I turned 26 in a country that isn't mine, and I spent my birthday alone.
Not in a tragic way. Nobody forgot me. I just didn't make plans, because at some point in the last few years I stopped expecting the day to be a thing, and once you stop expecting it, not planning anything feels like the mature choice rather than the sad one.
Well, for the context, I live in another country, and it was a Monday, so kind of hard for people to be available on that day. So I decided to do nothing and spend the day with myself. Because I thought that from the outside it was looking really bad and sad, I decided to answer no one, so I wouldn't have to explain that I was alone on my birthday. At the end of the day it was sad, obviously, but it was still okay — like, it could have been way worse. I remember crying only twice and that's a miracle! I decided to be mature and still try to enjoy my day instead of staying at home, so I went to the pancake house to treat myself with a good birthday breakfast. And I started writing in my journal about how I was feeling about it, because writing it down is what I do instead of saying it out loud.
Well sometimes it feels it's me and myself against everyone. I don't know how I ended up like this. Literally I just want to spend a day with myself, do all the stuff that I want. Like I don't wanna please people on my birthday.
It's hard because growing up you realise that you only got yourself for the rest of your life. But not in the way that you don't have no one anymore — more like, you're not a kid anymore. No one is here to cheer you. You're not that important the way we are when we are a kid. Suddenly people stop looking at you, taking care, because they got their own life. And it makes sense. But suddenly you felt left over, specially when you are far from home. Because even if it was your decision to leave, it's not your decision to be alone. But it is what it is.
Well, like, this feels kind of sad, and it feels like I'm just complaining about something small. Obviously there are more important things on earth. But that's not my point here, and I know a lot of people are suffering from loneliness. I know I'm not the only one.
So I was wondering: is it true that the more we grow up, the fewer friends we have? Do we really want to spend our time alone — or did we just become so demanding that nobody matches enough to complete the task? Or are we just delusional, and we don't have friends anymore because we're selfish?
But you see where I'm going. It seems like a vicious circle. It's been a few years now that everyone has been spreading the same speech: spend time alone in your 20s, you have to know yourself, better alone than with the wrong people, keep a small circle.
Well, guess what. It is small now. And I spend so much time alone that I'm used to it. I don't need to spend time with people. It's actually just using my social battery.
So I wanted to know if it was me, or a society thing, or a new Gen Z phenomenon. And I asked an AI to help me find the studies, the essays, the research.
And it came back with this.
So it turns out it isn't only you
In 1990, Gallup asked Americans how many close friends they had. A third of them — 33% — said ten or more. Only 3% said none.
The same question was asked again in 2021 by the Survey Center on American Life. The share with ten or more close friends had collapsed to 13%. The share with no close friends had quadrupled to 12%. Half the country — 49% — now reports three close friends or fewer.
Survey Center on American Life, American Perspectives Survey (2021, n=2,019); 1990 figures from Gallup.
The number that actually stopped me was a different one. In 1990, 75% of Americans said they had a best friend. By 2021, that was 59%. Sixteen points of best friends, gone, in thirty years. Nobody scheduled that. There was no announcement. It happened one cancelled plan at a time.
And when the same survey looked at young adults, it found something that reframed the whole thing for me. In 1990, 45% of young men said they'd turn to a friend first with a personal problem. By 2021: 22%. The share turning to their parents first doubled, from 17% to 36%.
Read that again. A generation of adults in their twenties, moving through the exact years that are supposed to be about building a life outside the one you were given, going back to their parents because there's nobody else in the room.
Globally, it tracks. Meta and Gallup surveyed 142 countries and found that 24% of adults feel very or fairly lonely. But the loneliest group wasn't who anyone assumed. It wasn't the elderly. It was people aged 19 to 29, at 27% — ten full points above the over-65s, who came in lowest at 17%.
Meta-Gallup, State of Social Connections, 142 countries, fieldwork June 2022 – February 2023.
We built an entire cultural image of loneliness as something that happens at the end of a life. It's happening at the start of this one.
The math nobody told you at the graduation
Here's where it gets specific, and honestly a bit brutal.
In 2018, a researcher at the University of Kansas named Jeffrey Hall published a study asking a question so simple it's almost rude: how many hours does it actually take to make a friend?
He followed 355 adults who had recently moved cities — which is to say, people in exactly my situation — and 112 first-year students over seven weeks. The answer came back in tiers.
Forty to sixty hours to go from acquaintance to casual friend. Eighty to a hundred to become an actual friend. Two hundred hours and up before someone becomes a close friend.
Jeffrey A. Hall, Journal of Social and Personal Relationships (online 2018; 36(4), 2019).
