The Formula_How Algorithms Solve All Our Problems... and Create More

by Luke Dormehl

  Such devices don’t have to remain the sole province of bored teenagers in Tokyo’s Harajuku district, of course. In the Republic of Iceland, the Islendiga-App (“App of Icelanders”) applies similar technology to the problem of solving the issue of accidental incest in a country of 320,000, where almost everyone is distantly related to one another. By accessing an online database of residents and their family trees stretching back 1,200 years, and then using an algorithm to determine the shortest path between two points, the app is able to inform users how closely related they are to the person they might be considering sleeping with. The app is activated by “bumping” phones, which in turn triggers an “incest prevention alarm” in situations where the number of genealogical steps is sufficiently small as to cause potential distress at future family gatherings. In the words of the app’s admittedly catchy slogan: “Bump the app before you bump in bed.”28

  Today, a large number of proximity-based social-discovery apps—from Grindr to Crowded Room—vie for prominence in the marketplace. Among the more interesting of those I came across during my research was Anomo, an app that lays out its ambitions to “democratize the way we socialize” in a way that combines both the actual and virtual worlds. Cofounded by a former runner-up on the U.S. version of The Apprentice, Anomo asks its users to create video-game-style avatars for themselves, which then become their virtual proxies as they negotiate their way through the actual world. “A lot of existing social discovery apps are about real profiles,” says cofounder Benjamin Liu. “You put your real picture and name on there. The problem with that is that people’s first reaction to it is often, ‘Wow, that’s really creepy.’” Anomo users do enter their real information to the app, but this is hidden until users choose to reveal it in a series of quid pro quo, “you show me yours, I’ll show you mine” exchanges. Prior to that, users have the option of either chatting to one another using their avatars (which consist of a short description, and one of around 100 different cartoon representations), or else playing a series of “icebreaker” games to determine how much they have in common with a person without having to endure a potentially awkward chat first. By comparing responses to questions such as “Do you prefer beer or wine?” or “Is a bicycle built for two cheesy, cool, or dangerous?,” Anomo’s app provides a “compatibility rating” ranking your answers next to those given by the other person. “Interacting anonymously changes the playing field,” states Anomo’s promotional literature. “Suddenly first impressions are not based on photos, but via a genuine connection.”

  Tapping the Scene

  For the most part, social discovery works along the same lines as Instant Messenger. Rather than focusing on finding “the one” users are instead dealing with a pool of available participants that exists in a constant state of flux. It doesn’t matter if a particular message is not returned, or the person we originally wished to “chat” with is no longer present. A person simply moves to the next user in line and begins the process over again, since there are always enough people online (at least in the case of those apps that find success) to counteract the loneliness of any given moment.

  While most of these apps require that people consciously engage with them, this rule is by no means an absolute. In his most recent book, Who Owns the Future?, computer scientist and virtual reality pioneer Jaron Lanier recalls a panel he served on at UC Berkeley, judging start-up proposals submitted by graduate engineering students enrolled in an entrepreneurial program. A group of three students presented a concept for quantifying nights out so as to ensure maximum romantic success for those involved. “Suppose you’re darting around San Francisco bars and hot spots on a Saturday night,” Lanier remembers the group pitching. “You land in a bar and there are a bounteous number of seemingly accessible, lovely, and unattached young women hanging out looking for attention . . . Well, you whip your mobile out and alert the network.”29 Such an idea would, of course, never work, Lanier observes, since the data would invariably be inaccurate and the scheme would wind up running on hope.

  His words of warning have not been enough to derail the technology’s core concept, however. SceneTap—previously known as BarTabbers—started life in Chicago, although it has since expanded to cover San Francisco, Austin, Columbus, New York, Boston and Miami. With cameras installed in more than 400 drinking establishments, SceneTap uses facial-recognition technology and people-counting algorithms to help bar-hoppers decide which locations to hit up on a particular night out. Currently, the tool can provide real-time information on crowd sizes, gender ratios and the average age of patrons in any given location—although that is not everything that’s planned. In 2012, the start-up filed a patent that, in the words of Forbes.com, “[crosses] the creepy line . . . and then keeps running and spikes the creepy football in the creepy end zone.”30 In a nutshell, the patent is designed to allow SceneTap to collect much more detailed data, including bar-goers’ race, height, weight, attractiveness, hair color, clothing type and the presence of identifying features like facial hair or glasses. In other words, one might access the app and be presented with information along the lines of: “The Raven is 73% full. Its crowd is made up of 33% natural blondes, 57% brunettes, 3% redheads, 5% bottle blondes, and 2% other. The men on average are 5.8 feet tall, and 70% are dressed in business casual. Women on average weigh 154.7 pounds, and 24% wear short skirts. 73% of patrons are white, 21% Asian and 6% black. Attractiveness average for the location is 7 out of 10.”

