“‘Now you’ve got nothing left that’s private, nothing that’s yours and yours alone. Centillion owns all of you. You don’t even know who you are anymore. You buy what Centillion wants you to buy; you read what Centillion suggests you read; you date who Centillion thinks you should date. But are you really happy?’
“The Perfect Match” by Ken Liu
‘That’s an outdated way to look at it. Everything Tilly suggests to me has been scientifically proven to fit my taste profile, to be something I’d like.’”
Under the romantic snow flurries of a New Haven Valentine’s Day, Datamatch’s algorithm matched Yale students to either their true loves, most compatible friends, or someone else who signed up just for “shits and giggles.”
Datamatch isn’t quite Tilly, the all-knowing, match-making algorithm that controls everyone’s lives in Ken Liu’s short-story “The Perfect Match.” It is, however, a reflection of the new role technology plays in our social lives, especially in the midst of a global pandemic that has inhibited in-person social interactions.
All algorithms are reflections of our desires: We want cars that drive themselves, predictable ways to make money in the stock market, and people to love who will love us back. But Ed Finn, founding director of the Center for Science and the Imagination (CSI) at Arizona State University, considers the allure of algorithms on a more fundamental level.
“There is a desire embedded in the whole idea of the algorithm itself, which is to make the whole world computable,” he shared in an interview with The Politic.
“We desperately want the answers, and we want to believe in the godlike, rational power of the machine,” he continued. “We end up putting our faith in the machine [to a degree] that we don’t even trust ourselves [nor do we] trust other humans.”
Because of this desire to believe they can provide what humans cannot, algorithms are powerful—even when, mathematically, they aren’t.
Datamatch doesn’t need to become more accurate or feel more threatening for us to confront the underlying question of algorithms’ role in social relationships. With Datamatch, as with other algorithmic programs, it’s not truly about whether we’ll create a more accurate version, or even if we should. It’s about why we want to have a match-making algorithm, if we actually do want to have it, and what that means for our relationships with each other.
Finn founded CSI over a decade ago to be a space for science fiction writers, artists, scientists, engineers, and researchers to “come up with technically grounded, optimistic visions of the future [and] to start working together to create more inspiring visions.”
Shaping our future is a concern in any era, but it is especially pressing in today’s environment—where accessible and advancing technology accelerates innovation and grants any individual the power to affect the future. Everything, from casual student-made dating apps to government-funded algorithms, is reflective of what we want and what kind of future we are creating.
***
“…Ridiculous… Just the kind of pseudo-intellectual anti-technology rant that people like her mistake for profundity.”
“The Perfect Match” by Ken Liu
It may seem “pseudo-intellectual” to draw connections between Datamatch and dystopian sci-fi stories, as if one might become the other. To the Yalie opening their email on an ordinary Sunday evening, Datamatch can seem more like a meme than a transformative cornerstone of change. You fill out a questionnaire, choosing which Yale library best represents you or describing your Wednesday night, and you don’t expect anything serious to come from it.
Alice Mao ’24 filled out her survey with this level of skepticism and levity. “I didn’t really expect to find a soulmate there or anything. If I did actually really get along with somebody, then that would be like a bonus,” she said.
Datamatch does not feel imposingly powerful because, unlike Liu’s fictional algorithm Tilly, Datamatch is just a fallible, inconsequential pastime that is neither all-knowing nor all-powerful. Mao explained, “I can’t imagine it being a long-term dating option. The charm is its novelty.”
“The Perfect Match,” like any other science fiction story, does not intend to predict the future or even depict a possible future. Rather, these stories give readers and writers a playground and vocabulary to imagine solutions, problems, changes, and reactions.
Science fiction is a tool, Finn told The Politic, that is “uniquely positioned in this feedback loop between technological progress and our creative imagination.”
“Imagination,” he said, “is something that we tend to kind of ignore or take for granted until it goes missing. And then we lament failures of imagination.”
