t cross a line into entrapment which will foil any potential prosecution”, he said.

That’s not to say I don’t think Negobot shouldn’t be used at all. There’s a perfect use for it in sexual technology because it demonstrates that a lifelike erotic AI is feasible.

If there are realistic sexbots in the future, they’re going to need not only a good AI that mimics human behavior, but also an AI that can recognize sexual innuendo. Negobot does that perfectly. If it can detect when a chatter says something that has a sexual connotation, then it will have no problem mimicking the first stage of human foreplay: verbal innuendo.


And lets not forget that a walking, talking sexbot of the future wouldn’t just be good for those with more eclectic sexual tastes, but such robots could also be non-sexual companions to the old, lonely, or physically handicapped–especially ones that have an AI that is so sensitive it can detect and reply in innuendo.

Negobot is a brilliant example of the merging of sex and tech. It was just designed with the wrong goal in mind.

Why We’re Tracking The Ever-Closer World of Sexual Computing

Sex is kind of a big deal for us humans. It drives a lot of what we do, and influences a large part of our world outside of the bedroom. It’s why some men seek high-paying, high-status jobs. It’s why some religions feel they need to issue moral laws. And whether we admit it or not, it can be a big part of our recipe for self-worth.


Sex has been synced with technology for hundreds, if not thousands, of years. If you’ve ever been in the Sex Machines Museum in Prague, you’ll know what I’m talking about. The museum holds some interesting mechanical (some even steam-powered) pieces of kit that were designed to enhance a user’s sexual experiences. When the technology of moving pictures was invented, pornos were some of the first things shot. Matter of fact, the porn industry arose hand-in-hand with the tech industry. Porn companies were some of the first to adopt the new technologies of VHS, DVD, and streaming video. And how long did it take porn companies to jump on Google Glass? About two minutes.

But sex and tech increasingly aren’t only linked by new mediums to show skin flicks. Outside of the medical industry, the most common places you’ll find research into synthetic skins are at sex toy companies. And while many extrapolate their fear of today’s drones and see them morphing into Terminator-like killing robots in the not-too-distant future, if history has shown us anything, we’re much more likely to see fully functioning sexbots first. Sound crazy? It’s not. After all, a robot getting pounded in the bedroom needs to absorb a lot less stress than one getting pounded in the battlefield.

And mark my words: One day we will be fucking robots.


I’m not trying to get a cheap laugh. It will happen. I know this because some very bright engineers across the globe are already working on technologies that will enable future sexbots to exist: synthetic skins with breathable pores that are capable of sweating; artificial intelligence that understands emotional context; human-accurate voice synthesis and facial recognition. And those are the hard parts. Believe it or not, building a robot that can mimic human movements isn’t that far off.

Before anyone jumps the gun and says that sexbots will lead to the moral decay of humanity and kill all human interaction, let’s not forget that sex with robots isn’t necessarily just for kinky cyber fun. It could have mental and social health benefits–like by allowing disabled people to have physical relationships or by eliminating the risk of sexually transmitted diseases from popular sex tourism areas across the globe, such as Las Vegas or Thailand.

If you’re interested in the ever-closer world of sexual computing, be sure to follow this tracker. Here we’ll explore the latest hardware and software advances that will one day change sex forever. Whether that’s a good or a bad thing, we’ll discover as we go. And if you’re in the business of merging sex and tech, get in touch with the author @michaelgrothaus to let him know what you’re up to.


Apparently, Wearable Fitness Devices Can Reveal If She’s Faking Orgasm

July 12, 2013

Clip-on fitness trackers and bracelets are already becoming the norm among the tech savvy, and soon eyewear like Google Glass will come to consumers, followed by (and perhaps swallowed by) the first true smartwatches–led by Apple, or course–and then consumable nano-devices to inspect your insides. Always-on “quantified self” devices can teach us a lot the world around us, and the world inside us. And when it comes to sex, the world inside someone else.

Gregory Ferenstein over at TechCrunch has written an interesting piece about how current technology–his BodyMedia armband–can actually alert his significant others to whether he is cheating on them or not. He discovered that his various sporting activities, including cycling, weight lifting, yoga, and more, actually produce different line graphs depending on the exercise. That is, each type of physical exertion has a unique signature (a series of spikes on the graph)–including sex.


See, Ferenstein wears his BodyMedia armband 24 hours a day, and when he gave his BodyMedia armband data to a friend, his friend–to Ferenstein’s surprise–was able to tell when he had sex. As Ferenstein writes:

Indeed, I inadvertently discovered that people knew whether I was engaging in sexual congress after I gave my health tracker data to a friend and he wryly quibbed about my night time activities […] Were I married, my wife might like to know why I burned 100 calories between 1:07 to 2:00 am, without taking a single step, and fell asleep right afterwards. Many married couples hold joint online accounts for Facebook and email, and even more share their passwords. Anyone looking at my exercise readout that night would instantly know that I was getting a sweaty workout.

