In partnership with
by Cenk Sidar
The middle is disappearing. Not in one place, in all of them at once. Middle income work, centrism in politics, the plain competent middle of almost any craft. I have stopped treating these as separate stories. They are one story.
Start with money, because it is the clearest. The economy is turning into a barbell. Capital sits at one end. Hard service work that pays little sits at the other. The middle is the part that goes.
Be clear about what that middle is. It is a kind of work: accountants, coders, paralegals, marketers, mid level analysts. These are the office jobs that once paid enough to own a home and raise a family on one salary.
That layer is going because value no longer needs it. In the AI era, a handful of people can build a company worth a billion dollars. Huge value now gets made without the people it used to take. The wealth still gets made. It just stops being shared.
The same hollowing hits the work itself, not just the pay. AI takes the average, formulaic middle of a craft first, the passable copy, the stock music, the routine reporting. The truly original becomes more valuable. The reliable professional who was better than average but not the best has the least room left.
What it leaves is a brutal rule. Be the best or be nothing. There is no living left in being good but not great, because a machine does good for free. You are either at the top of your field or competing with software that never sleeps.
I know this from the inside. I grew up in a middle income family, the kind this essay is about, and I was the mediocre middle for a long time during my educations. Never near the top of my class, and in high school closer to the worst grades than the best. A late correction before the national exams got me into a good liberal arts school in Istanbul, though not the best one. Nothing on my record said I would end up where I am.
What changed was not that I got good at everything. In college I found the one thing I truly cared about, the intersection of global affairs and business, and went all in. I got into the top program in my field, SAIS at Johns Hopkins, and everything changed. I climbed out not by being broadly good but on the exuberance of loving what I did. I will not turn that into a slogan, because I had luck and support, and a society cannot assume everyone finds their one thing.
It is a harsh way to run a society, and most of us lose under it. By definition, most people are not the best at what they do, and never were. The old middle gave them a place anyway. A decent accountant, a steady reporter, an average but careful designer could build a real life without being a star, and that mattered more than we admit.
A best or nothing world is also fragile. It piles all the reward on a thin layer at the top and leaves nothing to catch the rest. It runs hot, it burns people out, and it breeds resentment in everyone who did good work and got pushed aside. That anger does not stay quiet. It empties the political center and fills the edges.
You see it most plainly in politics. A secure middle class is what makes a moderate politics possible. When people believe they are slowly getting ahead, they accept compromise, because it still leaves them better off next year. Lose that belief and the patience goes with it.
People who think the system has stopped working for them do not want a moderate fix. They want someone who will break the thing that failed them. Economics is not the whole story, culture and identity push the same way, but a hollow middle is the ground all of it grows in. The populist does not invent the anger. He harvests what the economy already grew.
So three middles are going at once. Middle income work. The middle of quality. The political center. They look like separate problems, argued over by separate experts, but they run on one engine.
The engine is simple. AI rewards the people who work well with it and removes the layer that used to sit in the middle. The ones who pair human judgment with machine power pull ahead. The ones who competed by being solid and reliable fall behind, because solid and reliable is now cheap. When the gap between top and bottom grows, the middle is what you lose.
None of this is a law of nature, though. Daron Acemoglu won the economics Nobel for one stubborn idea: institutions decide whether a society shares its wealth or lets it pool in a few hands. New technology does not spread its gains on its own. The middle is not vanishing because physics demands it, but because our institutions are letting it.
Those institutions are built for the world that is leaving. Careers, pensions, parties, mortgages, the safety net, all of them assume a large, stable middle that absorbs shocks and rewards patience. That middle is thinning everywhere at once, and they keep behaving as if it is still there. That is why they keep getting surprised.
I do not think we get the old middle back by wishing for it. That world is going, the way landed gentry went once education replaced it. The real question is what holds anything together when the middle is gone and only the two ends remain. We would have to build it on purpose, through how we tax, train, and share what the machines produce, rather than wait for the market to refill a middle it is busy emptying. A system that tells most people to be the best or be nothing will not hold for long, and the cost is already showing, in our pay, our work, and our politics.
Cenk Sidar
You've seen the AI demos. Viktor does it without you watching.
The AI tool you tried last quarter waited for a prompt, hallucinated a number, then asked if you'd like a summary.
Viktor opened a PR at 2am, rebased it against main, ran your test suite, and posted a note in #eng: "Two flaky tests in payments service, both pre-existing. Recommended merging after fixing them." Then drafted the customer reply for the support ticket the bug created.
That's 619K autonomous actions per day across 20,000+ teams. Not chat replies. Real work shipped to GitHub, Stripe, Linear, Notion, and 3,000+ other tools, from inside Slack and Microsoft Teams.
You don't supervise him any more than you supervise a senior engineer.
SOC 2 certified. Your data never trains models.
"It's what you probably originally thought AI was going to be when you first heard of it in sci-fi movies." Tyler, CEO.
