Internet dating is all the rage and there are many apps out there that claim to help us hook up and/or find love based on more (or less) information about ourselves. But with more people than ever hunting for long-term relationships it seems that, if anything, lasting love is more elusive than ever! With this come worrying reports about the negative effects of computer-mediated communication, that it may be detrimental to our ability to communicate and that it foments feelings of dissatisfaction, envy and loneliness. Dr. Steve Carter is chief scientist at the world’s top dating site eHarmony, where he leads a team of data analysts and engineers. Speaking at London’s Institute of Art and Ideas, he admits that there are serious problems in society associated with virtual relationships. However, he says, serious dating apps such as eHarmony are helpful rather than harmful, in that they address a specific social problem that has always existed:
Steve Carter (American accent): What we’ve set out to do is to take a problem that people are demonstratively not good at solving, which is the long-term prediction of who they’re compatible with and, by taking data that we’ve collected, using best practices from a social-psychological-scientific standpoint, we’ve created models that provide a framework for solving that problem. So that when people come online and are solving the short-term problem, which is: ‘Do I find someone attractive? Do I have something to talk about with this person? Can I get them to come out and meet me?’ That those interactions lead to relationships that have a higher probability of working out over the long term. That is social engineering to the extent that it’s successful and we believe it has been successful.
IVANKA TRUMP
But does ‘dating by data’ take into account the complexity of human beings and the irrational nature of love? Carter suggests that we are more predictable than we would like to believe:
Steve Carter: You can always show how there’s an exception to any rule, but the fact of the matter is that while humans are very, very unique and very, very subjective and very, very chaotic, at the same time, you can predict better things for them to do often. Or you can predict what they’re going to do next, with enough frequency of correctness that you can create optimisations, so that we can know: ‘You’re more likely to buy another pair of shoes from me today if I show you a Jimmy Choo pair of shoes than an Ivanka Trump pair of shoes.’ And, in the same case, we’ve taken outcomes observed in long-term relationships and created models that do optimise the probability that people are going to have success.
DON’T DO IT
eHarmony was founded on extensive social and psychological research. Surprisingly, though, it was that into failed relationships that proved the most useful:
Steve Carter: eHarmony was a product founded by a psychologist. He had spent 25 years working with people who were in marriages that were thought unsuccessful enough that they’d sought therapeutic interventions, they’d sought help. And after a lifetime doing that he’d come to the conclusion that most of these problems were caused by the people marrying someone that they weren’t compatible with to begin with. That he could help them with communication strategies and with trying to reframe things... but that in reality that if he could have kept them from getting married to each other in many cases he would have been much more successful at finding a happy marriage for them.
PSYCHOMETRIC ROMANCE
And, says Carter, while you can optimise love by connecting with a pool of eligible partners, you can’t engineer it:
Steve Carter: I’m not trying to define love here. The question was, can you use that data that people are reporting about themselves and what’s important to them and how they feel to create psychometric profiles and then see that those psychometric profiles, and the way they interact with each other, predict which of those couples are in happy, satisfied relationships and which aren’t. And it turns out that you can statistically do that quite successfully. So then eHarmony was built around the idea of taking that statistical remodelling and turning it into a product that would then normatively or predictively say: ‘Here are people that you should talk to. If you’re going to fall in love with them, fantastic.’
UNIVERSE OF ONE
But if your desired relationship doesn’t work out then who is to blame? Carter says that people rarely blame the technology. When they do have issues, they want more command over the selection procedure – something discouraged at eHarmony.
Steve Carter: I’ve never heard anyone blame their calculator for solving a math problem wrong or blame their car for driving into a curb. I have spoken to a lot of people who’ve used eHarmony and both been successful, met their spouses, gotten married or not been successful, and they don’t blame the algorithms. They blame lots of other things, things that they want in addition: people want to have more subjective control by buttons and knobs: choose people based on their income or their weight, that you try to keep them from doing because they paint themselves into a corner very quickly and end up in a universe of one! But I’ve never had anyone blame an algorithm for their relationship outcome.
LEAP OFFLINE
And what about the interactive future of tech dating? Carter believes that the confusion of virtual and physical is dangerous:
Steve Carter: I think there’s a real danger of this. There is [are] a large number of companies in the world today who are in the act of trying to create products that are going to let you interact with a human being who’s not in your presence. I don’t think this is a good direction to go. We see lots of reasons why lack of face-to-face human contact is bad. We need to create tools that are enhancing face to face and that our emphasis at eHarmony is to promote that idea. We’re trying to get people to interact successfully online but to interact quickly online and then leap offline.