The Six Functions

TAM-RWR.5-01 · The Reshaped World, Arc 5: The Learning Civilization · The Approximate Mind

Robert Acheson has been president of a regional university in the upper Midwest for six years. In that time enrollment has dropped eighteen percent, which is severe but not unusual for institutions of his type and geography. He has tried most of what there is to try. New programs in data analytics and health informatics. A partnership with a regional employer for guaranteed internship placement. A marketing campaign that cost more than it returned. A strategic plan with the word “innovation” appearing forty-three times, which he approved because the board needed to see the word and the plan needed to be approved.

He keeps a framed photograph on the bookshelf behind his desk. Not of his family, though there is one of those too. This one is of the 1987 graduating class of the college he attended as an undergraduate, a small liberal arts school in Pennsylvania that closed in 2024. He is in the fourth row, barely visible, holding his cap at an angle because the wind was blowing. The school had 1,200 students when he graduated. It had 340 when it closed. The buildings were sold to a healthcare company that is converting the dormitories into assisted living units.

He does not display the photograph to make a point. He displays it because it is the last physical evidence of an institution that formed him, and because the institution no longer exists to testify on its own behalf.

The Bundle

#

The university is a bundle. It has been a bundle for so long that the bundling has become invisible, the way a rope’s individual fibers are invisible when the rope is taut. You see the rope. You trust its strength. You do not think about the fibers until they begin to separate.

The bundle contains at least six functions that arrived at the same institution by historical accumulation rather than by design. Each function has its own logic, its own constituency, its own funding stream, and its own relationship to the AI transition. They were never meant to be in the same institution. They ended up there because the medieval European university was a convenient container, and the convenience hardened into tradition, and the tradition hardened into assumption, and the assumption is now being tested by a technology that can perform some of the functions without the container.

The first function is research. The systematic production of new knowledge through disciplined inquiry. This is what the public and the faculty both think the university is primarily for, though they are both partly wrong. Research is the university’s prestige function, the one that generates rankings and reputation and the external funding that subsidizes everything else. AI is changing research profoundly: accelerating literature review, enabling pattern detection across datasets too large for human processing, generating hypotheses, and in some fields beginning to design experiments. The effect is not to make research unnecessary. It is to make the university’s monopoly on research infrastructure less absolute. A well-equipped lab still requires a physical institution. A well-equipped computational environment does not.

The second function is credentialing. The certification that a person has completed a course of study and demonstrated competence at a level the credential represents. This is what most students are actually purchasing, and what most employers are actually screening for, regardless of what either party says about learning or development. The credential is a signal. Its value depends on the signal’s reliability, and the signal’s reliability depends on the assumption that what the university tested is what the employer needs. That assumption is weakening. Not because universities test the wrong things, though some do. Because the relationship between what can be tested in an academic setting and what is needed in a professional setting is becoming less stable as professional requirements change faster than academic programs can follow.

The credential worked when what it certified changed slowly. It is strained when what it certifies is in motion.

The third function is professional training. The development of specific competencies for specific professional roles: engineering, nursing, accounting, education, law. This is the function most visibly disrupted by AI, because AI can now perform portions of what professionals do, which means the competencies being trained are not the same competencies the profession will require when the student arrives. The nursing program that trains assessment skills is doing essential work. The accounting program that trains students to prepare tax returns is training for a task that AI handles more accurately and more cheaply than a human. The programs know this. The accreditation bodies, which determine what programs must teach, change more slowly than the programs they accredit.

The fourth function is coming-of-age. The structured transition from adolescence to adulthood, conducted in a residential setting with age peers, away from the family of origin, over a defined period. This function is so deeply embedded in American culture that questioning it feels like questioning adulthood itself. But it is worth observing that the coming-of-age function is historically unusual. Most societies in most periods managed the transition to adulthood without a four-year residential interlude, and many of those societies produced competent adults. The coming-of-age function is not universal. It is a specific cultural solution to a specific developmental need, and it is expensive. It requires dormitories, dining halls, student life staff, counseling services, recreational facilities, and the full apparatus of a small city maintained for the purpose of housing people between the ages of eighteen and twenty-two. The question is not whether this function is valuable. It is whether it is so valuable that it justifies the cost structure it requires, for the populations that currently cannot afford it.

