More than two-thirds of men recruited for a sexbot study say they would give sex robots try. About two-thirds of women in the same study say they would not try a sex robot. Those findings come from the first exploratory survey of human attitudes toward sex robots. Such research has huge implications beyond whether humanity ends up using robots for sexual satisfaction—it can also reveal gender differences in how people view modern human relationships.
Debates about sex robots typically focus on either the crude robotic sex toys of today or Hollywood’s science fiction fantasies such as “A.I.” or “Ex Machina.” One U.K. researcher made headlines by calling for a ban on sex robot technology. But there has been surprisingly little effort to find out what people think about robots and sex in the real world. That is why researchers at Tufts University in Medford, Massachusetts set out to discover out what people think sex robots should look like and what uses of such robots would be considered appropriate. They hope their research can help create a future where robots prove more beneficial than harmful for human psychology and relationships.
“I think it’s very important to realize that sex robots and companion robots are all instances of social robots that have an effect on people,” says Matthias Scheutz, a computer scientist at Tufts University. “Especially when it comes to the potential of these machines to cause emotional harm to humans.”
Previous attempts to poll public opinion on sexbots have usually asked just several basic questions about whether or not people would have sex with a robot. Scheutz and Thomas Arnold, a research associate at Tufts University, went for a more complex survey by having people rank answers to a wide variety of questions on a 7-point Likert scale with 1 meaning “completely inappropriate” and 7 meaning “completely appropriate.”
The university researchers recruited 57 males and 43 females through the Amazon Mechanical Turk online service in an effort to get a more representative national sampling of the U.S. population. Their work was presented at the International Conference on Human-Robot Interaction(HRI 2016) on March 9.
More Like Masturbation Than Human Sex
Men and women shared a common understanding of sex robot capabilities and how sex with a robot should be classified in comparison with human relationships. For example, both male and female participants agreed that sex with a robot was more like masturbation than sex between humans. But men typically had greater enthusiasm than women for the different possible uses of sex robots.
One of the greatest differences in opinion came up regarding use of sex robots for sex offenders. Women showed disapproval on average by giving an “inappropriate” rating of 3.7 on the 7-point scale, whereas men gave a more favorable “appropriate” rating of 4.88 on average. Men and women also diverged in their “appropriate” versus “inappropriate” ratings for the case of using sex robots to practice abstinence.
On the other hand, both women and men generally agreed that using sex robots was more appropriate than hiring a human prostitute. They also agreed on sex robots being appropriate for use by disabled people and for reducing the risk of sexually transmitted diseases.
“We need a larger discussion on relationships and intimacy,” Scheutz says. “Some differences really may be the results of more complex differences between male and female attitudes toward human relationships.”
Significant gender differences in ratings sometimes appeared even when both men and women were generally in agreement. For example, men gave higher “appropriate” ratings to using a sex robot instead of cheating on a partner, to improve self-esteem, for making porn films, for group sex involving both humans and robots, to engage in unusual sex practices such as rough sex or sadistic behavior, and for sex education. Women also gave ratings that classified such uses as “appropriate” on average, but with lower levels of approval.
Men and women most closely agreed on using sex robots to maintain a relationship between humans, to assist training for the sake of preventing sexual harassment, and in isolated environments where normal human relationships are not available. The latter suggests that people probably won’t begrudge Mars mission astronauts or Arctic researchers their more intimate moments with future robots.
Pushing the Boundaries of Sex Robots
In terms of sex robot form and appearance, both men and women strongly agreed that sex robots should not look like a human child. That finding may reflect general disapproval of pedophile behavior. But more significant gender differences emerged for every other possible form of sex robots presented in the survey. Once again, men consistently gave higher ratings than women to every possible form of sex robots.
Sex robots that look like an adult human received the highest approval ratings from both men and women. Men gave a very high “appropriate” rating of almost 6.5 on the 7-point scale. Women also gave such sex robots a reasonably high rating of almost 5.2.
The adult human form of sex robots was followed in descending order of approval rating by “a fantasy creature,” “any recognizable life form,” “a celebrity” and “one’s current partner.” All those sex robot forms generally received an enthusiastic approval rating of more than 5 from men.
But women gave a much more tepid response with most ratings hovering between 4.5 and 4. The average rating from women even dipped slightly below 4—more disapproval than approval—for sex robots that resembled celebrities. It’s not entirely clear whether the female ambivalence about celebrity sex robots directed more toward Jude Law’s Gigolo Joe character from “A.I.” or the idea of men lusting after robots shaped like Hollywood’s latest female stars.
The differences of opinion were even greater for other possible sex robot forms. Men gave fairly high approval ratings above 5 for sex robots shaped like “one’s deceased spouse” and “one’s friend.” Women seemed to disapprove of such sex robots with average ratings between 4 and 3.
Besides the child sex robots, only two other sex robot forms drew universal disapproval from both men and women. Sex robots shaped like “one’s family member” had ratings of 3.3 from men and just below 2.2 from women, which suggests neither gender seems to approve of sex robots resembling siblings or parents. Similarly, sex robots shaped like animals drew general disapproval with an average “inappropriate” rating of 3.7 from men and an even lower rating of 2.6 from women.
Great Expectations for Relationships with Robots
Today’s sex robot technology remains fairly primitive because robotic technology is still stumbling out of infancy. But Scheutz expects the porn industry and others to continue developing such technologies until they become more sophisticated than crude sex toys. Humans have already shown a tendency to easily form emotional attachments to relatively simple machines such as Aibo robot dogs or cuddly Paro seal robots. That makes it likely that more humans could form emotional attachments with either sex robots or other future robots capable of more sophisticated social interactions.
Some people have formed emotional attachments to virtual characters that don’t even have a physical body. Certain games or smartphone apps offer a virtual boyfriend or girlfriend experience. In China, a Microsoft chatbot program called “Xiaoice” has become a beloved virtual friend for millions of Chinese social media users who message “her” to share their hopes and fears.
Hollywood examined the idea of a virtual companion to life in the 2013 film “Her,” directed by Spike Jonze. (WARNING: SPOILERS FOLLOW.) The near-future story featured Joaquin Phoenix forming a romantic relationship with an artificial intelligence named Samantha and voiced by actress Scarlett Johansson. Scheutz pointed to one of the film’s more clever twists toward the end when Phoenix’s character discovers that he’s not exactly in a monogamous relationship with his beloved Samantha. Instead, it turns out the AI Samantha has been having hundreds of romantic relationships with people all over the world.
“In addition to understanding what triggers our evolutionary buttons, we have to understand how we react to the possible capabilities of robots that are very different than our experiences with humans,” Scheutz says.
A smart AI with an attractive physical body could prove even more alluring and confusing for humans. But Scheutz does not believe a ban on sex robot technology will prove helpful in navigating future relationships with robots and AI. Instead, he wants to help improve understanding of how social robots in general impact both human psychology and human relationships.
Scheutz’s lab has already conducted a follow-up online study with slightly different questions that surveyed 200 people through Amazon Mechanical Turk. But someday, he hopes to arrange long-term experiments that could show how humans interact with social robots over weeks or possibly months. From a practical standpoint, such experiments will rely on robots becoming advanced enough to perform their own behaviors without being remote controlled by researchers.
“Once we actually understand how these interactions work out, whether they are harmful or helpful to relationships or human psychological wellbeing, then we can step in legally,” Scheutz explains. “But we need a basis for such legal decisions, so we need to gather data and do experiments.”