skynesher/E+/Getty Images
The conceptual value of higher education feels as if it is being quickly washed away by the tidal wave of AI. A machine can write the essay, pass the exam, finish the problem set before a student has finished reading it— and the more it can do, the more the worth of a degree seems to recede behind it.
But I think we have misread what the wave is taking. It is not removing the value of what we do as educators: Rather, it is taking away the work we relied on to implicitly carry what we value. We assigned the essay, the problem set, the lab report and assumed the capacities we were looking for —such as reasoning, judgment, the patience to work a hard problem through—would develop in the doing. The task was the vehicle, and the capacity rode along unnamed. Now that a machine can do many of these tasks, an assumption that was always unreliable becomes impossible to ignore.
Most of us rarely teach those capacities on purpose. We name them as course objectives and then deliver them by implication—in a comment on a draft, a seminar question left open, the way an instructor takes apart a weak argument in real time. Part of the reason is straightforward: These capacities are harder to define than disciplinary content, and we have never been confident we know what the thing is, much less how to teach it. We keep reaching for new terms—soft skills, durable skills, transferable competencies, 21st-century skills, career-readiness competencies—and that restlessness is itself a tell. Having never resolved the definition problem, we have let the difficulty of measuring these capacities harden into an excuse not to teach them directly.
A faculty survey we recently ran at Cornell University confirmed as much: When asked about teaching critical thinking, instructors across disciplines named deficits in their own skills and knowledge as their most significant obstacle—they want to develop these capacities in students but don’t feel prepared to define, teach or assess them directly.
But left implicit, these objectives tend to reach only the student already equipped to receive them. Many students cannot practice what no one has named for them, and those from first-generation and underresourced backgrounds pay the steepest price.
Making the implicit explicit doesn’t require a new course or a major redesign. It takes naming, out loud, the human capacity an assignment is meant to build—and I recently watched a roomful of instructors discover how easily that can be done.
Earlier this summer, I ran a two-day workshop, Teaching for Tomorrow: Highlighting the Value of Higher Education, for 28 faculty and staff from across my college—postdocs and lecturers, tenured professors, librarians, extension associates. The goal was to help instructors do three things they rarely get structured time to do: identify what their students are learning to do as people, not just as disciplinary learners; make the transferable dimensions of their teaching more explicit and intentional; and articulate the value of their courses beyond content and credentials.
Most of them arrived with AI on their minds. Asked about their goals before the workshop began, they described questions dominating faculty conversations everywhere: how to combat AI use in assessments, how to incorporate it responsibly, how to design assignments it cannot simply finish. Those are legitimate concerns. But over two days, something shifted —the technology had receded, and what came forward instead were the people in the room and the people in their classrooms.
I posed the question: Five years from now, what do you hope a former student carries forward from your course, beyond content knowledge? Their answers were strikingly consistent. Across disciplines—microbiology, ornithology, soil science, global development, communication—they reached for the same handful of things: comfort with uncertainty. Intellectual honesty about the limits of one’s knowledge. The ability to hear a confident claim and instinctively ask what it leaves out. Curiosity that persists when an experiment fails. The capacity to navigate disagreement without damaging a relationship. One wanted students “comfortable working in the gray, to handle and navigate uncertainty through confidence in their ability to identify and work with information to move forward.” Another hoped they would learn “to stay curious rather than defeated” when things go wrong, and to “bring collaborative, low-ego energy” to the work.
Without exception, what they most wanted to endure was a way of being in the world—curiosity, honesty, resiliency and the capability of working with others through genuine difficulty.
After two days exploring practical ways to teach these capacities more deliberately, I asked each participant to name one thing they would carry into their own class. The answers were concrete and low-cost. One was rewriting a group-project rubric to name the capacities it had always quietly built—leadership, collaboration, giving and receiving criticism, communication—rather than leaving students to infer them. Another was revising prompts to make students name the capacity they were practicing and point to evidence of using it.
None of this is a new course. None of it requires a major redesign. All of it is a decision to say out loud what we have always been doing and why—and the evidence suggests it works. My colleagues and I developed a teaching framework that defined and highlighted a list of teachable critical thinking skills and dispositions, piloted across 2,600 students, with measurable gains largest among those with the least prior exposure. These capacities yield to direct instruction when we decide to provide it.
The work begins in individual classrooms. Pick one course this term and make one implicit thing explicit—before students start an assignment, tell them which human capacity it is meant to build and why it will still matter when the content is long forgotten. This is a five-minute change, but the student who knows what they are practicing is the one who carries it forward. For those who lead institutions, the Cornell faculty survey finding points directly to what is needed: Instructors want to develop these capacities in students but don’t feel equipped to do so. The appetite for development is there, and what faculty need is the occasion and the support, not a mandate but an invitation backed by real resources.
The question of what college is for has never been more prevalent. Families scrutinize the cost with a skepticism that would have been unthinkable a generation ago, and politicians have been quick to exploit that anxiety, flattening the purpose of education into the production of employable units. But if we cannot say clearly what we develop in students beyond content knowledge, we risk leaving the field open to those who would define it for us. The answer has been in front of us all along, in every course we teach. What remains is the decision to say it, loudly and on purpose.
Most faculty still hold real authority over what happens inside their own courses, and we should use it while we have it. No one is better positioned to shape the future than we are. The researchers, lawmakers and leaders of the next generation sit in our classrooms now, and what we teach them to value, they carry into the world.
The list of things a machine can do for us is long, and it grows longer every day. The list of things it cannot do is short, but what is on it has always been the real point of an education. All the machine has done is sharpen the question we were always meant to answer: not what our students should know, but whom they should become.
