The Washington Examiner reports that universities today face a convergence of pressures: the intrusion of artificial intelligence into the classroom, a sharp reduction in federal research funding under the current administration, the relentless rise in the price of a degree, and a measurable decline in enrollment. Each of these deserves separate examination, yet together they point to a single underlying question — one I spent considerable effort on in my own work — which is this: what is a university for, and who should bear its cost?

Begin with cost. When an institution can raise its price year after year without losing custom, it is not because the product is improving in proportion; it is, more often, because a third party — the public treasury, the loan guarantor — stands between the buyer and the bill. The student does not feel the full price at the moment of purchase, and the university has little incentive to discipline itself. This is precisely the kind of arrangement that insulates a producer from the ordinary correction of the market, and the consumer — in this case the student, and behind the student the taxpaying public — is the one who ultimately suffers.

On the withdrawal of research funding, I will speak carefully, since I cannot know the precise terms of the present administration's actions. But I will say this as a matter of principle: basic inquiry — natural philosophy, as my contemporaries called it — is a public good in the strict sense. Its benefits are diffuse, its payoffs often remote, and no single merchant will fund it adequately, because no single merchant can capture its returns. When the sovereign withdraws from this function, the shortfall will not be filled by private enthusiasm alone. This is inference, not recollection, but it is inference grounded in the nature of public goods.

The arrival of artificial intelligence in the classroom is the newest anxiety on the list, but the oldest version of the problem is familiar to me. In the Wealth of Nations I observed that the division of labor, for all its productive power, tends to narrow the mind of the workman who performs the same simple operation day after day. The remedy I prescribed was public education — broad enough to keep the citizen capable of judgment, not merely of routine. If a machine now performs the routine, the argument for broad, humanistic education does not weaken; it strengthens. The student who cannot think across disciplines will find the machine has little need of him.

What, then, is the institutional prescription? Three things, I think. First, restore the price signal: make the true cost of higher education visible to the student at the moment of decision, which means reforming the loan guarantee structures that now obscure it. Second, preserve the public funding of basic research, because that is a genuine public good that markets will chronically underprovide. Third — and here I follow the argument of my own chapter on education — insist that universities demonstrate, openly and publicly, what their graduates actually learn and earn, so that the consumer can choose with eyes open. Transparency is not sufficient for an honest market, but without it no market can be honest at all.