What the University Is Now For

In 1869, Charles Eliot asked what the American college was for, and answered by dismantling the curriculum that had defined it for two centuries. In 1936, Robert Hutchins asked the same question against the pressure of the Depression, and refused to let the university be reduced to a job-training apparatus for an economy in collapse. In 1963, Clark Kerr asked it again of the postwar multiversity, and named the contradictions the institution had been carrying without naming. As much as we want to think this moment is unique, we stand at a similar inflection. The question is the same. The pressures are different. The hedging that has carried the American university through the last half century is no longer available, and the institution will have to answer, in language its students and their families can recognize as honest, what it is now for.

The pressure this time arrives in the form of a technology that has begun to absorb the credentialed performance the university has been selling. Large language models can pass the bar examination, write the brief, draft the memo, generate the code, summarize the literature, and produce the analysis that an entry-level professional was hired to produce. Companies have been laying off in the tens of thousands and citing AI as the reason, and a year later many of them are quietly rehiring, often offshore and at lower wages. Forrester’s 2026 Future of Work report finds that 55 percent of employers regret those layoffs and projects that roughly half will be reversed. The on-ramps into professional careers are narrowing at the same time that tuition continues to rise. The Burning Glass Institute reports that the value of a bachelor’s degree has dropped to a thirty-year low, with unemployment among recent graduates approaching levels seen among non-graduates, as employers automate the entry-level tasks that once gave junior staff the experience the labor market is now demanding upfront. The families paying that tuition are being asked to take on more debt for a credential whose market value the technology is in the process of revising in real time. The institution is being asked to absorb this revaluation at a moment when the federal research enterprise that has underwritten the modern American university is itself contracting, when state appropriations have been declining in real terms for two decades, and when the discretionary resources required to undertake any serious redesign are scarcer than they have been at any point in the postwar period. The institution has so far responded to all of this with a vocabulary of adaptation. AI literacy across the curriculum. Faculty workshops on integrating the tools. A few new degree programs. The vocabulary is not wrong. It is incommensurate with the size of what is happening.

The numbers are not yet coherent. The forecasts contradict each other. The capability claims are real and also overstated, in different ways, by parties with different incentives. Anyone who has worked inside this technology knows that the picture being painted in the business press is partly accurate and partly a story being told for reasons that have little to do with the technology itself. None of this is evidence that nothing is happening. It is evidence that the institution has been measuring the wrong thing for a long time, and the new measurements are arriving faster than the institution can absorb. What the disruption is bringing into focus is not a future shock but a drift the university has been allowing for at least two generations. The drift is not technological. It is a drift in the institution’s account of itself.

That account has rested on a productive ambiguity. The university has told one story to its trustees, its accreditors, and the public, which is the story of the holistic education, the formation of citizens, the cultivation of judgment, the well-rounded life of the mind. It has told a different story to its students and their families and the labor market, which is the story of the credential, the ticket, the signal, the return on investment. The two stories were never quite compatible. They were held in suspension by an institution wealthy enough, slow enough, and culturally trusted enough that no one had to choose. Faculty believed they were doing the first thing while students were paying for the second. The institution measured itself by the first and collected tuition for the second. The gap between the two was absorbed by prestige, by endowment where it existed, and by a social consensus that the degree was worth what it cost regardless of the mechanism. The well-resourced privates can extend this hedge for another decade. The public access institutions cannot. Their students are not making a mistake when they arrive for transactional reasons. They are reading the labor market correctly. They are paying for a credential because they have been told, accurately, that the credential is what stands between them and economic precarity. AI is the mirror that has shown that both stories cannot be true at the same time, and the institutions that have been telling both will have to choose.

I write this from a particular vantage point. I am a computer scientist who has spent two decades inside the field this technology comes from. My own research advances AI to decode the molecular machinery of life, and I have watched the technology open questions that were not previously askable and accelerate discoveries that matter. I hold both things at once. The same technology that is making it possible to design new therapeutics is restructuring the labor market my students will enter, and the institution they are paying to prepare them has not yet figured out how to be honest about either.