The fifth function is social sorting. The production and reproduction of class membership, professional networks, romantic partnerships, and social capital. This is the function that dare not speak its name. No university president will say: we exist to sort young people into social categories and connect them with the people they will hire, marry, and do business with for the rest of their lives. But the evidence that this is what the university does, and that this function accounts for a significant portion of the economic return to a degree, is robust. The alumni network, the fraternity, the seminar with twelve people who will become partners at law firms: these are not incidental features of the university experience. For many students, they are the experience, and the classroom is the incidental feature.

The sixth function is community. The provision of belonging, shared identity, shared struggle, and the experience of being embedded in a group organized around something larger than individual preference. This is the function closest to what Robert’s college in Pennsylvania provided and what he cannot quite name when he tries to explain why its closure felt like a death rather than a closure. Community is what the university provides that is hardest to replicate, hardest to price, and hardest to defend in a budget meeting. It is also what students report missing most when they attend entirely online.

What AI Does to Each Fiber

#

The six functions are not equally vulnerable. AI does not dissolve the bundle uniformly. It pulls at specific fibers while leaving others intact, and the fibers it pulls are not always the ones the institution is defending.

Research is being transformed but not displaced. The tools are new. The enterprise is not. What is changing is who can participate: the computational infrastructure required for many forms of research is becoming available outside the university, which means the university’s role as gatekeeper of research capacity is weakening. The independent researcher, the corporate lab, the government institute, the well-funded nonprofit: all of these can now conduct research that previously required a university’s infrastructure. The university retains the advantage of disciplinary communities, peer review structures, and the freedom that tenure provides. These are real advantages. They are not monopoly advantages.

Credentialing is under the most direct pressure, though the pressure is slow. Employers are beginning to experiment with alternative signals: portfolio assessment, skill-based hiring, AI-evaluated work samples, micro-credentials from non-university providers. The experiments are tentative. The four-year degree remains the default screening mechanism for most professional hiring. But the default is being questioned, and the questioning accelerates each year the gap widens between what the degree signals and what the employer needs.

Professional training faces the paradox the distillation thesis identified across all professions: AI absorbs the skill scaffolding and reveals the vocational core. The nursing program that trains clinical assessment is training the core. The law school that trains legal research is training the scaffolding. Both are in the same institution, subject to the same accreditation standards, offering the same credential. The distinction between what is core and what is scaffolding is the distinction the institution has no mechanism to make, because making it would require admitting that some of what it teaches is being superseded while it is still being taught.

Coming-of-age is largely untouched by AI, which is why it may be the function around which the surviving university reorganizes. The residential experience of being among age peers, navigating independence, making mistakes at manageable scale, forming an identity away from the family of origin: these require physical co-presence, and physical co-presence is what AI cannot provide. The irony is that the function the university never primarily organized itself around may be the one that justifies its continued existence as a physical institution.

I wonder whether Robert’s institution, and the hundreds like it, will eventually stop pretending they are primarily in the knowledge business and acknowledge that they are primarily in the formation business. The knowledge is increasingly available without them. The formation is not.

Social sorting is accelerated by AI in ways that are difficult to discuss publicly. AI matching algorithms already do a version of what the university’s social sorting function does: connect people with complementary interests, skills, and social positions. LinkedIn is a social sorting mechanism. So is every professional networking platform. The university’s advantage is that its sorting happens through sustained proximity over four years, which produces deeper connections than algorithmic matching. The disadvantage is that the sorting is expensive, geographically constrained, and accessible only to those who can afford the admission price. AI-mediated sorting is less deep but far more accessible. The question is whether depth or accessibility matters more, and for whom.

Community is the fiber that holds when everything else frays. Not because community is inherently strong. Because community is what remains when the instrumental justifications for the institution have been met by other means. If you can get the knowledge elsewhere, the credential elsewhere, the professional training elsewhere, the social network elsewhere, what is left? The experience of being in a room with people who are struggling with the same questions you are, guided by someone who has struggled with them longer, in a setting where the struggle is the point rather than an obstacle to the point.

That is a real thing. It is also a difficult thing to put on a billboard.