  The patent also allows for microphones to be placed in cameras in order to pick up on what customers are saying, as well as for the facial recognition technology to identify people and link them with their social networking profiles to determine “relationship status, intelligence, education and income.”

  No doubt STD tracking could be added at a later date.

  Your Sex Life with Models

  “I’ve been in a relationship with just one woman my entire adult life, ever since I was in high school,” says Kevin Conboy, a soft-spoken computer programmer with a mop of brown, curly hair and a Brian Blessed beard. About ten years ago, when Conboy was in his mid-twenties, working as a user-interface engineer, he tried to work out how many times he and his wife had had sex in their time together. The idea grew in scale and, before long, Conboy was spending his evenings working on an application to keep tabs on his sex life: methodically cataloguing and modeling everything from the duration and frequency of sex to its quality and average time frame.

  “I strive for honesty in my life as much as possible, but I thought that this was one thing that was better to build in secret and then to show my wife once it was completed to ask for permission,” he says. “At first, she thought it was kind of weird, but she also understood that to me code is a form of self-expression. She asked me for the link to view it, and she would check up every now and then to see how our sex life was doing. It got to the point where she would tell her friends about what I had created, and it sort of became this talking point.”

  It was a friend of Conboy’s wife who eventually convinced him to open up the app for public use. Giving it the name Bedpost, and the tagline “Ever wonder how often you get busy?,” the app advises users to “simply log in after every time [they] have sex and fill out a few simple fields. Pretty soon, you’ll have a rolling history of your sex life on which to reflect.”31

  This mass of data can be visualized in a variety of forms—including pie charts and scatter plots—with heat maps showing different intensities of color based on the quantity and quality of sex a particular user is having. “The amount of data you can attach to sexual activity is uncapped,” Conboy says. “For instance, if you’re on your phone, there is no reason you can’t record the GPS data. You might be having sex all over the world and it could be fun to look back at all that information.”

  The goal of his work, Conboy says, “is to get you to think about your sex life in a way that you hadn’t previously.” Its success hints at somethi
ng interesting: that the act of measuring and quantifying sex in the form of zeroes and ones can in itself become an erotic act. After all, if text-based cybersex based entirely around semiotic interaction can be arousing then why can’t the code that underpins it?

  When I spoke with Conboy for the first time, he was busy rethinking the Bedpost user interface, which had remained largely unchanged for several years. A number of the site’s best suggestions, he said, had come from the app’s users themselves. The popular feature of menstruation tracking, for example, was something he says he never would have come up with by himself. In particular, he was fretting over whether or not to add an orgasm counter to the list of available features. There was also the question of adding greater social-networking capabilities (“This might freak some people out, but I like the idea of being able to pull people in via your Facebook and Twitter accounts”), along with the ability for users’ partners to log in and add their own “tags”—thereby creating a type of wiki criticism of “sex-as-performance,” which could take place long after the act had taken place.

  The ability to share the experience of sex is one that is particularly fascinating. What if we were able to get closer to the other’s sexual experience not simply by asking, “How was it for you?” but by actually delving into the data for evidence of the other person’s satisfaction? This wouldn’t necessarily have to rely on anything as subjective as user tags. A number of pieces of wearable tech have already shown themselves to be uniquely valuable when it comes to quantifying the sex experience. By recording heart rate, perspiration output and motion patterns, the BodyMedia FIT armband can, for instance, recognize whether wearers are having sex at any particular time. More intrusively, it is even able to gauge whether wearers are having good sex or not, since analyzing spikes in the various metrics can reveal if a partner has faked an orgasm.32

  Awkward conversations aside, this available data has the potential to raise a number of issues. For example, the spouse who suspects his or her significant other of cheating could access their data and question why 100 calories had been burned off between 0:13 and 2:00 A.M. on a particular evening, without their taking a single step, and with them falling asleep immediately afterward. Removing said tech (or else “forgetting” to wear it) might be grounds for suspicion in itself. Even arranging for illicit encounters to take place during daylight hours, when data anomalies would theoretically be more easily explained away, would be of little use since sexual activities look different in terms of fitness data than other energetic activities like weight-lifting, jogging, yoga, martial arts and cycling.

  For Conboy, Bedpost helps to normalize him within the sexual spectrum. “The aggregate data gives me a sexual confidence,” he says. “It’s nice to know that you have had sex a certain number of times this month, or this year. It’s a reminder that my sex life is healthy. The numbers help me to relax about whatever I’m obsessing over at the time. I don’t think I could have an erotic sex life without the confidence the data gives me.”

  Of course, this raises yet more questions—not least how much data is enough?

