Robin Dunbar Never Scrolled: What Happens When Your 'Real' Friend List Hits 2,000

Back in the nineties, British anthropologist Robin Dunbar crunched some numbers on primate brain sizes and landed on a figure that would haunt the internet for decades: 150. That's roughly how many stable relationships a human brain can juggle, he argued. Your Christmas card list, your old regiment, your actual functioning social circle — all hovering around that number.

Then we built apps with unlimited friend counts, and everything Dunbar predicted got put through a blender.

The Ceiling Didn't Break. The Definition Did.

Here's the thing nobody says out loud: your follower count was never a friend count. When someone on OranSNS has 4,000 connections, they haven't mentally catalogued 4,000 people. What they've actually done is split their social world into layers — and this is where it gets interesting.

Dunbar himself described concentric circles: about 5 intimate contacts, 15 close ones, 50 meaningful ones, and 150 total. Platform designers, whether they admit it or not, have been rebuilding those rings in software. Close Friends lists. Inner circles. Private groups inside bigger communities. The vocabulary is different, but the architecture is pure anthropology.

I spoke (informally, at a Seattle tech meetup, over decidedly average coffee) with a product designer who worked on community tools at a mid-size platform. Her take: "We stopped fighting the 150 limit. We designed around it. The feed is for the many, the groups are for the middle ring, and DMs are for the five."

That's the honest answer. Platforms didn't shatter Dunbar's number — they gave every ring its own room.

The Algorithm Changed Who Counts as a Friend

There's a second force at play, though, and it's messier. Algorithmic feeds quietly redefined what a relationship requires. Under the old model, maintaining a friendship meant effort: a call, a coffee, a birthday remembered. In a curated feed, the algorithm does the maintenance for you. Your college roommate's life updates surface automatically. You react, occasionally. Relationship sustained — technically.

Sociologists call these "ambient ties." Weak by definition, but not worthless. Research out of US universities has repeatedly shown that loose ties — the former coworker, the guy from your intramural league — are disproportionately where job leads, new ideas, and opportunities come from. Platforms turned ambient ties from an occasional happy accident into a standing infrastructure.

Is that genuine closeness? No. Is it nothing? Also no. It's a third thing we haven't fully named yet.

Where It Gets Gross — and Where It Doesn't

Let's be honest about the failure modes. Scale without structure produces the worst of both worlds: a thousand shallow connections and zero intimacy, with a feed optimized to keep you scrolling rather than connecting. If your app's business model depends on engagement, ambient ties are a monetization goldmine and real closeness is a rounding error.

But the platforms doing it right share a few tells:

• They make small groups easy. The best social products treat the 10-person group chat as the atomic unit, not the public broadcast.
• They reward showing up, not performing. Features that nudge you toward actual conversation — prompts, check-ins, shared activities — beat one-way clout metrics.
• They respect the rings. Not everything should go to everyone. Granular sharing isn't a privacy gimmick; it's how humans have always operated.

The tools exist. Whether people use them to build real community or just to collect acquaintances like Pokémon is, as always, on us.

So Is Dunbar Dead?

Not exactly. The 150 limit was never about how many people you could know — it was about how many you could maintain. That constraint is cognitive, and no app has repealed it. What's changed is that maintenance got cheaper, delegation to software got easier, and the payoff for a huge weak-tie network got bigger.

Dunbar's number isn't dead. It's been demoted — from a hard ceiling to one layer in a much taller building. The question for the next decade of social platforms isn't how many friends we can handle. It's whether the apps we use make the top floors worth climbing to.