Skeptical readers may claim that not all infidelity happens at night, and that clever philanderers could simply claim that they were hitting the gym, when they were actually knee-deep in sin hotel. But, as we’ve seen above, sex looks quite different than weight-lifting. In fact, the profile of sex looks distinct from any exercise I’ve recorded myself doing, including weight lifting, sprinting, yoga, martial arts (capoeira), TRX, spin class (stationary cycling), grocery shopping, and cleaning the house.

Now, as Ferenstein notes, not all people wear their health trackers 24 hours a day–and it’s easy to remove them whenever you want. But it is an interesting (and slightly frightening) discovery that given access to your fitness data, a third party can glean information about your sex life from a health tracker meant to track your fitness.

That got me thinking: Sharing is a major component of these devices. Many fitness trackers suggest their users do so as to keep themselves motivated. Published on the web or even left lying on the screen of an iPad, records of all your body movements and outputs could expose a cheating spouse, or even let parents know when their teenagers begin having sex. One might argue that tech used in this way is a cheater’s comeuppance or a useful tool for parenting, but I would argue that in either case it is a gross invasion of privacy.

But more than just telling if someone is having sex, Ferenstein also noted that he could tell from his data from another piece of fitness tech he owns (a Basis watch) if the person he was having sex with was faking orgasm. This is due to the monitoring of perspiration output and heart rate–things that spike during a real orgasm, but not during a fake one (in this case, Ferenstein’s partner would need to be wearing the watch).

From a data science perspective, this is fascinating (How many women are faking it?), but from a perspective that I hold–that everyone is entitled to their most intimate privacies–this is frightening. If a man found out his wife is faking it, it could lead to hurt feelings, guilt, confusion, and more. I doubt a woman would appreciate it from her side either.

And keep in mind: This sexual information can currently be gleaned from what will look like archaic tech in a few year’s time. Once wearable tech moves on to true smartwatches with more advanced biosensors, and then to stick-on patches that we apply to our skin and fabrics that we can wear as shirts (or even underwear), it is very possible that our most intimate sexual moments will become social. Is this a good thing?

No. This is the only time I can honestly say I don’t think tech will help enhance our sex lives. There’s too much room for invasion of privacy; it’s too likely that people will get hurt. Imagine a young gay teenager using a fitness tracker. His parent’s who don’t know he’s gay knows he’s hanging out with a male friend. They see his fitness data spike, yet it registers he isn’t taking any physical steps, which suggests he’s lying in bed. His tech could inadvertently out him before he is comfortable with it.

But just because our sex lives might become more accessible to third parties in the future isn’t a case against wearable computing. There will be a lot of good that comes from it (healthier lives; better monitoring of medical conditions; easier ways to find lost children). But when the Age of Wearable Computing does hit us, it will become more important than ever for users to be vigilant over their privacy settings or else we risk our sex becoming social.

The Forerunners Of Future Sexbots, Now

June 19,2013

To get to the stage where people can interact with human-like robotic companions, whatever their function, will require overcoming three main challenges: building an AI that is intelligent enough to not only speak like a human, but to process a human’s complex speech patterns to understand emotional context; building synthetic organs that look and feel indistinguishable from a human’s; and building an underlying mechanical skeleton structure that allows the robot to move like a human.

That last challenge is already well on its way to being overcome thanks to a group of researchers at the University of Tokyo who have built a robot called Kenshiro. The amazing thing about Kenshiro is that its movements don’t rely on cogs or pistons, but on a pulley-like system that mimics how actual muscles in the human body works (human muscles only pull, they don’t push). Kenshiro has over 160 “muscles.” That’s a far cry from the 640+ in a human body, but just take a look at what researchers in 2013 can do with just 160:

But what happens once you get the mechanics down? A robot, even one that moves and looks like us, won’t feel real until it can give the illusion of understand our own — and expressing its own — emotional complexities. The most advanced “AI” most of us have access to nowadays comes in the form of iOS’s Siri — and we all know how well that works. But even if true AI takes decades longer to achieve than human-like movement, that doesn’t mean we won’t have lifelike robots before then, as Japanese roboticist Hiroshi Ishiguro successfully showed off this week.

In a show in New York, Ishiguro took the stage next to a lifelike robot version of himself — called a “Geminoid.” The robot moved and spoke like the real Ishiguro, but that’s because it was “tele-operated” by a colleague offstage. The robot had to be remote controlled, of course, because of one big drawback — there’s currently no AI that could make it act and work like a human. However, this “drawback” also lends itself to demonstrating another benefit of pre-AI robots: if a human can control lifelike robots then we can push these surrogates out into the world to handle jobs that are too dangerous for flesh and blood.

However, given all the robotics advances in just the last year alone, we still are a ways off from having robotic companions. As David J. Hill writes for SingularityHub:

In the years to come, news of robotics development will only increase as we watch researchers put humans together much like Doctor Manhattan. Whether individual bones and muscles will work best or some other design will prove superior will only be resolved once each design can be tested. In the end, the greatest challenge may not be creating individual robotics systems or artificial intelligence, but packing all of these components into a single robot.

It is possible one day, however, to think that a much older version of yourself will be explaining to your live-in robot that their lineage started way back in the day with the creation of Kenshiro and the Ishiguro Geminoid.

[Image: Flickr user Sam Howzit]