In the midst of a pandemic and constant technological innovation, we are embroiled in a formative time of transformation, with our society wrestling with science, technology, and social change. Through sci-fi, Finn says, we can explore questions of “our hopes and fears about what we actually want to happen.”
***
“As predicted, it turned out they were into the same books, the same movies, the same music. They had compatible ideas about how hard one should work. They laughed at each other’s jokes. They fed off each other’s energy.
The Perfect Match” by Ken Liu
Four billion women on Earth, and Tilly seemed to have found the perfect match for him.”
What we actually want to happen when it comes to love and relationships is not a simple question. We consider emotions and social connections to be the very qualities that separate us from machines. And yet we also look to machines to guide us through romantic relationships.
Margaret Clark, a professor of psychology at Yale, wrote to The Politic, “People are not pieces of a puzzle in which only two pieces fit together just right. Rather, people come together, interact with one another, and the nature and feel of the relationships lies in the nature of that interaction. People adapt to one another and change as they do.”
Dating—which has never been simple to begin with—is made even more complicated with the rise of the modern internet. “The availability of dating sites,” Clark noted, “has widened the pool of possible partners by a lot.” However, while a large proportion of romantic relationships in recent years have begun online, she emphasizes that “immediate, physical proximity still plays an important role in whom we meet and form relationships with.”
In the past year, we have entered into somewhat of a social contradiction: Technology furnishes us with an expanded pool of people to interact with, but COVID-19 simultaneously depletes our relationships of the critical component of human interaction. Since last March, organic ways of falling in love have become not only inconvenient, but dangerous.
When it is difficult to navigate relationships ourselves, putting our faith in a system that can supposedly compute the correct answer is a seductive idea.
This algorithmic view, however, assumes something that Clark doesn’t believe is true: that a perfect match exists, that “there really is one person or one precise type of person (with characteristics 1, 2, 3, 4, etc.) who is one’s soulmate.”
“I just don’t think the soulmate idea makes sense. I think a wide variety of different types of pairings might be equally satisfying or good (or equally rotten),” she concluded.
***
“Although everything had gone exceedingly well, if he was being completely honest with himself, it wasn’t quite as exciting and lovely as he had expected. Everything was indeed going smoothly, but maybe just a tad too smoothly. It was as if they already knew everything there was to know about each other. There were no surprises, no thrill of finding the truly new.”
“The Perfect Match” by Ken Liu
Even if we had the capacity to create an algorithm with the accuracy of science fiction machines, do people actually want what algorithms for relationships promise? As American sociologist Edward O. Wilson once said, “the real problem of humanity is the following: we have paleolithic emotions; medieval institutions; and god-like technology.” Perhaps we create technology before we truly understand our own desires.
Clark’s lab conducts research on how emotional relationships unfold, and how people pursue the relationships they want. Self-promotion to a potential partner and self-protection from possible rejection are key components to how a person initiates a relationship. An algorithm promising an omnipotent perspective impacts this natural deliberation. Clark explained, “Your interest in another person may not be reciprocated and it’s painful when that occurs. Believing that [the] use of an algorithm can prevent that might be reassuring to some people.”
At the same time, some want the romanticized pain that heartbreaks, breakups, and unrequited loves bring—it can be pitfalls that make successful relationships all the more special. According to Meghan Laslocky, author of The Little Book of Heartbreak, “the pain is there to teach us something. It focuses our attention on significant social events and forces us to learn, correct, avoid, and move on.” Other research supports the idea that breakups are seminal moments of growth.
Even for someone who wants to find the perfect relationship right away, Clark emphasized that a “fit” between partners is not binary. Partners adapt as they interact, and are influenced both by one another but external factors as well.
In attempting to mathematize love, we overestimate what the algorithm can provide and underestimate what human variables can do. People and relationships are ever-changing, making it impossible for an algorithm to capture enough information to determine a perfect match. When we expect an algorithm to give us both certainty and fluidity, we are, as Finn put it, “[bending] over backwards to make them seem more magical and powerful than they really are.”