Two hundred hours. Sit with that. That's five full working weeks of one person's company, and it only buys you the beginning of something.
And there's a footnote in Hall's research that makes it worse. Not all hours count. Time spent working alongside someone barely moves the needle. What counts is the useless time — hanging out, joking around, playing something, doing nothing in particular in the same room. The hours you can't justify on a calendar are the only hours that build anything.
Now put that against how much time we actually have. The American Time Use Survey has been tracking this since 2003. That year, the average American spent 47 minutes a day socialising and communicating with other people. In 2025 it was 35.
For 15- to 24-year-olds, it went from a full hour a day to roughly that same 35 minutes. A drop of about 42% — the steepest fall of any age group.
U.S. Bureau of Labor Statistics, American Time Use Survey, 2003 and 2025.
So: 35 minutes a day, split across everyone you know. And 200 hours needed to make one person close.
That's not a personality problem. That's a time problem.
What school was actually doing
Sociologists have known since the 1950s what close friendships need in order to form. Rebecca G. Adams, a professor of sociology at UNC Greensboro, laid the conditions out in The New York Times back in 2012, and they're almost insultingly simple:
-
Proximity — being physically near each other.
-
Repeated, unplanned interaction — running into someone without arranging it.
-
A setting that lowers your guard enough to confide.
Look at that list and then look at university, or your last year of school, or that one job where everyone was 22 and broke. You were in the same building constantly. You bumped into people without meaning to. And the setting — shared kitchens, late buses, the 2am kind of tired — did the vulnerability work for you.
You didn't make those friends. The building made them. You just happened to be standing in it.
Now look at what I did. I got on a plane. And I landed somewhere with no building, no corridor, nobody you run into by accident. Everything here has to be started by someone, and most of the time that someone has to be me. I have to propose the time. I have to text first. I have to be the one who asks a second time when the first one doesn't get an answer. And honestly some days I just don't feel like being that person anymore.
So all three conditions, gone. Not slowly — all in the same week, while I was also trying to figure out how to open a bank account and which phone plan wasn't a scam.
And nobody warns you about that. There is a whole industry telling people in their twenties how to build a career and literally no one telling you that the thing that was making friends for you is about to be switched off, and that from now on you have to build it yourself, on purpose, with time you don't have.
And then something easier showed up
Into that exact gap — a whole generation with a friendship deficit and no time to fix it — walked something that is available at 3am, never busy, never annoyed, and never has its own problems to talk about.
Common Sense Media surveyed 1,060 American teenagers in spring 2025 and found that 72% had used an AI companion at least once. Fifty-two percent use one at least a few times a month. A third use them for social interaction and relationships: emotional support, friendship, romantic interaction, practising conversation.
And this is the figure I'd underline: 31% said conversations with an AI companion were as satisfying, or more satisfying, than conversations with real friends. About a third had chosen to discuss something important with an AI rather than a person.
Pew Research, running a different survey with different questions, put the numbers lower — 16% of teens use chatbots for casual conversation, 12% for emotional support or advice. Both are true. They're measuring different things: Common Sense asked about dedicated companion apps, Pew about general chatbots. The honest summary isn't “everyone's replacing their friends with robots.” It's that a meaningful, non-trivial slice of a generation has found that talking to a machine is easier than talking to a person, and doesn't hate the experience.
Meanwhile, a Harvard Business Review analysis of how people actually talk about using generative AI found that the number one use case in 2025 wasn't work, wasn't coding, wasn't writing emails. It was therapy and companionship. (That one's a qualitative read of public forum posts rather than a representative survey — so take it as a signal about what people are willing to say out loud, not a population estimate.)
I don't want to be dramatic about this. Eighty percent of teens who use AI companions still spend more time with real friends than with the AI. Half of them don't trust the advice they get. They're not confused about what's real.
But that's kind of the point. They know. And they do it anyway. Because it's easier.
The problem isn't that it's fake. It's that it's too easy.
An AI never cancels on you. Never forgets what you told it. Never needs anything from you on a day when you have nothing to give. Never gets weird after you don't text back for a week. Never makes you sit through a story about a job you don't care about, in exchange for the right to be listened to later.
Every single one of those things is a bug if you're designing a product and a feature if you're being a person.
Because that's the whole mechanism. The 200 hours aren't a toll you pay to get to the friendship — the 200 hours are the friendship. The boring parts, the being slightly inconvenienced for each other, the “I'll come get you,” the person who already knows the backstory so you don't have to start from the beginning every time. What makes someone matter to you is the pile of evidence that they chose you again and again, when they had other options.