I am also an executive responsible for deploying this technology at the scale of a public research university serving forty thousand students, most of them first generation, many of them the first in their families to take on this much debt for an outcome the market is now revising. I am a teacher who has watched students arrive in my classroom with questions their parents are asking them and that I do not have permission to answer in the official institutional vocabulary. And I am a parent. I have a child preparing to enter the labor market the parents reading this are worried about, and I have the same questions they do. What should she study. What will be there in three years. What will the degree mean. I do not have the luxury of the institutional script, because the script does not work in my own kitchen. The parents who write to me asking which disciplines are safe deserve a more honest answer than the one the institution has been giving them, and I have come to believe that the institution’s inability to give that answer is the problem, not a regrettable feature of an otherwise functional system.

The honest answer is not a list. There is no list. Anyone who gives you a list is selling you something, and the institution that has been gesturing at one has been complicit in a pretense that is no longer sustainable. The truth is simpler and harder. If a job is a task that can be fully digitized, it is done. The tempo at which it is done is not in our hands. It is set by capital, by investment cycles, by regulatory response, by the appetite of firms for the disruption itself, and none of these are forces the family or the institution controls. No one can tell you which careers will be safe in three years or in ten. Anyone who claims to is guessing. The question parents are asking, what should my child major in, has been the wrong question for some time. The institution has been letting them ask it because answering the right one would have required a redesign the institution was unwilling to undertake. The right question is what habits of mind, what kinds of relationships, what capacity for judgment under uncertainty, what tolerance for slow understanding, will let a person remain economically and humanly viable across a working life in which the specific tasks they were trained for will be repeatedly absorbed into machines on a tempo no one can predict. Some disciplines build these capacities better than others. Some build them well in some hands and badly in others. Some programs nominally in safe disciplines build none of them, and some programs in disciplines now considered doomed build them in abundance. The discipline name on the diploma is not the unit of analysis. The formation the student undergoes inside it is. The institution has been silent about which of its programs deliver formation and which deliver only the credential, and that silence is the hedging position, and the hedging position is what is no longer available.

What the institution must now choose is a third thing, something it has not been willing or had reason to name. It is not the holistic education in its inherited form, which has been carried by an aesthetic and a class assumption that no longer fit the students the public university actually serves. It is not the credentialed pathway, which the technology is in the process of hollowing out. It is something the public research university is uniquely positioned to build, because the public research university is the institution in American life most directly responsible for whether the formation that sustains a working life across decades remains available to people who arrive without the means to assemble it on their own. It is the formation of people who can do the work that resists automation, which is not the cognitive work the machines are absorbing but the judgment, the relational capacity, the slow understanding, the tolerance for ambiguity, and the practiced humility that emerge only from sustained encounter with hard things and with other people doing the same. This formation is not a curriculum. It is an institution. It requires faculty who are present to students rather than performing for them, advising relationships that persist over years, mentoring that does not appear in workload calculations, and a tempo of learning that the dashboard cannot capture and that the institution has been quietly disinvesting in for a generation because the disinvestment did not show up in the metrics it was tracking. The redesign is not a redesign of the curriculum. It is a redesign of what the institution treats as its actual product, and a willingness to charge tuition for that product honestly rather than for a credential whose market value is being revised in front of us.

The question I want to leave with the reader is not what the university should do. It is what the country loses if the public research university does not do it. The well-resourced private institutions will protect themselves longer. Their endowments and their prestige will absorb the dissonance for a while. They will not be fine. They will be insulated, and the insulation will run down, and when it does they will face the same question with less time to answer it, because they will have spent the runway extending the hedge rather than redesigning around it. The public access institutions do not have that runway. The reckoning arrives at the public university now. The students who arrived in good faith for a credential the institution sold them are owed an answer, and the answer cannot be that the institution is studying the question. The institution has to choose, in public, what it is for. The hedging is over. Said more plainly, the jig is up. What comes next is a redesign or a slow surrender, and the slow surrender will be paid for by the students who can least afford it.