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

by Luke Dormehl

  Certainly, Internet dating offers quantitatively better odds of finding a partner. Upward of 30 percent of the 7 billion people currently alive on Earth have access to the Internet, with this figure rising dramatically to around 80 percent in the United Kingdom and North America, where usage is among the highest. Where previously the romantic “field of eligibles” available to us (as sociologist Alan Kerckhoff once phrased it) was limited to those people we were likely to come into contact with as part of our everyday social network, the Internet now provides access to an unprecedented quantity of people, numbering around 2 billion in total.13 Of these, a growing percentage have shown interest in using the Internet to find love. In a typical month, almost 25 million unique users from around the world access some form of online dating site.14

  Does it offer a qualitatively better experience, however? This is, after all, the mission statement of websites like eHarmony—which jealously guard not only their user base of customers, but more importantly the proprietary algorithms used to match them. Generally speaking, sites that offer matching services tend to rely on similarity algorithms (with a certain amount of complementarity matching added in). Partners are matched on numerous personality traits and values, reflecting not only what the creators of a particular algorithm deem important, but also what is prized by its customers. This, in turn, has allowed a broad range of matching sites to pop up around eHarmony, each one catering to a different clientele who stress the importance of different aspects of the romantic experience.

  For those interested in playing the numbers game, there is OkCupid, whose tagline borrows from eHarmony’s in-your-face scientism by promising that “we use math to get you dates.”15 For the ultra-scientific (or perhaps the latent Eugenist) there is GenePartner.com, a website that claims to have “developed a formula to match men and women for a romantic relationship based on their genes.”16 For the low, low price of just $249, would-be daters can order a kit with which to swab the inside of their mouths. This saliva sample is then mailed back to the lab at GenePartner, where it is analyzed, and the results fed back to the user. For an additional fee, GenePartner can even match genetically similar users with their supposed soul mates. “With genetically highly compatible people we feel that rare sensation of perfect chemistry,” the company’s press materials state. “This is the body’s receptive and welcoming response when immune systems harmonize and fit well together. Genetic compatibility results in an increased likelihood of forming an enduring and successful relationship, a more satisfying sex life, [and] higher fertility rates.”

  For the narcissistic, meanwhile, there is FindYourFaceMate .com, which hails itself as a “revolutionary new online dating site that employs sophisticated facial recognition software and a proprietary algorithm to identify partners more likely to ignite real passion and compatibility.” Based on the theory that we are innately drawn to those people with features that resemble our own, subscribers to FindYourFaceMate upload profile pictures, which are then analyzed on nine different facial parameters (eyes, ears, nose, chin and various parts of the mouth) in order to find suitable “face mates.”17

  Then there is BeautifulPeople.com—described by its founder as a “gated community for the aesthetically blessed”—which marries skin-deep shallowness with a striking free-market ideology, summed up by its promise that “beauty lies in the eyes of the voter.”18 For the plus-sized dater, there is LargeAndLovely .com (“Where Singles of Size & Their Admirers Meet”) and, on the other end of the spectrum, FitnessSingles.com (“Where Relationships Workout”). There are sites like SeaCaptainDate .com (“Find Your First Mate”) and TrekPassions.com (“Love Long & Prosper”); those that match based upon your particular strain of vegetarianism (VeggieDate.org); and those that pair people on a shared taste in books (ALikeWise.com) or a fondness for those who work in uniform (UniformDating.com).

  The idea, essentially, is to drill down until we discover the particular weighted node that best captures our fancy. For some, their vegetarianism might be a defining characteristic and, therefore, a “must have” demand. For others it is nonessential, or even incidental. Online, not only is everyone a formula, as argued in the previous chapter, there is also a formula for everyone.

  Love in the Time of Algorithms

  There is a scene in the 2009 comedy film Up in the Air in which Natalie, an upwardly mobile businesswoman just out of college, describes the qualities she is looking for in a partner. She calls this her “type.” “You know, white collar,” she says, listing her ideal mate’s attributions. “College grad. Loves dogs. Likes funny movies. Six foot one. Brown hair. Kind eyes. Works in finance but is outdoorsy, you know, on the weekends. I always imagined he’d have a single-syllable name like Matt or John or Dave. In a perfect world, he drives a 4Runner and the only thing he loves more than me is his golden Lab. Oh . . . and a nice smile.”

