American higher education is a bundle of incompatible activities held together by custom and accident, and three successive crises-the COVID-19 pandemic, the arrival of generative AI, and the second Trump administration's research-funding cuts-have exposed that the bundle makes no sense. The combined effect has not created a new problem so much as revealed a structural truth: undergraduate credentialing, coming-of-age rituals, and scientific research are separate enterprises that no longer need to occupy the same campus, even if the institutions are too entrenched to separate.
Artificial intelligence arrived on campus as a crisis but has since become banal. Students can now complete nearly all their schoolwork automatically-solving scientific problem sets, writing computer code, and drafting essays. Faculty have no reliable way to know whether they are teaching or performing for students who are also performing their role. Some professors offload the drudgery of writing recommendation letters, grant proposals, and even scientific papers to AI. At the same time, the technology proved enormously useful for knowledge production in the natural and applied sciences, including protein-structure prediction, materials design, and drug discovery. Pressured by donors, boards, employers, and parents, administrators announced AI initiatives, invested in AI infrastructure, and established AI degree programs.
These two forces-students using AI to their detriment and colleges using it to their benefit-might seem to be in conflict. Campuses are defending themselves against AI as an interloper while retooling for an AI future. But the deeper damage predates AI. The technology combined with the pandemic and the funding cuts to amplify a truth that has been building for decades.
The two promises that never really fit together
College in America makes two promises. First, it is a credentialing service: it awards a degree that affirms knowledge and sends a graduate into a career. Second, it is a coming-of-age service that forms identity through social exploration and intellectual development. Partying, Greek life, sports, and other forms of college life always risked undermining the educational project, and vice versa. But the two categories coexisted well enough-until they began to unbundle.
The coming-of-age experience has become less dependent on the physical campus. Globalization and the internet brought much of the wider world to students long before they arrive. College football went national, abandoning regional rivalries. Greek rush became an outsourced industry. Campuses function as hospital systems, major local employers, and real-estate-investment operations as much as they are schools. Meanwhile, the economy demanded more from students outside the classroom-internships, networking, publications, volunteering. In a brutal job-search environment, undergraduates may feel that their older professors do not understand the gap between the delivery of a credential and the career it promises.
COVID-19 then revealed that the credentialing process could continue if campus life was reduced or eliminated. Some students sued expensive universities, arguing a closed campus was not worth full tuition. But credentialing proved entirely possible with a compromised or absent version of campus life.
AI severs the last connection between learning and living
AI has now separated the two purposes almost entirely. Writing an essay previously required doing something resembling the assigned academic work. Because AI makes it possible to complete coursework without intellectual labor, the output is disconnected from the process-if students choose the shortcut. Suffering through a term paper no longer inherently creates opportunities to get late-night coffee with fellow students, commiserate over a difficult professor, or reach an unexpected realization while drunk at a house party. Social and academic progress are no longer woven together.
A combination of shifting values, educational entrepreneurship, regulatory arbitrage, and the speed of technological change has altered what a diploma means. Yet schools have had no trouble continuing to award degrees to students who spent their undergraduate careers on campuses that tout the same traditional look on quads and inside libraries, but that AI has mangled beneath the surface. For educators navigating these shifts, AI for Education resources have become increasingly relevant to understanding how the technology reshapes both teaching and assessment.
Research and teaching: a structural mismatch
The Trump administration's attacks on higher education hit campuses where it hurt because they exposed a different incompatibility: the one between education and research. Since World War II, the American research-university sector has become an enormous, distributed, scientific-research industry. The funding cuts disrupted tens of billions of dollars of federal money that pays for that research, including graduate education. The fact that big universities are driven by research was not a secret, but it was also not well understood by the American public, including parents and students.
At research universities, undergraduate teaching is seen-if not described-as a necessary act to keep tuition dollars flowing and to maintain the appearance that undergraduate education is predominant. University administrators often argue that the research and teaching missions bolster each other: a star researcher can bring the latest knowledge to the classroom. But those purposes sit more in tension than in unity. A revealing term helps explain how tenuous the relationship really is: a researcher who gets a big grant can use part of those funds for a "teaching buyout," paying their way out of the classroom. Success at one of the university's missions literally purchases relief from the other. There is no research buyout.
Campuses such as Washington University in St. Louis take in $1 billion or more of research funding every year. That money pays for lab work, equipment, buildings, and staff. It forms the foundation for metrics-grants won and research expenditures-that campuses use to compete for faculty, graduate students, philanthropic gifts, and prestige. Threaten the flow of research money and you destabilize the financial and organizational structure of the whole sector.
Every major college touts the liberal arts as the core of an undergraduate education, but the humanities enjoy mostly vestigial value that does not match the underlying economics. Small liberal-arts colleges focus almost exclusively on undergraduate education and are earnestly committed to the arts and letters. That is one reason they are better positioned to weather funding attacks. But these schools serve smaller populations and remain tied to the broader university system. Even at these schools, the humanities are particularly vulnerable to AI, because the intellectual labor that the technology can now eliminate-wrestling with the meaning of texts, figuring out how to express complex ideas-is the fundamental point of that type of education. Teachers looking to adapt their methods can explore structured paths such as AI for Teachers.
Why this matters for educators
The incompatible purposes of college will not resolve themselves. Universities are purpose-built to resist change, and leaders have repeatedly adopted a "wait and see" approach, treating disruption as interruption. The degree will become ever more a goal and less a process of self-discovery. The research and teaching missions will further retract from each other. For educators, the practical implication is that the classroom is no longer a space where academic progress and personal formation are guaranteed to happen together. Faculty who treat teaching as their primary work will need to design courses that acknowledge AI's presence-not by policing it in vain, but by structuring assignments so that intellectual labor remains visible and assessable. The appearance of college will endure, but the work of educating young people now requires rebuilding the connection between credentialing and formation that the institution itself no longer provides.
Your membership also unlocks:
