The Gate That Walked Away

———

The college is not in crisis because artificial intelligence arrived. The college is in crisis because the institution that used to catch what the college could not finish — the first employer — quietly stopped catching, somewhere between 2000 and 2020, and nobody adjusted the upstream chain. The credential survived. The formation did not. AI did not break the sequence. AI made the break visible. The fourth gate walked away, and now the third gate is being asked to deliver a finished worker it was never built to finish. That’s the argument.

Adam Harris, writing in The Atlantic this week, names the symptom precisely. Tuition looks like a dubious investment. Enrollments are softening. The eighteen-year-old cliff is two years away. Harris reaches for the standard diagnosis — colleges are in identity crisis — and he is correct that the institution is breaking. But the diagnosis is shallow. The break is not in the college. The break is downstream of the college, and the college is only the most visible casualty.

To see this clearly, you have to hold all four gates of formation in view at once.

———

From 1989 to 1991, I was a project engineer in General Motors’ Corporate Maintenance Management department, inside the Advanced Manufacturing Group. The work was road shows. We traveled to plants, paired up with the skilled trades — millwrights, electricians, pipefitters — and transformed plant maintenance from reactive (fix it when it breaks) to planned (catch it before it breaks). It was Japanese-pressure-era work. The Big Three were finally learning what Toyota had known for two decades, and we were the ones standing in front of tradesmen who had been pulling overnight breakdown shifts for thirty years, telling them their job was about to change.

Halfway through that assignment, I was halfway through a PhD at Pittsburgh. GM put me on half-time work at full-time pay so I could drive Detroit to Pittsburgh and finish the last six months to my defense. No contract required this. The company had no claim on the doctorate. I could have walked the day I defended. They did it anyway, because that was what a first employer did in 1989. The gate was operating at full strength: real work with skilled trades, real stakes in real plants, real apprenticeship on the floor, real wage — and the gate was willing to fund formation past the point of immediate return.

Eleven years later, Jayanthi and I would be standing on the same line at Lansing Grand River — she the quality engineer, I the trim shift leader — when the plant won J.D. Power Gold during the Cadillac CTS launch. Same gate. Still working.

That gate has since walked away. Not all at once, and not everywhere, but enough to change the chain.

———

The Four Formations are sequential, and each gate inherits the deficits of the gates before it.

The first gate is Home — where order, attention, and the small disciplines of daily life are deposited. The Germans call it Ordnung. In incomplete homes it does not happen, and the second gate has to make it up.

The second gate is School. Through age sixteen or eighteen, the school is meant to transfer literacy and numeracy and to finish the formation work the home left undone. The American school has been recruited into a single track — college preparation — and so it cannot do the second job. Anyone not bound for the third gate is treated as a residual.

The third gate is College-Vocation. The American system narrows further: vocation is the abandoned half of the gate’s name. What remains is the college, and the college is, by its own honest accounting, a knowledge-transfer institution. It can prepare. It cannot complete.

The fourth gate is First Employer. This is the only gate that combines real work, real stakes, real apprenticeship, and a real wage at the same time. It is the only gate that can finish a worker. Lukas, the young German tradesman I have written about before, spent his first five years out of his Ausbildung at Siemens. By age thirty he had roughly twenty-five thousand hours of formed work behind him — the gold standard not because Germany is special, but because all four gates were intact and in sequence.

———

What Harris is seeing in the American college is the consequence of the fourth gate walking away.

When the first employer was still forming workers — when GM in 1989 paid a project engineer’s full salary for half-time work so he could finish a PhD, when Honda still ran Waigaya, when Siemens still ran a five-year first-employer track — the college could afford to be a sorting and credentialing mechanism. It could hand forward a partially-finished product because the fourth gate was going to finish the work and then some. The sequence Harris names — you go to college, you graduate, you get a good job — was never really about the college. It was about the first employer finishing what the prior three gates left undone.

When the fourth gate began to defect — when first employers stopped apprenticing, stopped catching new graduates and finishing them, started extracting from already-formed workers brought in from elsewhere — the chain began to collapse upstream. The college was the first institution to look broken, because the college was the last institution before the fourth gate. But the college did not change. What changed is that there was no longer a formation institution waiting for the credential to be cashed in.

AI did not cause this. AI accelerated it. The entry-level white-collar work that used to be the first employer’s apprenticeship terrain — the analyst role, the associate role, the junior engineer role — is exactly the work AI tools can now do at a tenth of the wage. The first employer, already drifting toward extraction, now has economic permission to finis