The Same Everywhere

In short: Homogenization, societies becoming more alike, has always helped people coordinate through shared language, time and standards. What's new is its global scale, its speed and the feedback loops of platforms and AI, which push everything toward the middle. The task is to standardize what connects people and protect what distinguishes them.

You know the coffee shop. Reclaimed wood on the counter, white subway tile, a bare bulb hanging over a hand-lettered menu, a flat white in a ceramic cup with no handle. You've been in it in Grand Rapids. You could walk into it tomorrow in Lisbon, Seoul or Melbourne and not be able to tell which city you were in until you stepped back outside.

The writer Kyle Chayka gave this a name years ago: AirSpace, the frictionless, interchangeable aesthetic that follows the internet into physical places. Once you notice it, you see it everywhere. The same fonts on every startup's homepage. The same beige minimalism on every new apartment building. The same three song structures on every chart. The same logos, redrawn into the same clean sans-serif.

The easy reaction is to call this a new disease of the digital age. I don't think that's quite right. Sameness isn't new. In some ways it's one of the oldest things societies do. What's new is that it now happens to everyone at once.

Why Do Societies Become More Alike?

Every large society has had to make people more alike in order to work at all.

Take language. At the time of the French Revolution, a large share of people living in France didn't speak French. They spoke Breton, Occitan, Basque, Flemish and dozens of local dialects. The historian Eugen Weber described how it took schools, conscription, roads and newspapers most of the following century to turn those people into what his book's title calls Peasants into Frenchmen. A nation needed a shared language to run courts, armies and markets, so it made one.

Take time. Until the 1880s, American towns kept their own local time, set by the sun. Noon in one city was a few minutes off from noon in the next. That was fine until railroads needed timetables that worked across hundreds of miles. In 1883 the railroads agreed on standard time zones, and the country slowly set its clocks to them. Local time disappeared because coordination mattered more.

Take print. The political scientist Benedict Anderson argued that printed books and newspapers helped create nations by letting strangers read the same words in the same form on the same day. Spelling settled. Dialects narrowed. People who would never meet began to imagine themselves as one community.

The pattern repeats through mass production, broadcast radio, television and the shopping mall. Each one made people more alike, and each one made something larger possible: trade, democracy, shared culture, the ability of millions of strangers to cooperate.

So homogenization isn't a bug in how societies work. It's a feature. Coordination needs common ground, and common ground means giving up some difference. Organizations work the same way: they exist to coordinate, and coordination runs on shared language. Societies are the same, just bigger.

What Changed?

If sameness is old, why does it feel so new? Three things shifted.

Scale. Every earlier wave of homogenization happened inside a boundary. France standardized French, but French stayed different from German. American towns adopted standard time, but American culture stayed different from Japanese culture. Sameness grew within each society while variety survived between them. That variety between places was the world's reserve of different ideas. Global platforms remove the boundary. When the same few apps carry the same feeds to billions of people, the homogenization that used to happen within a country now happens across all of them at once, and the variety between places starts to thin out too.

Speed. It took France roughly a century to make its people speak one language. A visual trend can now cross every continent in a few weeks. Local styles used to have generations to resist, adapt and absorb outside influence on their own terms. Now they get a season.

The feedback loop. Earlier forces were mostly pushed from the center: a ministry, a railroad, a broadcaster. Today's is a loop. Recommendation systems show people more of what other people already engaged with. Creators study what performs and make more of it. The system rewards what performs, which teaches everyone to make what performs. Chayka called this world, in a later book, Filterworld. Nobody decided the coffee shop should look that way. It converged, one optimized choice at a time.

The old homogenizers had intentions you could argue with. A loop doesn't. It just keeps finding the middle.

How Does AI Speed Up Homogenization?

Generative AI is the most powerful version of that loop yet.

A language model learns by absorbing an enormous amount of what people have already made, and it produces what is most likely given all of it. That's what makes it useful: the output is reasonable, fluent and safe. It's also what makes it a homogenizer. Ask a model for a tagline, a logo direction, a strategy deck or a short story, and you get something close to the center of everything it has seen. Ask a million people's models the same thing and you get a million variations on that center.

You can already see it. Marketing copy that all sounds like the same friendly expert. Brand identities redrawn into the same geometric minimalism. Cover letters that hiring managers can now spot from the first sentence. None of it is bad. That's the problem. It's competent, and it's everywhere, and it's the same.

Then the loop closes. Researchers have shown that models trained repeatedly on their own output lose the rare cases and drift toward the average. As more of what's published is generated, more of what the next models learn from is generated too. The center gets stronger and the edges get thinner, not because anyone chose that, but because the math rewards the likely over the unusual.

AI can work as a mirror that shows us our own thinking, or as an oracle that hands out answers. Homogenization is what happens when a whole society uses it as an oracle. Everyone asks, everyone gets the probable answer, and the probable answer becomes the only one anyone sees.

Why Does Difference Matter?

None of this means we should fight sameness everywhere. Shared standards are why your phone charger works in another country, why a pilot and a control tower understand each other, why a stranger can read your contract. Some things should be the same. That's what makes cooperation cheap.

The risk is in losing the balance. Variety is how systems adapt. In the first half of the twentieth century, the world's export banana trade ran almost entirely on one variety, the Gros Michel. A fungus wiped it out commercially, and the industry had to switch to the Cavendish, which is now the same kind of monoculture facing the same kind of threat. Cultures and economies work the same way. Odd local ideas, unfashionable methods and stubborn regional styles are where the next useful idea usually comes from. When everyone optimizes toward the same center, there's nothing left at the edges to draw on when the center stops working.

There's a human cost too, and it shows up as longing. The more everything looks the same, the more we notice what doesn't: the specific place, the specific voice, the person who actually knows us. Sameness doesn't create the hunger for real connection. People were lonely long before the feeds. But it makes that hunger harder to ignore, because the generic version of everything is always within reach and never quite enough.

So the question isn't sameness or difference. It's where each belongs.

Standardize what's meant to connect. Protocols, measurements, accessibility rules, the plumbing that lets strangers cooperate. Sameness there is a gift.

Protect what's meant to distinguish. A brand's voice, a city's character, a team's way of thinking, an individual's taste. Sameness there is a slow loss you don't notice until it's gone.

Use the average as a starting point, not a destination. When a model hands you the probable answer, treat it the way you'd treat a first draft from someone who has read everything and lived nothing. Then add what only you know.

Homogenization has always been part of how people build something larger than themselves. What's changed is that it no longer stops at the border, and it no longer needs anyone to decide it. That makes the choice more personal than it used to be. In a world that keeps finding the middle, the people and places that stay specific will be the ones everyone else eventually comes looking for.

Frequently Asked Questions

What is cultural homogenization? The process by which places, products and ideas become more alike. It has long helped societies coordinate, but global platforms and AI now drive it faster and across borders.

What is AirSpace? Writer Kyle Chayka's term for the interchangeable aesthetic, like the same coffee shop design in every city, that spreads from the internet into physical places.

How does AI contribute to homogenization? Language models produce the most likely output given what they've learned, which pulls work toward the average. Models trained repeatedly on generated content can drift further toward the middle.

Is homogenization always bad? No. Shared standards make cooperation cheap. The goal is to standardize what connects people and protect what distinguishes them.

Evan Prout is a strategist in Grand Rapids, Michigan, who writes about how organizations think and decide with AI.