Click and connect: the rise of ‘artificial intimacies’
The first thing I noticed about Roxxxy was not her improbable proportions but the look frozen on her face, somewhere between pleasant surprise and utter horror.
The first thing I noticed about Roxxxy was not her Photoshop-smooth complexion, the slouching way her body had been arranged on the chair, or the flimsy black garment straining against her improbably proportioned torso. The first thing I noticed was the enigmatic look frozen on her face, somewhere between pleasant surprise and utter horror. There could be no mistaking Roxxxy Gold True Companion – her full name – for a human being. Indeed, one could reach no other conclusion than that she was an object created to satisfy a prosaic heterosexual male fantasy. And yet, here I am, unable to decide whether to call Roxxxy “it” or “her”.
On first impression, there seems little difference between Roxxxy and the sex dolls that have gradually grown more lifelike in recent decades. But when Roxxxy debuted in 2010, her creator Douglas Hines, formerly an artificial intelligence engineer at Bell Labs, touted her as “the world’s first sex robot”. Not because she moves about freely – her movements remain limited to a few simple arcs – but because traces of artificial intelligence (AI) breathe a semblance of life into Roxxxy and her male stablemate Rocky. They can recognise and generate speech, “have an orgasm”, and supposedly they have personalities.
Despite enormous popular and media interest in such technology in recent years, and worries that it will, variously, displace us in the bedroom and corrupt our morality, many people will be relieved to know that today’s sex robots really aren’t that good. Roxxxy and Rocky remain pale imitations of the living, breathing companions that most people seek. There is a long way to go before anybody can manufacture something like the warm-blooded, flawed, fellow human beings we know, love, and, occasionally, take to bed.
But while we gawk at the sex robots, trying to process them in all their uncanny weirdness, other technologies based in artificial intelligence and virtual reality are insinuating themselves into human interactions, quite likely with more profound effects. Collectively, these are the “artificial intimacies”: technologies that engage our human needs for connection, intimacy and sexual satisfaction. Machines that can help us make and maintain friendships in a world of cognitive overload. Machines that can help us feel better. And machines built to feed back to us whatever it is that they need us to see, hear or feel. Many artificial intimacies will simply refine technologies that already exist, including social media and video games. Others will look entirely new. Together, they will likely transform the quality of human life.
Digital lovers don’t have to be as conspicuous as gynoid or android sex robots. They just have to channel the power of AI to provide better sexual experiences. Smart sex toys are just getting started, but they are undeniably a big part of sex tech’s future. Contemporary “teledildonic” sex toys are already most of the way there, allowing a friend or lover to control a user’s favourite toy remotely – hence “tele” – from anywhere with an internet connection via co-ordinated apps.
Imagine being separated from your lover during an extended period of overseas work or, heaven forbid, a global pandemic. What if the two of you could, in the comfort of your distant homes, enjoy full immersive interactivity as your VR avatars have sex with one another? Every move, gesture and gasp could be felt, seen and heard by the other through headsets, teledildonic devices and haptic suits.
When I was at university in the early ’90s, the only legit way to meet somebody was to encounter them in the flesh. We met our girlfriends and boyfriends at parties, in class, and through friends. But within a generation people of all ages would routinely advertise and find each other on a bewildering ecosystem of online platforms such as Grindr, OkCupid, Bumble, Happn, Silver Singles, Hinge, Tinder and even Christian Connection.
Those matchmakers broker a solid proportion of new hookups and romances. The most recent wedding I attended was of two wonderfully well-matched people who met on Tinder. Three in five Australians surveyed in 2017 had used online dating platforms. Online matchmakers have proliferated because AI has made them good at identifying people who are physically near to each other and could possibly hit it off. They work by presenting users with one another’s profiles, and then, if they both like what they see, declaring a “match” and putting them in touch with one another. That sounds simple, but it takes programming genius and plenty of data. The data-rich environment of profile information, swipes, likes and matches creates the perfect soil in which machine learning can flourish. The matchmakers with the best algorithms will shape who meets, who matches, and, ultimately, who mates.
