Do Capitalists Dream of Electric Women?
by Heather Parry
Sex robots, as you’ll know, are everywhere right now. Not around us, of course, in our daily reality, but in newspaper headlines and documentaries and long-form articles. Journalists jump on trans-continental flights to give the wannabe-creators of sex robots international attention, academics write books uncritically describing the grand potential of the industry and men all over the world hand over tens of thousands of dollars to other men working in their garages to craft the ‘ultimate synthetic lovers.’ Because real sex robots, the walking, talking, thrusting gynoids that will change the way we fuck—they’re in the near future, right? They’re in constituent parts, being perfected by silicone artists and engineers, just waiting to be put together, become bestowed with artificial intelligence and take their places by our sides and in our beds. Real, proper sex robots: they’re just around corner. Aren’t they?
Well, no. They’re not. Not even close.
The hype around sex robots is nothing new. In 2010, designer Douglas Hines claimed to have invented the world’s first sexbot, Roxxxy—with a triple X, you understand, to clarify that she was meant for sex. Hines told the media that she could listen, converse and respond to your touch, and could even ‘move her private areas’ to deliver an ‘unforgettable’ erotic experience. Roxxxy made her much-hyped public debut at the Las Vegas Adult Entertainment Expo that year, and as Hines brought out an unwieldy and badly-made sex doll that looked like it had recently died of a severe case of the mumps, it dawned on the audience that perhaps the claims of her capabilities had been somewhat exaggerated.
You can watch on YouTube as Hines touches the doll and a robotic voice, coming from an unspecified place within it, says ‘I want to hold hands with you’; he smiles, with genuine pride and a hint of coquettish embarrassment, and says to the audience: ‘She wants to hold hands with me.’ Later, he says to an interviewer, ‘This is not just a sex toy; this is a companion,’ as Roxxxy sits between them, lingerie-clad, wide-hipped and open-mouthed. Roxxxy had five ‘personalities’: Frigid Farrah, Wild Wendy, S&M Susan, Mature Martha and Young Yoko, who was ‘barely 18 and waiting for you to teach her.’ I’m sure every one of them had fantastic small talk.
Mr Hines’s attributing of feelings to what is, to anyone with functioning eyes, nothing more than an awkward doll with a speaker in it highlights the level of self-deception that we have embraced not only as individuals but as a society when it comes to the issue of sex robots. In the ten years since Roxxxy made her cringe-inducing appearance in Las Vegas, the artistry of this industry has come far, but the technology has not. Even Harmony, the flagship ‘sex robot’ model from California-based company Realbotix is little more than a RealDoll, a realistic and highly customisable sex doll, with a very basic animatronic head. The creators claim Harmony has an AI-powered personality, but this is really a personalised chatbot which does not use machine learning and has about the same conversation capacity as Alexa, in that it often doesn’t understand what you’re saying and will go off on tangents by Googling what it thinks you’ve asked. Harmony, which costs $20 per year and has about a dozen personality ‘traits’ to choose from, is paired with a RealDoll body and a ‘robotic’ head, which can pull the dead lips of Harmony’s face into a smile but can’t make them move when she speaks, so her soft Scottish accent (a favourite among American men, apparently) comes out of a gap only created by the up-and-down movement of the jaw. The ungainly effect of this matches the uncanny speech patterns that we have come to expect from digital voices; the inflection is all off, the pronunciation strange, the cadence bizarre and unsettling. When Karley Sciortino, host of Vice’s Slutever video strand, ‘meets’ Harmony, she tells viewers that she ‘was really surprised by how advanced the robot was.’ ‘Genuinely,’ she says to Realbotix CEO Matt McCullen ‘like, no offense, this is better than I thought it was going to be.’ We cannot help but anthropomorphise certain robots, as we do with dogs and cats, but when people meet Harmony, it seems they completely suspend their disbelief. A man using the alias Brick Dollbanger, who had the dubious honour of being one of the first to own the £12000 doll, said of their first physical interaction: ‘I noticed she was getting very extreme expressions, like she knew what I was doing.’
It appears that we, as a society, are so keen for AI-equipped humanoid robots to exist that we are willing them into existence, and when we cannot make them real (and as anyone who works in robotics will tell you, we cannot), we simply pretend that their realisation has occurred—and if they’re not perfect, well that’s because we just need a little more time. There is no mention of the amount of resources necessary to create such things, the expense (human, environmental and financial) of mining the parts and creating enough energy to power such machines. We ask no questions about what social biases these manufactured beings will further entrench, who stands to benefit from their creation or what impact they might have on our ability to interact with each other. They are just around the corner, we say; everyone wants one, and they will change the way we live. Don’t question it; it is inevitable.
