On Friday, Grant Sanderson posted a video to his channel 3Blue1Brown titled “There’s never been a better time to go into math.” By this morning it had passed 1.02 million views and was the most-commented link on Hacker News. Nine days earlier, Stockton University announced 21 faculty and staff layoffs against a $20 million structural deficit.

These two facts are usually presented as a paradox. They aren’t. They’re the same fact, viewed from opposite ends of one ledger.

Sanderson’s argument — that AI will make mathematical research more productive, and that mathematical training retains its value even as machines get better at proofs — is probably correct. What neither he nor the 499 commenters say out loud is that the value of mathematics and the solvency of a mathematics department have come unhooked. The first is climbing. The second is being cut.

The Calculus Sequence Is the Business

A state math department does not run on its graduate seminar. It runs on the 300-seat lecture hall full of engineers, pre-meds, and business majors — the service course, staffed by graduate students paid partly out of the tuition those sections generate. The pure mathematicians upstairs are, financially speaking, a division of the intro sequence. Sell the intro sequence, and you can afford to think about manifolds.

Two things are eating that subsidy simultaneously. Enrollment is falling — Stockton’s deficit is an enrollment story before it is a budget story. And the lecture section is now the most substitutable product in American higher education. A motivated eighteen-year-old can learn single-variable calculus from the best explainer alive, for free, at 1.5x speed, without buying a clicker.

That explainer is Grant Sanderson. I mean this with no disparagement whatsoever: his channel is the finest calculus curriculum ever assembled, and it is, structurally, the competition. Every hour a sophomore spends with an animated epsilon-delta proof is an hour not spent generating credit hours in a building that has a dean attached to it. The profession’s greatest living popularizer is, on the revenue side, its most effective competitor.

“We spent a decade trying to get the failure rate down and the section count up,” a lecturer who has taught the calculus sequence at a Midwestern state university for nine years told me, leaning against the kitchenette counter after a curriculum meeting. “The section count is the problem now.”

The Safe Major Isn’t

The governing assumption in statehouses and on op-ed pages for fifteen years has been that the university’s fat sits in the humanities, and that STEM is where the money should go because STEM is where the jobs are. That assumption is now doing real damage to the people who hold it.

The University of Utah said in September that it could cut up to 15% of its staff ahead of future enrollment declines, describing conditions as among “the most significant challenges in its 176-year history.” That is not a philosophy department problem. When a public flagship with a 176-year run starts talking about fifteen percent of its staff, it is not sorting the useful disciplines from the useless ones. It is repricing everything, because the revenue model underneath all of it — enrollment, multiplied by sections, multiplied by tuition — has stopped clearing.

Conservatives who spent a decade arguing that the academy should be run more like a business are about to find out what that actually looks like. A business does not keep a low-margin division out of sentiment. It looks at which lines of instruction pay for themselves and which ones are carried by the ones that do, and then it cuts the carrier when the carrier stops carrying. Mathematics is not being defunded for ideological reasons. It is being defunded for the reason a chain closes its least profitable store: the traffic stopped.

The Asset Nobody Is Defending

Sanderson makes one observation in the video that deserves more attention than the headline claim: that AI could change how academic contributions receive recognition. Sit with that. The university’s real product was never the proof. Proofs are cheap now. The product was adjudicated credit — the machinery that decides whose name goes on the theorem, which journal accepts it, which committee counts it toward tenure.

That machinery is the thing worth a billion dollars, and it is the thing no one is building a defense of. Math departments spent the last decade arguing for their relevance on the grounds that they teach reasoning and supply industry with quantitative talent. Both claims are true and both are worthless in a budget meeting, because neither explains who pays the light bill.

Here is the awkward part for the profession: the most effective mathematics instruction on earth is funded directly by viewers, not by tuition, not by grants, not by the state legislature. An 8.7-million-subscriber channel with a Patreon link has out-competed every lecture hall in the country at the one thing lecture halls were built to do. That is a remarkable achievement and a catastrophic piece of evidence for anyone testifying before a board of trustees.

Sanderson is right that there has never been a better time to go into math. He said nothing about getting paid to do it, and the deans have already done that arithmetic.

Sources