***
“Having Tilly around was like having the world’s best assistant…”
“The Perfect Match” by Ken Liu
“‘You see? Without Tilly, you can’t do your job, you can’t remember your life, you can’t even call your mother. We are now a race of cyborgs. We long ago began to spread our minds into the electronic realm, and it is no longer possible to squeeze all of ourselves back into our brains.’”
Algorithms are ingrained in nearly every aspect of our lives—search engines, social media, navigation tools, ads, college admissions, insurance, and even jail time. The image of dangerous algorithms taking over the world looks more mundane in reality than in novels, but it is no less significant.
Since Ada Lovelace created the first computer algorithm in 1843 to calculate the Bernoulli numbers, humans have worked to expand the reach of algorithms. Finn’s 2017 book What Algorithms Want explains the advent and growth of algorithms in human society. The word “algorithm” and its modern meaning is founded on the notion of “effective computability”—when a question is solvable and has a knowable answer. If an outcome is effectively computable, the only thing to worry about is finding the answers, or creating algorithms with the computing power to find them. When Alan Turing, Alonzo Church, and other mathematicians first came up with algorithmic proofs in the mid-20th century, people were startled by how few questions were actually “solvable.”
“From that time to this,” Finn said, “what humans have worked to do is continually expand the space of effective computability, so that it now includes things like driving cars and piloting airplanes and telling you who to date.”
We are always looking to expand our space of effective computability, because we want to know more about our world. However, having complex algorithms provide answers grants us an additional advantage as well: outsourcing decision-making, and thereby outsourcing culpability.
But devolving responsibility catches up to us. We still need to grapple with the implications of algorithms, romantic or otherwise, and determine who is accountable for the answers they compute, especially when those answers may encroach on our data and privacy or perpetuate biases and power imbalances.
***
“Everything Centillion did was arguably legal. The wireless transmissions were floating in public space, for example, so there was no violation of privacy. And the end user agreement could be read to allow everything Centillion did to ‘make things better’ for you.”
“The Perfect Match” by Ken Liu
The year 2021 is set to be a critical point for AI governance. Major developments are necessary: current regulation of algorithms in America is haphazard at best. Developers argue that their algorithms are neutral; they only become harmful because of biased data or improper use by consumers, but users claim no responsibility because they don’t understand the complicated, inscrutable workings of the technology they use.
Datamatch imagines the possibility of reinterpreting the nature of human relationships using technology. “The Perfect Match” imagines how people react—positively and negatively—in a world that has successfully reinterpreted that human-technology relationship.
Finn sees science fiction as a powerful equalizer:
“The everyday general public, a roboticist, a poet, can all read the same science fiction story and have a really helpful discussion about our future and the technological and social possibilities of that future, without necessarily having to get graduate degrees or do a lot of research in preparation,” he explained. “Setting out these possible futures as a shared space that leverages different kinds of expertise is really, really important as a way to make these futures more inclusive.”
The technology we create in the present will shape the world we experience in the future. Before implementing technology that impacts our social interactions on a near molecular level, we need to answer questions about what we want from our social relationships. We need to imagine our future as we play with these “inconsequential” algorithms today.
Datamatch isn’t a slippery-slope into an oppressive, controlling, algorithm-defined society, but it is an example of our desire to extend algorithms into social, emotional terrain. It is built into the idea of an algorithm that with its use, we can have a more straightforward world—one with more objective answers. But algorithms affect real people, and as we push them into new domains, they are forced to make decisions on increasingly intricate things. We make algorithms to give us answers, but behind every algorithm created to answer one question is a rabbit hole of deeper questions that algorithms cannot shoulder for us. When we want machines to “effectively compute” the decisions of our lives, questions about human desires and biases become all the more pressing to answer.
“‘I never really thought of you as my type,’ she said.
“The Perfect Match” by Ken Liu
Sai’s heart sank like a stone.
‘But who thinks only in terms of ‘types’ except Tilly?’ she said quickly, then smiled and pulled him closer.”