  No matter how scientific the intention, the moment we start measuring we also begin limiting, not just based on what the measurement tool is designed to capture, but on those metrics we consider worthy of measurement at the time. In the case of Bedpost, Conboy might add an orgasm counter, but how about respiratory rate? Or if he adds respiratory measurements, what about perspiration levels and heart rate? And so on it goes—the ideal and meaning of what we are measuring receding further and further from view, like advancing on an Impressionist painting and seeing what from a distance had been a convincing re-creation of a landscape dissipate into a sea of painted dots.

  Night of the Loving Dead

  If we accept the (highly questionable) view that true love, like opportunity, knocks only once, then how about the idea that love must automatically be separated by its deathly binary opposite? It has long been a technological pipe dream that man should be able to transport his “spirit” beyond the corporeal body and into the metallic lattices of a computer. In The Enchanted Loom: Mind in the Universe, NASA futurist Robert Jastrow waxes lyrical about his hopes for a future in which the human brain could be “ensconced in a computer . . . liberated from the weakness of the mortal flesh . . . in control of its own destiny.”33 As with natural birth, such a form of reproduction would ensure our immortality through the continuation of our perceived wisdom, good and happiness. To frame this as a question about love, what if apparent death didn’t therefore part us from our loved ones but rather, as with a butterfly emerging from a chrysalis, simply meant a change in form factor?

  It is into this space that applications such as IfIDie, DeadSocial and LivesOn enter the frame. All three apps exist in various magnitudes of complexity. DeadSocial and IfIDie function as what can essentially be described as netherworldly “out of office” services: allowing nothing more complicated than the writing and recording of timed Facebook messages, activated upon a particular user’s death, and which can then be sent out to lovers, friends and family for years to come. In a video designed to promote the former service, we are told how the widow of a deceased DeadSocial user can even be “the recipient of an inappropriate joke from time to time.”34 IfIDie, meanwhile, recommends the recording of a webcam video in the event that one wishes to bid a fond farewell or (accompanied in literature by an image of a heavenly cloud in the shape of a raised middle finger) “settle . . . an old score” with a member of the living.35

  More complex, both ethically and technologically, is LivesOn—which carries the distinctly Twilight Zone tagline, “When your heart stops beating, you’ll keep tweeting.”36 LivesOn uses machine-learning processes to sift through your past social-network feeds, looking for the subjects, likes and news articles that interested you during life, so that similar subjects can be tweeted about on your behalf after death. Users are encouraged to help their LivesOn “train and grow” by following a person’s existing Twitter feed and analyzing it, learning tastes, likes, dislikes and eventually even syntax. “The goal is to get it to almost become like a twin,” says creator Dave Bedwood.

  That isn’t all. Several years ago, news broke that the U.S. Department of Defense was developing a project designed to create “a highly interactive PC or web-based application to allow family members to verbally interact with virtual renditions of deployed Service Members.” Using high-resolution 3-D rendering, the high-level concept was to allow a child or spouse to engage in a simulated conversation with an absent parent or partner, and receive responses to stock phrases including “I love you” and “I miss you.” “I’m not saying this kind of ghost is for everyone,” commented one writer for Slate, with what is perhaps something of an understatement,

  but I dare you to tell a child who has lost her father in Iraq or Afghanistan that she can’t keep a virtual rendition of him to help her go to sleep. And I dare you to stop the millions of others who will want ghosts of their own when today’s military project becomes, once again, tomorrow’s mass market.37

  While it would be the harsh writer who would willingly inflict pain on a person who has lost their loved one (although it seems odd to blame said writer for the absence of the parent/spouse to begin with), one cannot help but think that Slate has not properly thought through its position here. A computer program that is able to convincingly replicate the appearance, voice and sentence structure of an absent or deceased relative is neither quantitatively or qualitatively the same thing as having said person there—much as a dead body is not the same as a live one, even if it might look no different in a still photograph. To presume that the person we fall in love with, or even just the objects that hold personal value for us, are simply a composite of their individual properties is incorrect. Studies have shown that taking away the security blanket or teddy bear of a young child and substituting it with a duplicate will not see the replacement accepted in place of the original. W
e may well react in the same way if our Rolex watch was replaced with an unofficial replica, no matter how convincing a knockoff this might prove to be. Taken to extremes, it would be the remarkably unfeeling doctor who would reassure the parent of a dead child, or the widowed spouse of a dead husband or wife, by telling them that their loss needn’t matter since we can clone the person in question—or better yet, that they have a twin who is still alive. In his writing, the French phenomenological philosopher Maurice Merleau-Ponty explored this idea, by distinguishing between what he views as the living and dead properties of an object. It is in The Formula’s empirical desire to reduce perception and feeling to that which is observable, he argues, that all elements of “mystery” are lost.38