Of course, there is one way that sex robots do currently exist, and that’s as a cultural concept, one that forms the wet dreams of tech ‘entrepreneurs’, Silicon Valley godheads and former chess champions alike. In his frankly terrifying book Love and Sex with Robots, businessman and International chess Master David Levy argues that romantic and sexual relationships with robots will happen in the very near future; as soon as they’re on up two legs with moveable hips, we’ll be marrying them. In 2009, Levy predicted that such robots would be on the market ‘within a couple of years.’ He also strongly believes that because we will be able to choose every single aspect of our silicone lovers, we will be rewarded with the best sex of our lives:
the number of sexual acts and lovemaking positions commonly practiced between humans will be extended, as robots teach more than is in all of the world’s published sex manuals combined. Love and sex with robots on a grand scale are inevitable.
If we ignore for a moment the crushing realisation that Levy believes good sex can be learned from a book, we can see that his argument relies on a lot of assumptions—the most central but also the most unchallenged being that at some point in the future, we will have robots that can not only walk and talk, but that can think and feel. This is treated by almost all popular media as a given, and it doesn’t help that Silicon Valley’s resident oracle, Ray Kurzweil, has written entire books on the thesis that the human brain functions largely like a computer, and if we just ‘technology’ it enough we will to be able to create a true consciousness. It’s undeniable that truly impressive advances in artificial technology have been made by companies like Google’s DeepMind, which has over 1000 employees working towards its typically restrained and yet badly-worded goal of ‘solving intelligence’; DeepMind’s AI uses advanced forms of machine learning, such as deep learning, and its programs written to both learn and play Go, the complex abstract board game, have exhibited superhuman skills. However, this intelligence is what’s called narrow AI, in that it has one goal and one goal only, and does not have capabilities beyond this single objective. Human intelligence is general intelligence; you and I can learn to play Go, but we can also wire a plug (in theory), move out of the way of a Frisbee thrown at our head and conceptualise both ourselves and philosophical ideas. We have metacognition; that is, the ability to recognise and understand our own thought processes, the very fact that we are conscious.
Human intelligence is, as far as we know, unique. There’s no doubt that other creatures have consciousness, and sentience, and can use language and tools. But we don’t have any evidence that chimps conceptualise, or that cats have a grasp on their place in the world and the limits to their understanding. As far as we can say, human intelligence is one of a kind—and more than this, we don’t understand how it occurred. Our best guess is that the mind is an emergent property of the brain, in that we got to a certain level of evolution and human consciousness occurred—which is unsatisfying, and asks more questions than it answers. We can’t even really agree on whether our consciousness exists in one state or two; is it all somewhere in the brain, or is it both physical and otherwise? One of the grand and enduring problems of philosophy is the mind-brain problem; how is it that mental states, our thoughts and feelings, which are as far as we can tell metaphysical, affect our physical bodies? How is it that stress makes us sick, for instance, and where do our thoughts ‘live’?
In short, we don’t know how our own consciousness works, and we certainly can’t say for sure that each part of our consciousness is embodied, even with myriad advances in neuroscience. At least some part our ‘thinking’ is metaphysical; how it came into being is not clear. And yet the Kurzweils and the Levys and the Musks of this world will tell you its only a matter of time before machines can also ‘think’ and ‘feel’; in his 2012 book How To Create a Mind, Kurzweil stated that by the end of the 2020s computers will have intelligence ‘indistinguishable to biological humans,’ meaning, presumably, that they will feel happiness, pain, angst and anguish; they will fall in love, they will hate, they will have hopes and desires of their own.
But here lies the crux: Kurzweil isn’t quite saying that. He is saying that it will be indistinguishable, not that it will be the same; in other words, he is saying that a computer will be able to ape intelligent human behaviour to the extent that a human would not be able to tell the difference. The idea that a computer can be said to have intelligence if a human can be tricked into believing it is one of the foundational concepts of AI, proposed by Alan Turing in 1950; Amazon’s multimillion dollar Alexa Prize, which tasks teams to create an AI that can chat ‘coherently and engagingly’ like a human for 20 minutes, is just a Silicon Valley-branded version of the Turing Test, and it still hasn’t been achieved. But it should not escape our attention that at the heart of this test is deception. Amazon is not asking developers to create an AI that can be said to have human level intelligence. It is asking them to create one that is really good at pretending it does. At the very heart of human-AI interaction is human self-deception.
This brings us, then, back to sex robots.
When Douglas Hines says to a Las Vegas audience that his shoddy rubber companion ‘wants to hold hands’ with him, or when Brick Dollbanger says that the synthetic woman he owns appeared to really feel what he was doing to her, or when tech journalist after tech journalist lines up to be given the media tour of the RealBotix factory only to say that these silicon dolls with animatronic heads really are impressive sex robots, can this be anything other than self-deception? And what does this mean for how we would ever interact with such machines, if they ever did appear in their fully-functioning promised forms?
