Nobody has to work to keep people sorted by class, background, or belief. Sorting is what happens automatically whenever exposure is restricted -- no gatekeeper required. What takes actual, deliberate work is the opposite: building a shelf, a room, or a network wide enough that the sort never gets the chance to happen.

Watson Health's core failure wasn't ambition — it was training a system built for language pattern-matching on clean, hypothetical cancer cases, then marketing it as ready for real patients with real comorbidities and incomplete records. IBM sold the remains in 2022, a fraction of what it spent building it.

A grad student late to class solved two famous unsolved theorems as homework in 1939 because nobody told him they were unsolved. The same mechanism -- no institutional knowledge of what a field has declared closed -- explains an amateur finding new pentagon tilings, an obscure lecturer proving a decades-old prime conjecture, and two New Orleans teenagers publishing a proof mathematicians called structurally impossible.

Most people have to read what a room requires, and the work of doing it correctly is real and measurable. A small number of people don't -- not because of what they wear, but because of what they control. The wardrobe is a symptom. The control is the actual variable.

Louis Vuitton, Hermes, and Bottega Veneta cluster in a small list of money-center metros and, for at least one brand, gate their scarcest product behind a documented purchase history. FUBU, Sean John, and Rocawear built comparable fortunes the opposite way -- mass retail, nationwide, explicitly designed to include rather than exclude. Both proved real belonging can be built. Only one needed a velvet rope to do it.

A tier of retail exists that most of the country has never had reason to know about -- not because it costs too much, but because the stores themselves were never nearby. And for at least one brand, proximity to the store turns out to be the first step in a documented, litigated system for deciding who gets access to what's actually scarce.

Einstein published general relativity in 1915 and predicted starlight would visibly bend near the sun -- a claim nothing on Earth could test until a 1919 solar eclipse proved him right. AI ran the same wait-for-the-instrument experiment twice, on a much longer clock, and almost nobody framed it that way at the time.

Max Weber described what happens to some people inside a large rule-bound structure: they self-rationalize into it, becoming, in his words, a cog in the machinery -- not a flaw, a fit. The real problem was never bureaucracy itself; it's never knowing in advance whether you're one of the people it works for.

Economist Alfred Marshall described it in 1890 without fully explaining it: when an industry concentrates in one place, "the mysteries of the trade become no mysteries... but are as it were in the air." Route 128 had three times Silicon Valley's tech workforce in 1965 -- and lost the lead within a decade anyway.

Benjamin Franklin started a weekly debate club at 21 because he and his friends couldn't individually afford the books to argue from real reading. Pooling their money instead created the first successful lending library in the American colonies in 1731 -- practice creating the demand that built the institution.

The night before Challenger launched, engineer Roger Boisjoly warned NASA the O-rings could fail in cold weather. He was right, and he was overruled anyway. Two weeks later, Richard Feynman dropped the same O-ring material into ice water on live television and squeezed it -- the same fact Boisjoly already had, in a form nobody in the room could argue with.

Real competence has an unglamorous final stage: the steps disappear. Cognitive science has named the pattern twice -- Polanyi's "we know more than we can tell" in 1966, and a 1967 three-stage skill model whose final stage is defined by losing conscious access to how you got there.