  We laugh at the scene because of the speed and exactness with which Natalie is able to recite all of these details. But the joke also rings true because all of us have met someone like Natalie. We may even be someone like her.

  In one sense, the idea that selecting the perfect lover is the result of finding the individual whose list of attributes best measures up against our own wish list seems like the most natural thing in the world. Whether we are looking for a spouse, a holiday or a new laptop, all of us make mental lists of the qualities we are searching for and then match up whichever potential offerings we come across with our checklist of minimum demands. If a potential relationship is deemed not attractive enough for us on some level, a holiday is too expensive, or a laptop won’t carry out the tasks we are buying it for, we dismiss it and move on to the next option. However, is this really the right way to think about love? In his book How the Mind Works, the experimental psychologist and author Steven Pinker poses a question very similar to the one asked in the Stable Marriage algorithm described at the start of this chapter: namely, how can a person be sure in a relationship that their partner will not leave them the moment that it is rational to do so? Pinker gives the potentially problematic example of a more physically attractive “10-out-of-10” neighbor moving in next door to us. The answer economists David Gale and Lloyd Shapley would give us is that we are in the clear just so long as this neighbor is already paired with someone more preferable to themselves than our spouse. The answer Pinker instead presents us with is the more humanistic suggestion that we ought not to accept a partner who wants us for any rational reason to begin with—but rather who is committed to staying because of who we are.

  Murmuring that your lover’s looks, earning power, and IQ meet your minimal standards would probably kill the romantic mood, even though the statement is statistically true. The way to a person’s heart is to declare the opposite—that you’re in love because you can’t help it.19

  This antirational view of love is one that may be breaking down in the age of The Formula. In Love in the Time of Algorithms, author Dan Slater relates the story of “Jacob,” a thirtysomething online dater who meets a 22-year-old secretary named Rachel on the Internet. After dating for a few months and moving in together, the couple split up after coming to the conclusion that they want different things from life. So far, so normal. What is different, however, are the comments made by Jacob in the wake of the relationship’s failure. “I’m about 95 percent certain that if I’d met Rachel offline, and if I’d never done online dating, I would’ve married her,” he tells Slater. “At that point in my life I would’ve overlooked everything else and done whatever it took to make things work. Did online dating change my perception of permanence? No doubt. When I sensed the breakup coming, I was okay with it. It didn’t seem like there was going to be much of a mourning period, where you stare at your wall thinking you’re destined to be alone and all that. I was eager to see what else was out there.” In other words, there are plenty more fish in the sea—or, in the words of dating website PlentyofFish, at least “145 million [unique] monthly vi
sitors.”

  Are You Sure You Want to Delete This Relationship?

  Jacob’s words offer a perfect summation of what the Polish sociologist Zygmunt Bauman refers to as “virtual relationships” in his book Liquid Love: On the Frailty of Human Bonds. Categorizing a more scientific approach to love alongside quick fixes, foolproof recipes, all-risk insurance and money-back guarantees, Bauman describes how the ultimate promise of virtual relationships is to “take the waiting out of wanting, [the] sweat out of effort and [the] effort out of results.”

  Unlike “real relationships,” “virtual relationships” are easy to enter and to exit. They look smart and clean, feel easy to use and user-friendly, when compared with the heavy, slow-moving, inert messy “real stuff.” A twenty-eight-year-old man from Bath, interviewed in connection with the rapidly growing popularity of computer dating at the expense of singles bars and lonely heart columns, pointed to one decisive advantage of electronic relation: “You can always press ‘delete.’”20

  It is explanations such as this suggestion of “liquid love” that might mean—in the words of FreeDating website founder Dan Winchester—that the future will be made up of “better relationships but more divorce.” Although this seems a paradoxical statement, it is something that could be the end result of better and better algorithms, Winchester suggests. “I often wonder,” he says, “whether matching you up with great people is getting so efficient, and the process so enjoyable, that marriage will [eventually] become obsolete.”