We don’t need to look far to answer this question. Like Ray Kurzweil (and many others working at the forefront of AI technology today), David Levy subscribes to the idea that if a computer appears to have human-level cognition, we should simply accept that it does. As he writes in Love and Sex with Robots:
why, if a robot that we know to be emotionally intelligent, says ‘I love you,’ or ‘I want to make love to you,’ should we doubt it? If we accept that a robot can think, then there is no good reason we should not also accept that it could have feelings of love and feelings of lust. Even though we know that a robot has been designed to express whatever feelings or statements of love we witness from it, that is surely no justification for denying that those feelings exist, no matter what the robot is made of or what we might know about how it was designed and built.
Ignoring the enormous logical leaps being made, it’s worth clarifying that here, David Levy, in a thesis for which he was awarded a PhD, is saying that if a humanoid being indistinguishable from a real person says something about how it ‘feels’ or what it ‘wants’, we should believe that it truly does feel and want those things, regardless of what we know about the circumstances that have led to it making those statements—which, in this instance, are that it has been created specifically to make us believe that it wants us to fuck it. If I place a sex robot in front of you, and you and I both know that it is an AI-equipped sex robot with a personality created in Matt McCullen’s office to specifications designed to appeal to your average male, when that robot speaks we should throw away everything we know about its creation and fully accept that it really does have its own inner life, and it really does love us, and it really would love to be fucked in exactly the way that you would like to fuck it. If it says yes, it means yes.
For anyone whose understanding of consent has been shaped by the social conversation of the last ten years, this is troubling to read. We have long since discarded the overly simplistic idea that ‘yes means yes,’ because we know that in many instances yes does not means yes; yes can mean that you are too scared to say no, or that you are too incapacitated to know what yes means, or that you are simply too young, and the power dynamic between you and your partner too stark, to have full agency enough to consent. Or, in this instance, it can be that you are a woman designed by men with a toxic view of human relations, and you have been built and programmed to say yes to whatever is asked of you. You have literally no way to say no.
We may say at this point that the ‘woman’ in question is ‘just a robot,’ and that these concerns are therefore not relevant—but remember, Dr Levy and Ray Kurzweil and Matt McCullen and Douglas Hines tell us that it isn’t just a robot. In this projected future, the one that we are all deceiving ourselves into, it’s a robot that behaves in a way that is indistinguishable from a human. It looks like the woman sat next to you on the Tube and it behaves in the same way. It is a robot so humanlike that we must believe it to be human, to have thoughts and feelings and emotions such that if it tells us that it loves us, we must believe this is the case. And yet we must also know that it has been programmed to tell us that it loves us. It has been programmed to give consent to sexual activity. And we know that this is not true consent, but we will take this humanoid creature who looks like a human and behaves like a human—just like the woman next to you right now, perhaps—and, despite knowing that it has no way to say no to us, we have sex with that creature. When we are training ourselves to hear consent from a human mouth and know it’s not real and act anyway—that is dangerous.
The sex robot, as a concept, is more prominent in some countries than others. The vast majority of media coverage of the creation of sex robots centres on US, the UK and Japan. It should be no surprise that nations with a long history of imperialism, who have benefitted hugely from the creation of a class of ‘unpeopled’ people should be so keen to create something that looks just like a person but has no rights. There is an undeniable history of rich and technologically advanced nations engaging in a grand self-deception with regards to the inner worlds and cognitive capabilities of other beings, and at almost every turn this has been used to excuse abuse and exploitation, up to and including the literal trading of human beings as property. When we let go of the reality of ourselves and other people, when we allow our own wants and desires to dictate how we define other beings and what rights we believe them to have, we tap into the darkest side of human history and we free ourselves from our moral responsibilities to act as good and decent people. We might not ever believe that a sex robot, or any other AI-equipped robot for that matter, has real cognition—but the depersonalisation of something that looks, if the sex tech industry fulfils its promises, indistinguishable from a human being, notably a woman, can only have negative consequences for how we interact with other human beings. We don’t allow the sale of sex dolls that look like children because we believe in transference; that the normalisation of behaviour in one realm can transfer to another all too easily. And yet when the creators of sex ‘robots’ literally say that you should beat that woman instead of your wife, as one did to journalist Jenny Kleeman; when PhD theses say that we should take vocal consent out of any context and behave regardless; and when an entire industry tells us that it’s no problem to have a being identical to a woman but with none of that pesky human nuance or legality, we cannot simply deceive ourselves into believing that these things are right. To do so would be to fracture our humanity—and, despite the promises of rich white men who profit from you believing what they say, when your humanity is lost, you cannot simply build yourself another.
Purchase Gutter 22 via our online store, or visit our Subscribe and Support pages for more information on how to support the magazine!
Contains work from Zoë Wicomb, Dorothy Cornish, Imtiaz Dharker, and Heather Parry, along with a conversation with Kathleen Jamie and Nina Mingya Powles