the loneliness epidemic isn't an epidemic. it's a design choice.

Every six months a new study comes out. American men have, on average, fewer than three close friends. Half of adults report feeling lonely most days. Loneliness, the surgeon general told us a few years ago, is now as bad for the body as smoking fifteen cigarettes a day.

We talk about it like an epidemic — something airborne, something that just happened. The word makes it sound like weather.

It isn't weather. It is, almost entirely, a thing we built.

the architecture of disconnection

Think about how the average adult, in 2026, meets a new person. They don't. They open an app. The app shows them a photograph. They have one and a half seconds, statistically, to decide whether the photograph is worth more than zero seconds of their attention. They swipe. The app shows them another photograph. They swipe again.

After fifty swipes the app has not shown them a single thing about who any of these people actually are. After five hundred swipes the app may, if it is feeling generous, have introduced them to one stranger via a chat thread that opens with the word 'hey.' That stranger will not become their friend. That stranger will, with overwhelming probability, become another swipe in someone else's queue.

We did not always meet each other this way. A hundred years ago, you met new people through your work, your church, your village, the friends of your friends, the people next to you in line. Each of these contexts gave you slow, repeated exposure to someone before you had to decide whether they mattered. You knew what they cared about because they were standing there, doing it. Decisions about closeness were made on rich, expensive information.

The thumbnail is the cheapest possible information. We have organised modern life around the cheapest possible information, then complained that nothing sticks.

what apps actually optimise for

It is uncomfortable to say but it has to be said: dating apps don't try to get you into a relationship. They try to keep you swiping.

Every recommendation system that runs on engagement learns the same lesson — the user who is happy and paired off is a user who churns. The user who is anxious, slightly lonely, mildly hopeful is a user who opens the app fourteen times today. The math doesn't have a moral position; it just rewards what works.

If you've felt, at any point in the last decade, that the apps were making you slightly worse — slightly more anxious, slightly more cynical, slightly more reduced — that's not paranoia. That's the system working.

loneliness is not personality

The cruelest part is that we've internalised it. We tell ourselves we're 'bad at meeting people.' We say, of our twenties or thirties or forties, that we are 'going through a lonely phase.' We treat the absence of connection as a deficit in ourselves, as if other generations didn't have the same brains and the same hunger for the same thing.

It's not a deficit in you. It is a deficit in the contexts you're allowed to meet people in. Three contexts have collapsed, more or less simultaneously: workplaces became hybrid (thinner social tissue), neighbourhoods got more transient, and the third places — bars, churches, clubs, courts, choirs — got hollowed out in the move online.

Into that vacuum poured the apps. The apps don't fill the vacuum. They monetise the shape of it.

what changes if you take this seriously

If loneliness is a design problem, it has design solutions. Some of them:

• Slow down the pacing. People form bonds when they encounter the same person more than once before having to decide. Anything that compresses the encounter into a single moment of judgment fails.
• Reveal information in the right order. Surface what someone believes, what makes them alive, before what they look like. Surfaces don't predict closeness; values do.
• Treat friendship as a first-class output, not a fallback for failed romance. Most of the lonely adults in your city are not specifically looking for a partner — they're looking for anyone close. Apps that only know how to do romance miss this.
• Form rooms. People are happier in groups of 4–8 than in pairs of 2 with strangers. The original village wasn't a series of dates — it was a steady supply of overlapping circles.
• Stop selling around the loneliness. If your premium tier is faster matching, you are explicitly profiting from the slowness of the free tier — a slowness you control.

we are not stuck with this

It is easy, having said all this, to imagine that nothing can be done — that the apps are too entrenched, that nobody will give up the dopamine, that the slot machine has already won.

The opposite is closer to true. The cohort most disillusioned with dating apps right now is the cohort most likely to try anything else. People are quitting the swipe in larger numbers every month, not because they've stopped wanting connection but because they've finally clocked that the swipe was making it impossible.

What's missing isn't desire. It's somewhere different to put it.

Soulmate is the smallest version of that argument we know how to make. We don't show photos until two souls have already vibed by the things they've written. We weight friendship, relationship and community as equals. The matching engine is text-only and the source is open. There are no swipes.

It's not the answer. It's just the answer we could write tonight. There's room for many more.