Outsourcing matchmaking to AI sounds cutting-edge, but there's some irony baked into the origin story here: the person declaring war on swiping is none other than Justin McLeod—the Hinge founder who built his name on the swipe mechanic in the first place. His new app "Overtone" ditches the endless right-swipe interface entirely, letting AI directly curate matches deemed "worth introducing." McLeod's pitch is that the system is "grounded in relationship science and careful judgment." One user on X wasn't buying it: "This is literally arranged marriage, except your parents got swapped out for an AI."
Using algorithms to calculate true love isn't a new idea. Back in the 1960s, a system called TACT (Technical Automated Compatibility Testing) once matched a man with his own biological sister—high similarity clearly doesn't mean compatibility. Half a century later, "compatibility" remains one of the toughest nuts to crack in psychology.
Paul Eastwick, psychology professor at UC Davis and author of "Bonded by Evolution," has spent the past decade using machine learning to study exactly this. He told Dazed: "You can use data to predict who's more popular, who's pickier—but you can't predict who'll click with whom." He didn't hold back on Overtone's "relationship science" claims either: "If 'relationship science' means having people fill out a bunch of questionnaires, then analyzing their preferences and personality traits to match them up—we've tried that countless times, and the results are about as good as flipping a coin." Overtone's spokesperson, for their part, has only said the app is "based on research into how people connect and what makes long-term relationships work," without disclosing the actual mechanism.
Ironically, even knowing the predictions probably won't hold up, Overtone still appeals to people burned out on swiping. Indigo, 28, admits she hates the current dating apps but is willing to give Overtone a shot: both sides provide more information, so matches should theoretically be more deliberate—and more importantly, "you don't have to keep thinking about it," freed from the tyranny of endless options. That temptation—handing over the decision-making and giving your thumb a rest—is itself a product being sold.
But research also shows that people are pretty unreliable judges of their own stated preferences—one study found that 74% of people were willing to date someone who violated two of their supposed "absolute dealbreakers." In other words, we think we know what we want, but reality often plays out differently. This is the most fascinating part of Eastwick's research: compatibility isn't some fixed number—it's something that only gradually emerges through real interaction between two people. "That's exactly why so many people find relationships even when they're not the most obviously appealing person in the room," he says. Whether or not AI ends up doing the introducing, you'll probably still have to walk into that room yourself.