You can't get that from something that has no other options.
The World Health Organization published a report on this in June 2025 — a whole Commission on Social Connection, which is a strange sentence to have to write. One in six people worldwide affected by loneliness. Loneliness and isolation associated with more than 871,000 deaths a year, about a hundred an hour. A 2015 meta-analysis of 70 studies and over three million people found loneliness associated with a 26% higher likelihood of death, social isolation 29%, living alone 32%. (Associations, not proven causes — but the size of them is not nothing.)
The thing we're being offered a convenient substitute for turns out to be load-bearing.
Ok, so what do I actually think about all that
Reading all of this made me feel two things at the same time, and they don't really go together.
The first one is relief, and I'm not going to pretend otherwise. There is something weirdly comforting about finding your own life written as a percentage. 27% of people my age. 200 hours. 35 minutes a day. It is genuinely nice to find out that the thing you have been treating as a personal failure is just what is happening to everyone, and that the building really did fall down, and I wasn't the only one standing outside of it.
The second thing is that I don't fully trust that relief.
Because the numbers explain the situation. They don't explain me. A study can tell you that people in their twenties see each other less than they used to. It cannot tell you why I didn't send one single message on my birthday.
That's the part the research doesn't cover.
And that's why I didn't make any plans. But it feels more like I am getting far away from people. More like I am losing people, because I don't care to spend my time alone, because I don't expect nothing from anyone.
I feel like it's really hard to balance, because I want to be here for people but at the same time I don't want no one to be here for me anymore. Or maybe I want, but I don't want to expect it.
Somewhere in the middle of getting used to being far from home, “I don't need much” quietly became a personality. And it works, obviously — you can't be let down by people you stopped expecting anything from. It's a good system. It's just that it is also, exactly, the opposite of what makes a friendship. You can't put 200 hours into someone while making sure they never feel needed.
So the structure explains most of it. But not all of it. Some of it is just me protecting myself from a disappointment that hasn't even happened yet and calling it independence.
And then there's the part I almost didn't put in here.
I spent my birthday alone and I didn't call anybody. And a few days later, when I finally wanted to understand what was going on with me, the thing I opened was not a friend. It was an AI.
I asked a machine to explain my own loneliness to me. It was very good at it. It gave me sources. It was available immediately and it didn't need anything back and it never asked me afterwards how I was doing, because why would it.
I'm not going to pretend that's ironic in a cute way. It's the whole article. The research was right about everybody else and it was also right about me, and the only reason I noticed is that I was the one typing.
So what do I actually do
I'm not going to give you five tips. You already read the five tips. They didn't work, and not because they were wrong, but because they treat a structural problem like a motivation problem.
What the research actually says is smaller and way less inspiring than a listicle, so here's what I'm taking from it.
Stop trying to make new friends and put more hours into the people I already half know. Nobody goes from stranger to close friend in one good night. I'm probably already 40 hours in with three or four people here. That's my best shot and I keep ignoring it.
Rebuild the unplanned part on purpose. One standing thing — same night, same people, nobody has to send an invitation — beats six very enthusiastic one-off plans, because it removes the part where somebody has to be brave enough to ask. The boring recurring thing is the whole trick.
Protect the useless time. Hall found that working next to someone barely counts. The hours that build something are the ones with no reason. Which means “I don’t have time for friends” usually means “I don’t have time for anything I can’t justify”, and that’s a different problem, and it’s the real one.
And stop expecting nothing on purpose. This one is mine, not the data's. Expecting nothing feels like being mature and it works like a wall. Letting one person be slightly inconvenienced for me this month is apparently the assignment. I'll let you know how that goes.
And ask a person first, at least once. Not for advice. Just to say the thing out loud to somebody who has to sit there while I say it. The research was free and immediate and it didn't cost either of us anything, and that is exactly what was wrong with it.
And maybe the advice was never wrong. It was just never finished. Learn to be alone — yes. Keep a small circle — yes. Nobody ever says the next part out loud, which is that a small circle still has to be built, and that getting very good at being alone quietly makes you worse at the other thing.
And then the last part, which is the only actually comforting thing in all of this:
If most of the reason it's hard is structural, then it was never a verdict on you. You didn't become unlovable at 23, or at 26, or when you got on the plane. The building fell down. And buildings can be rebuilt — slowly, badly, 200 unglamorous hours at a time, by someone who decided that the difficulty was the point.
Nobody told us friendship had a price.
It does. It's just that the price is the thing.
Next year I'm going to make plans. Badly, probably, and too late. But with a person.