Love in the Age of the Algorithm

The matchmaker became a machine

For most of human history, the people you might love were a function of geography and kin: the village, the congregation, the workplace, the friend of a friend. Within about two decades, that changed. Meeting online went from a faintly embarrassing last resort to one of the most common ways couples form — and for some groups, the most common. A Pew Research Center survey found roughly three in ten U.S. adults have ever used a dating site or app, rising to about half of those under thirty.

What’s new isn’t introduction-at-a-distance — lonely-hearts columns are centuries old. What’s new is that the introduction is now mediated by a ranking system. An algorithm decides whose face you see, in what order, and how often, optimised less for your lasting happiness than for your continued engagement. The matchmaker used to be an aunt who wanted you settled. Now it is a product whose business depends, quietly, on your staying single enough to keep opening the app.

The paradox of the endless menu

Psychologist Barry Schwartz argued in The Paradox of Choice (2004) that beyond a certain point, more options don’t liberate us — they paralyse us. Faced with too many, we postpone deciding, expect more from whatever we pick, and feel the sharper sting of everything we passed up. Schwartz distinguished maximizers, who must find the very best, from satisficers, who choose what is good enough; maximizers, his research suggested, end up less satisfied even when their choices are objectively better.

A dating app is the paradox of choice rendered as a deck of cards that never runs out. When the next profile is always one thumb-flick away, the person in front of you is silently graded against an imaginary better match still in the stack. It can corrode the very thing it promises — the felt sense that this person, chosen, is enough — and replace it with a low hum of “what else is out there.” Abundance, it turns out, is not the same as ease.

Designed to be played, not finished

The swipe is not a neutral gesture. It borrows directly from the psychology of slot machines: a simple repeated action, a variable and unpredictable reward, the little jolt of a match arriving when you least expect it. That intermittent payoff is one of the most habit-forming patterns known to behavioural science, and it is the engine room of the modern dating app. The point of a game is to keep playing; a dating app that worked perfectly would, paradoxically, lose its player to a relationship.

So the experience tilts toward the gathering of matches rather than the meeting of people. Hence dating burnout — the specific exhaustion of effort without arrival, of conversations that fizzle before coffee, of feeling like a product being browsed as much as a person being met. The fatigue is real and widely reported, yet most people return: in the same survey, users described deleting and re-downloading the apps several times a year, a cycle of quitting and relapse the design quietly anticipates.

Falling for people who can’t see you

The algorithm reshaped more than dating; it reshaped longing itself. Sociologists Donald Horton and Richard Wohl coined the term parasocial interaction in 1956 to describe the one-sided intimacy a viewer feels toward a TV personality — an illusion of friendship the performer cannot return. Streaming, short video and the always-on feed have poured jet fuel on that illusion. A creator who speaks to the camera each morning, a streamer you watch for hours, a character you adore — the brain registers the daily, intimate, face-to-face cadence as friendship, even though the relationship runs one way only.

This isn’t a pathology, and it isn’t new — people have loved authors and film stars for generations. But its scale is. For some, parasocial bonds are warm, low-stakes companionship; for others, they quietly substitute for the riskier, reciprocal kind, offering the feeling of connection without the vulnerability of being truly known. The research here is still young and genuinely mixed, and it is worth resisting easy moralising. The honest question is simply whether a given bond is adding to your life or standing in for the relationships that could.

A new grammar of ambiguity

Every era coins words for how it loves, and this one minted a vocabulary of the in-between. A situationship is a romantic or sexual bond that deliberately resists a label. A soft launch is hinting at a new partner online — a sleeve, a shadow, a second coffee cup — before naming them. Breadcrumbing is feeding someone just enough attention to keep them interested with no intention of more. And ghosting is the now-ordinary act of ending things by simply vanishing — replies stop, and no explanation comes.

It’s tempting to read this lexicon as proof that people have grown colder, but that’s too neat. The ambiguity is partly structural: when the next option is always a swipe away, defining anything feels like a cost, and disappearing carries no social price because you may never cross paths again. The words are not the disease; they are an honest map of a landscape the technology built. And a map can be used to navigate toward clarity — to notice you’re being breadcrumbed, to ask for the conversation, to refuse to dissolve out of someone’s life without a word.

The data, and the next machine

Before the verdict, two findings that cut against the gloom. The first: for all the burnout, the apps quietly won. By 2017, meeting online had become the single most common way American couples meet — about 39%, up from 22% in 2009 — overtaking friends, family, school and church. The sociologist Michael Rosenfeld called it disintermediation: the screen quietly fired the aunt, the roommate, the colleague who used to make the introduction. The second is stranger still. Despite the cliché that app-love runs shallow, a large 2013 study found that marriages which began online were slightly less likely to break up, and slightly more satisfying, than those that began offline — a small gap, but pointing the opposite way from the doom. The medium feels corrosive and performs, on average, just fine.