  What Winchester is expressing is not unique, although it might well be a new phenomenon. In his 2004 book, The Paradox of Choice: Why More Is Less, the American psychologist Barry Schwartz argues that the overwhelming amount of available choice in everything from shopping to, yes, dating has become for many people a source of anxiety in itself.21 In terms of relationships, this “paradox of choice” is dealt with by subjecting individual lovers to segmentation: an industrial term that denotes how efficiency can be gained by dividing up and isolating the means of production. In an age of mass customization, relationships become just any commodity to be shaped according to fads, changing desires and flux-like whims.

  Such a postindustrial approach to dating runs counter to what we have been culturally conditioned to believe. The Lover is supposed to be unique; not just a combination of answers to set questions, each one to be answered correctly or incorrectly. Anyone who has ever scanned through page after page of Internet dating “matches,” however, will soon find themselves, to quote the French novelist Marcel Proust, “unable to isolate and identify . . . the successive phrases, no sooner distinguished than forgotten.”22 Profiles become a seemingly never-ending repetition of terms like “cute,” “fun-loving,” “outgoing,” “romantic” and “adventurous”: each striving to break the formulaic mold of uniformity, but all ending up drawing from the same cultural scripts of desirable characteristics nonetheless. A similarly autistic approach to human relations can come across in the profile-building questions asked by dating websites. “Does money turn you on less or more than thunderstorms?” “What about body piercings versus erotica?” “Would you rather your potential spouse has power or long hair?” Since everything is given equal weighting on an interface level, and so many of the algorithm’s inner workings are obscured, there is no way of knowing just what affects our final score.

  This issue is one that was touched upon in a 2013 article for the Guardian’s “Datablog” (whose tagline reminds us that “Facts are sacred”). Recounting her experiences of algorithmic matchmaking, journalist Amy Webb writes, “When I first started online dating, I was faced with an endless stream of questions. In response, I was blunt, honest, and direct. Then my patience started to wear thin, so I clicked on what I thought sounded good.” Before long, Webb began to second-guess the answers that she was being asked to enter. Certainly, she liked strong men who work with their hands, but was this a veiled attempt to ask whether she would date a lumberjack? After all, “they’re strong and work with their hands,” she says. “But I don’t want to marry a lumberjack. I don’t even like trees that much.”23

  Perhaps even worse is the existential crisis that results from a seemingly objective algorithm determining that it has scoured the Internet and that there is no one out there for you. “About 16 percent of the people that apply to us for membership we don’t allow to participate on our site,” Neil Clark Warren says. “We have seven different reasons for excluding people. If they are depressed—because depression is highly correlated with other pathologies—we don’t let people participate. If they’ve been married more than three times we don’t let them participate, which eliminates 15.5 percent of the marriages in America, which involve at least one person that’s been married three times previously. In addition, we have something called ‘Obstreperousness,’ who are people that you just can’t satisfy. You give them one person and they’ll criticize them for being too assertive. You give them another and they’ll claim they’re too shy. We have ways of measuring that, and we ask that those people don’t continue on the site.”

  Taking the Chance Out of Chance

  Gary Shteyngart’s 2010 novel Super Sad True Love Story is a coming-of-age fable set in a dystopian New York City of the near future. Because of the world’s total informational transparency, no scrap of personal information is kept secret. All a character has to do—as occurs during one scene in which the novel’s bumbling protagonist, Lenny Abramov, visits a Staten Island nightclub with his friends—is to set the “community parameters” of their iPhone-like device to a particular physical space and hit a button. At this point, every aspect of a person’s profile is revealed, including their “fuckability” and “personality” scores (both ranked on a scale of 800), along with their ranked “anal/oral/vaginal” preferences. There is even a recommender system incorporated, so that a user’s history of romantic relationships can be scrutinized for insights in much the same way that a person’s previous orders on Amazon might dictate what they will be interested in next. As one of Abramov’s friends notes, “This girl [has] a long multimedia thing on how her father abused her . . . Like, you’ve dated a lot of abused girls, so it knows you’re into that shit.”24