Beyond the birds and the bees, similar matchmaking algorithms can find us friends and mentors. Whenever humans and machines interact, and those designing the machines use AI to tailor that interaction, there exists the possibility of artificial intimacy. Most of us have already met at least one of the “big five” AI assistants: Apple’s Siri, Amazon’s Alexa, Microsoft’s Cortana, Baidu’s DuerOS and Google’s Assistant. When I say, “Hey Siri, please define artificial intelligence”, my iPhone records my question, sends it to a cloud-based processing centre to be processed and actioned entirely by computer, and the chosen response is sent back. Within little more than the usual human pause between question and response, I hear Siri’s reassuringly calm voice answering, “The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.”
All of the virtual assistants can help users recall information, find a recipe for tonight’s meal of cacio e pepe, remind us to pick up pasta and pecorino on the way home, and kick Spotify into action to play that Lana Del Rey song about Norman Rockwell. People ask, or used to ask, those kinds of mundane things of their friends, housemates, or lovers during the course of a regular day.
What about the other things we do for one another in our relationships? If Alexa can learn to give us what we need vis-à-vis cooking and grocery advice, will it also learn to give us what we want even when we don’t ask? Smartphones have so many sensors: microphones for telephony and voice commands, barometers for air pressure and altitude, motion sensors, ambient light sensors, moisture sensors, three-axis gyroscopes, touch ID sensors, and, of course, cameras that can identify a user’s face and record video. When does all this sensory ability turn into sensitivity? The hardware is in place. Coupled with all manner of artificially intelligent applications, we can expect devices to learn to infer our mental states, frustrations and desires, and then to act on them. How long will it be before your virtual assistant can sense from your tone of voice, the speed with which you shift your phone from hand to hand, and the levels of moisture on your fingertips, that your anxiety is off the charts? And wouldn’t it be a good thing if it can learn what you need, in order to manage your own particular brand of anxiety, and reassure you by playing a personalised track of autonomous sensory meridian response (ASMR) trigger sounds designed and optimised by machine learning to mitigate your particular form of anxiousness?
Perhaps it will learn to sense when you slip towards depression and assess from the symptoms whether you need to talk, or you just need somebody to talk quietly to you (in which case it could then start playing Ralph Fiennes reading The English Patient). Digital therapists are already picking up their virtual pads and pencils. The first chatbot, ELIZA, emulated a psychotherapist. Today, all manner of therapy chatbots exist to provide very real therapeutic help.
Likewise, a slew of chatbot apps encourage confession – both the Catholic type and the more secular forms of unburdening. Some, like Twitter’s Confession.bot, will broadcast your confessions anonymously. Confession: A Roman Catholic App ($1.99) offers a 21st-century alternative for those Catholics who are too busy, claustrophobic or priest-averse to go to regular confession. With a personalised examination of conscience for each user, password-protected profiles and a step-by-step guide to the sacrament, this app invites Catholics to prayerfully prepare for and participate in the Rite of Penance.
So far so good. Recipes, reminders and confessions might also be needs and, if well designed, applications like these will help people in their relationships with themselves and others. The potential for useful artificial intimacies that fill new and emerging gaps, that help people work through their issues and just get through all their tasks, seems almost beyond limitations.
Social distancing and isolation accelerated the shift to online social interaction, perhaps irreversibly. People spending more time online using a greater variety of apps are inadvertently hastening the development of artificial intimacy even as I write. More users equals more data from which machine learning algorithms can learn about human interaction.
The digital lovers, virtual friends and algorithmic matchmakers of the very near future are being designed and built right now. What artificial intimacies learn about human preferences, behaviours, worries, likes and dislikes will be used in coming years to develop newer and more compelling artificial intimacies.
When you dip your extremities into the pond of artificial intimacy the waves of data you create will ripple outward, far beyond the shores of your imagination. All manner of machine learning algorithms will analyse and learn from it, eventually serving up more compelling advertisements and content to you, and to people like you. As futurist and author of Team Human, Douglas Rushkoff, puts it: “Each one of us is not just up against whichever algorithm is attempting to control us, but up against them all.”
Rob Brooks is Scientia Professor of Evolution at UNSW. Edited extract from Artificial Intimacy – Virtual Friends, Digital Lovers and Algorithmic Matchmakers (NewSouth, $32.99), out now.