Everyone knows one. The sharpest person in your orbit — quick, curious, genuinely warm when you get them talking — who has, by any normal accounting, almost no social life.

They decline the party again. They’re not at the group dinner. And when you picture them home on a Friday — deep in a project, a book, a rabbit hole — the standard worry kicks in: shouldn’t somebody check on them?

Here’s the strange answer from one of the most talked-about happiness studies of the past decade: probably not. They may be the one person in your circle whose Friday night is working exactly as designed.

The study that found the exception

In 2016, evolutionary psychologists Norman Li and Satoshi Kanazawa analyzed survey data from about 15,000 young adults, looking at how socializing related to life satisfaction. The headline finding was unsurprising: for most people, more frequent socializing with friends meant higher life satisfaction. Friends are good. Film at eleven.

The interesting part was the exception. The social-time-equals-happiness link grew significantly weaker as intelligence rose — and among the most intelligent participants, it flipped entirely: more frequent socializing was associated with lower life satisfaction. The busier their social calendar, the less satisfied the highly intelligent reported being with their lives.

Honesty about the fine print, because this study gets waved around carelessly: it’s one dataset, it’s correlational, and it measured how often people socialized, not how many friends they had. But the pattern it captured — brilliant people voluntarily running small circles and being fine — is one everybody recognizes, which is exactly why the study went around the world. It put numbers under something the sharp-friend-at-home had never been able to explain without sounding arrogant:

for some minds, the standard dose of socializing isn’t medicine. It’s a bill.

Where the hours were going instead

Why would more brainpower make a full calendar feel like a cost? The researchers offered a few interpretations, and the most practical one has nothing to do with disliking people.

Highly intelligent people, they noted, are disproportionately likely to be consumed by long-horizon projects — the research, the craft, the business, the obsession that takes years to pay off. For a person carrying one of those, every evening is contested territory: the group dinner isn’t competing against loneliness, it’s competing against the thing — and the thing usually wins, because the thing is where that particular mind feels most alive. Frequent socializing, in this reading, registers as an opportunity cost paid in the only currency the project accepts.

Add the texture problem: a wide social circle runs, necessarily, on maintenance-grade interaction — the catch-ups, the small talk, the tenth retelling. Pleasant, low-density, and for a mind that’s happiest at depth, mildly starving. It’s not that the smart friend can’t do those evenings. It’s that they leave them the way you leave a meal that was all bread — technically fed, oddly emptier.

None of which makes them difficult. The same people are frequently superb company one-on-one, at depth, on subjects that matter. They didn’t opt out of people.

They opted out of the format.

Why it intensifies with age — for everyone, actually

Now the title’s second clause — the older they get, the less willing they become — and here’s where the story widens beyond IQ, because the aging part isn’t really an intelligence finding at all.

It’s a universal one. Decades of research on how people spend their social energy across the lifespan shows that as time horizons shorten, virtually everyone becomes ruthlessly selective — pruning the wide pleasant circle, reinvesting in the handful of ties that carry real weight. It’s one of the most reliable patterns in the psychology of aging: the meaningfulness threshold rises with every passing decade, in everyone.

Which reframes the intelligent small-circle person perfectly: they’re not a different species. They’re an early adopter. The audit everyone eventually runs at sixty — which of these hours actually feed me? — they started running at twenty-eight, because running audits on their own experience is rather the whole skill set. Then every year of results confirms the finding, the willingness to fund unfed hours drops a little further, and by fifty the calendar is exactly as small and as dense as the data always said it should be. Age didn’t change their preference.

It vindicated it — and quietly brought everyone else around to the same conclusion, twenty years late.

The fine print that keeps this honest

Two boundaries, before anyone laminates this article as a permission slip.

First: fewer was never none. The broader research on wellbeing is unambiguous that even the most self-sufficient minds need a small number of genuinely close ties — the one or two people who actually know them. The thriving version of the small-circle life has real depth inside the smallness; the failing version is a project and no witnesses, and those are different lives with different endings. The test isn’t circle size. It’s whether anyone’s inside it.

And second: this was never a ranking of people by headcount, in either direction. The extrovert with forty beloved friends isn’t shallow; the study’s whole point is that the same social dose lands differently in different minds. The error was only ever the universal prescription — the assumption that a full calendar is what flourishing looks like for everyone, and that the brilliant friend home on Friday is a problem to be solved.

So, about that friend. Stop scheduling the intervention. Send a different invitation instead — the one-on-one, the long walk, the dinner for two with an actual subject — because that’s the format they’ve been accepting all along, and inside it you’ll get a version of them the group dinner has never once seen.

They were never lonely back there with the project.

By the only measure that matters — theirs — they were at capacity: a small life, densely packed, with nothing in it that doesn’t feed them.