  The world presented by Super Sad True Love Story is, in many ways, closer than you might think. Several years ago, the Human Dynamics group at MIT created a mobile system designed to sound an alarm if it was within ten yards of an attractive potential date. According to two of the researchers, Nathan Eagle and Alex Pentland, the system was “developed to enable serendipitous interactions between people” and was thus given the name Serendipity. As per the promotional literature supplied by the team:

  In a crowded room you don’t even have to bother working out who takes your fancy. The phone does all that. If it spots another phone with a good match—male or female—the two handsets beep and exchange information using Bluetooth radio technology. The rest is up to you.25

  Apps like Serendipity are part of a new trend in technology called social discovery, which has grown out of social networking. Where social networking is about connecting with people already on your social graph, social discovery is all about meeting new people. There are few better examples in this book of The Formula in action than MIT’s Serendipity system. Here is a problem (“chance”) and a task (“making it more efficient”). Executed correctly, the computer might provide an answer to the question asked by Humphrey Bogart’s character in Casablanca. “Of all the gin joints in all the towns in all the world, why did such-and-such a person walk in to yours?” Because the device in their pocket beeped.

  Naturally, this raises as many questions as it answers. Serendipity is the occurrence of events by chance in such a way that winds up being beneficial to those involved. Equal weighting is placed on both the “random” and “beneficial” nodes, the absence of either one meaning that no matter what occurs, we may say with some certainty that it is no longer serendipitous. The distinction becomes more
pointed when love is involved. After all, for many people the essence of love (if it can be described as such) is the forging of a certain universal value out of pure randomness: the idea that an apparently meaningless, chance encounter brings with it the ultimate meaning.

  One of the best literary analyses of this phenomenon comes in Alain de Botton’s debut novel, On Love, in which the narrator becomes smitten by a woman he meets on a Paris-to-London flight. In an effort to inject a degree of rationality into the irrational, de Botton’s protagonist tries to calculate the odds of such a meeting taking place, eventually coming up with an answer of 1 in 5840.82. “And yet,” he writes, “it happened.”

  The calculation, far from convincing us of the rational arguments, only backed up the mythical interpretation of our fall into love. If the chances behind an event are enormously remote, yet the event occurs nevertheless, may one not be forgiven for invoking a fatalistic explanation? . . . [A] probability of one in 5840.82 [makes it seem] impossible, from within love at least, that [our meeting] could have been anything but fate.26

  Of course, as the late Steve Jobs might have said about fate: “There’s an app for that.” Serendipity’s creators proudly state, “Technology is changing the way we date. For the shy and single, it has been the biggest aid to romance since the creation of the red rose.”

  Wear Your Heart on Your Sleeve

  MIT’s Serendipity project is not the first investigation of its kind. In the late 1990s, a proximity matchmaking device called Lovegety briefly became all the rage in Japan, selling more than 1.3 million units at an approximate price of $21. The aim of Lovegety was to allow users to find potential dates in their vicinity. Users were asked to input responses to several questions, which in turn became their personal list of preferences. When a mutual match was discovered within range, the device alerted both parties to one another’s presence. Compared to eHarmony’s barrage of questions and 29-point compatibility scale, Lovegety’s preference options were admittedly sparse. Even with this being the case, however (and despite the preferences on offer being extremely superficial, such as a shared desire to partake in karaoke), a surprising number of users still described the experience as qualitatively different from that of being approached by a stranger. “When you’re picked up out of the blue, there is always an element of suspicion,” said one female customer. “But when you’re brought together through the Lovegety, you’re more at ease because you already have something in common. You already have something to talk about.”27 In other words, the technology was more than just an invisible mediator between two parties. Like the shared ownership of a Porsche, being part of the Lovegety club meant there was an automatic commonality between both parties. To paraphrase Marshall McLuhan, the medium really was the message.