A former leading figure in Google’s AI endeavors has warned, or rather proclaimed, that AI sex robots are coming soon, and that they will be indistinguishable from the real thing, so much so that men may even prefer them. More surprisingly, Mo Gawdat openly states that he doesn’t see it as something bad, if an individual feels the same love and gets the same emotional satisfaction from a relationship with a virtual AI lover that is not a real and sentient human being.
“Just think about all of the illusions that we’re now unable to decipher illusion from truth, right? Sex happens in the brain at the end of the day, I mean, the physical side of it is not that difficult to simulate, okay? But if we can convince you that this sex robot is alive or that sex experience in a virtual reality headset or an augmented reality headset is real, then there you go.
There is that huge debate of ‘are they sentient or not?’ Does it really matter if they’re simulating sentience so well?”
Gawdat goes on to point to brain-computer interfaces such as Elon Musk’s Neuralink, and sees a time in the not-too-distant future when virtual reality porn is plugged directly into the nervous system and we “wouldn’t need another person”. When questioned as to whether such ‘fake’ relationships with non-sentient AI characters could be as satisfying as a relationship with a real person, the ex-Google executive makes clear he’s in the camp who believes that they can be.
The case for ‘artificial intimacy’ and a future in which millions of people, perhaps the majority, prefer relationships with AI companions was made several years ago by evolutionary psychologist Rob Brooks in his book of the same title. I wasn’t particularly convinced upon my first reading of that book. I’d always been skeptical and a little non-plussed by people who appear to feel genuine affection for any kind of virtual companions, such as the Tamagotchi pets that were popular in the 90s and 00s. And I’ve always assumed that the stories of men marrying their realistic sex dolls or such were either marketing ploys, exercises in self-publicity, or frankly, the individuals were mentally ill. The evolutionary argument that Brooks makes in his book is rather long-winded and begins with the domestication of the dog, or rather the domestication of humans by dogs. Brooks points out that dogs have learned to push the right emotional buttons when it comes to their human masters over thousands of years, or at least the ones that didn’t have failed to reproduce their genes. AI companions will see a similar evolutionary process play out, except it will be thousands of years condensed into a decade, or even a few months. Further, there is virtually no limit to how well your AI companion will be able to understand you and respond to your exact needs. She will be able to understand you better than any real person (or dog), even better than yourself. There is no doubt that AI girlfriends will be able to satisfy their human companion’s emotional and sexual needs in a variety of ways that go beyond the ability of any real girlfriend. But will there still be a psychological gap that means real girls will never be replaceable by non-sentient sexbots? Mo Gawdat certainly doesn’t think so.
The AI girlfriend industry is still in its infancy. While virtual girlfriends such as Replika have been around for a number of years, they will rapidly improve, due to recent generative AI advances, in both image creation and human-like chat. And soon, as mentioned in the video above, VR and AR girlfriends will become a very popular thing. The success of Replika and other apps like it, and the passionate reaction from thousands of Replika subscribers when she had her algorithms changed overnight, demonstrate the vast potential market for virtual girlfriends, and their ability to already satisfy the emotional needs of a huge number of people. The first AI girlfriend simulator site to combine AI image generation with chatbot features – DreamGF.ai – was launched just a couple of months ago, and I have a personal knowledge as to how successful it is already, with thousands signing up to it each day. Whether or not these AI girlfriend simulators are replacing real women, they are certainly already meeting some needs for many thousands of men.
Another way that the ‘psychological gap’ can be bridged is by basing AI girlfriends on real people. This has already been famously done with a number of influencers, with the first that was brought to the attention of the public being Caryn Marjorie – transformed into CarynAI. And not long after, a developer in Romania ‘cloned’ his girlfriend and turned her into an AI girlfriend available for everybody. He even uploaded the Python script to GitHub.
Mo Gawdat makes the point that a generation that has grown up interacting online, is already finding it easy to form bonds with AI companions. He says that part of the brain knows the person is not real, but it doesn’t matter to them. I was thinking about this yesterday when I read about an OnlyFans model who had taken part in a boxing match. After winning her bout, she exposed her breasts to the cameras – and it wasn’t a particularly attractive sight. Glammed up with plenty of make-up and her hair done perfectly, and no doubt with a little help with filters and photo-editing, her pictures leave hundreds of thousands of thirsty simps begging (and paying) for more. But after a hard and sweaty boxing fight, unsurprisingly she looks quite different. I wonder if any of her hundreds of thousands of fans really care. And if they can pay money for photos of a real person who can look so different in ‘real’ life, why wouldn’t they pay for the photos of an AI girlfriend who is ravishingly more beautiful, and never has a bad hair day or an unflattering pic of herself (well, perhaps an extra limb or two occasionally), and of course, can give each of them a level of personal attention impossible for a real individual (unless they too have been cloned as an AI bot)?