The harder question is what replaces it. The swipe may be peaking — younger users report the deepest fatigue, and the apps are shedding paying customers — and the industry’s answer is to add more machine, not less. AI already drafts the opening line and coaches the hesitant; the next step is agentic. Bumble’s founder Whitney Wolfe Herd has floated a near future in which your AI “concierge” goes on dates with other people’s concierges, scanning the city and handing you a shortlist of three. Whatever you make of it, it sharpens this guide’s central question to a point. The first algorithm sorted the faces. The next one wants to write the words, screen the strangers, perhaps feel the first flickers on your behalf. The thing worth guarding is the part no concierge can outsource: the actual, unscalable risk of letting another person matter.

Both things are true at once

The fair verdict is a paradox, and it should be held as one. Apps genuinely widen the pool — a gift to anyone whose village was small, whose identity made the local market hostile, who is shy, or disabled, or starting over far from home. They have built real marriages and real families that would never otherwise have begun. Online daters themselves are split rather than damning: in Pew’s data about half call their experience positive — though that average hides a sharp gender gap, with women, especially younger women, far more likely to report harassment and unwanted explicit messages.

So the same tool widens the pool and isolates at once. It can be the bridge to a person you’d never have met, and the treadmill that keeps you from meeting anyone. The deciding variable is rarely the app and almost always the user’s intention: whether you treat the screen as a doorway you walk through into real, embodied attention, or as a room you live in. The algorithm sorts the faces. It cannot do the loving. That part is still, mercifully, yours.

Threads to

If the loneliness underneath the scroll is what you feel, read the loneliness epidemic next; to understand why some people ghost and others cling, the guide to attachment is the key. To sit honestly with the in-between, see the situationship in the Atlas, and the wider vocabulary lives in the Lexicon. To practise the opposite of the swipe, run the 36 Questions or let Ariadne point the way; if you’re early in all this, walk the Newly Curious path.

The structural conflict at the heart of matching

Dating apps are sometimes described as though they share users’ goal of helping them find a partner and leave. The economic structure of how they make money suggests something different. The US dating app market was valued at approximately $3.2 billion in 2022, with the largest revenue stream across most platforms being ongoing subscription fees. A user who finds a committed relationship and deletes the app is a churned subscription. A user who remains on the platform — cycling through matches, upgrading to see who liked them, paying for location boosts — is a retained subscriber. The platforms are, structurally, better served by extended searching than by successful ending. This is not conspiracy; it is business model. And it is worth naming plainly because it means the product is designed for a goal that is only partially aligned with yours.

A 2018 study by Elizabeth Bruch and M. E. J. Newman in Science Advances analyzed messaging patterns across four US cities using a major dating platform’s anonymized data. It found that the desirability hierarchies that emerge in online dating markets are steep and self-reinforcing: most messages go to a small number of highly-rated profiles, producing a competitive dynamic that keeps the majority of users in perpetual pursuit of partners who are, by the platform’s own engagement metrics, out of their range. The paper did not conclude that the platforms created this hierarchy intentionally — but the hierarchy benefits retention. A user who consistently receives responses from people they are interested in does not stay long. A user who is perpetually almost-succeeding — occasional responses, occasional dates, occasional promising conversations — is algorithmically, economically, and psychologically kept.

The future-facing version of this problem is escalating. As AI moves from writing your opening messages to engaging in preliminary conversations on your behalf, the layer of authentic interaction that precedes actual human contact thins further. The question that structural analysis raises is not whether any individual app is malicious — it is whether a technology ecosystem organized around monetized attention and retention is well-suited to the human project of genuine intimacy, which requires presence, discontinuity, and the kind of irreversible commitment that exits systems rather than staying in them. The machine will get better at keeping you. The exit, increasingly, requires a deliberate act.

Why romantics ghost: the destiny paradox

Ghosting — ending a connection by simply stopping all contact without explanation — is often understood as a failure of courage or empathy. Research adds a more unsettling mechanism: it tends to be perpetrated by the people with the most romantic ideals, not the least. Those who believe most strongly in soulmates, in meant-to-be connections, in fated matches — what psychologists call destiny beliefs — are also the most likely to find ghosting acceptable and to have actually used it.

Freedman, Powell, Le and Williams (2019) studied this paradox across two preregistered studies involving over 1,300 participants. Stronger destiny beliefs predicted more positive attitudes toward ghosting, stronger ghosting intentions, and prior ghosting behaviour — including in long-term relationships, not just early-stage connections. The mechanism is cognitive: if a relationship is either fated or not, its failure is equally fated, and a clean disappearance can feel like an appropriate verdict rather than a failure of care.

For the digital dating context, this matters structurally. Matching platforms are built on destiny aesthetics — the language of “chemistry,” “sparks,” and a curated other half — which may prime exactly the belief system most associated with contact abandonment. The most romantically attuned users may be simultaneously the most enthusiastic adopters and the most likely to vanish. The practical implication is grim: presenting yourself as a soulmate-seeker may attract partners whose exit strategy is invisible.