The Survivor’s Question

#

Robert’s strategic planning process has produced, across six months of committee work and stakeholder engagement, a document that says his university will become “a nationally recognized leader in applied learning and workforce development.” He approved the document. He does not believe the document. Not because it is wrong but because it describes a version of the university that any institution could become, and the competition to become it is a competition his institution, with its enrollment and its endowment and its geography, is unlikely to win.

What he believes, and has not yet found the language to say in a governance meeting, is that his institution’s actual value proposition is something closer to what his Pennsylvania college provided: a place where people are formed, not just trained. Where the struggle of learning is scaffolded by relationships with people who care whether you succeed. Where the coming-of-age happens in a community small enough that someone notices if you don’t show up.

He cannot call this “workforce development.” The board wants workforce development. The state legislature that partially funds the institution wants workforce development. The parents paying tuition want a return on investment that they can measure in starting salary.

The thing the university actually provides that nothing else provides is the thing no one is willing to pay for by name.

The formation function. The community function. The experience of being in a place where someone older and more experienced takes an interest in whether you become a person of judgment and not merely a person of skill.

Robert knows this. He does not know how to operationalize it. The institutional apparatus around him is designed to measure credits, enrollment, retention, graduation rates, and post-graduation employment. None of these measure formation. None of them capture whether the student who sat in the back of the seminar and said nothing for eight weeks and then asked a question that changed the direction of the conversation was formed by that experience in ways that will matter for the rest of her life. No metric captures that. The student herself may not know for twenty years.

The Geographic Divide

#

The unbundling does not distribute evenly across institutions. The research university with its $2 billion endowment and its global reputation will survive the unbundling because it never depended on any single function. It can afford to let the credential weaken because its brand carries independent value. It can afford to let professional training migrate because its research enterprise generates sufficient revenue and prestige. It can afford to maintain the coming-of-age and community functions because it has the financial base to subsidize them.

The regional university, Robert’s university, cannot afford any of this. It depended on the bundle. The bundle was the product. Each function subsidized the others in ways that only become visible when the bundle frays. The research function, modest as it was, generated the grants that funded the graduate assistants that staffed the undergraduate labs that kept the tuition-to-faculty ratio manageable. The credential function generated the enrollment that funded the dormitories that provided the coming-of-age experience that attracted the enrollment. The professional training function generated the employer partnerships that generated the internship placements that generated the post-graduation employment statistics that generated the next year’s applicant pool.

Remove any fiber and the others weaken. The bundle was load-bearing in the same way the friction was load-bearing in the service economy: invisibly, until its absence revealed the structure it had been supporting.

The institutions that survive will be the ones that understand which functions they actually provide and reorganize around those functions before the market makes the decision for them. The institutions that do not survive will be the ones that continued to sell the bundle while the bundle’s individual components became available elsewhere at lower cost.

Robert’s Pennsylvania college did not understand this in time. It continued to sell the bundle, at a price the bundle could no longer justify, to a population that had options the population of 1987 did not have. The closure was not sudden. It was a thirty-year compression, visible in the enrollment figures that declined two or three percent per year for so long that each individual year’s decline felt manageable, until the compound became terminal.

The Photograph

#

The 1987 graduating class had 278 people in it. Robert can name perhaps forty of them now. The ones he can name are not the ones who sat next to him in classes. They are the ones he ate with, argued with, stayed up late with, failed an exam alongside and commiserated with over bad coffee in a common room that smelled like carpet cleaner and microwave popcorn.

The formation happened there. Not in the lecture hall. In the common room, the dining hall, the walk across the quad at eleven at night when someone said something that rearranged how he thought about what he was studying. The formation happened in the spaces between the formal functions, in the interstices the institution provided without intending to, simply by putting people together in a place for four years and letting proximity do its work.

He has not been able to replicate this in an online format. He has tried. The Zoom seminar does not produce the walk across the quad. The discussion board does not produce the argument over bad coffee. The formation requires co-presence, and co-presence requires a physical institution, and a physical institution requires a financial model, and the financial model depends on the bundle, and the bundle is fraying.

He looks at the photograph. The wind is still blowing. The cap is still at an angle. The school is assisted living now.

He is trying to save his institution from the same trajectory. He is not sure the trajectory is avoidable. He is sure that if it is avoidable, the path runs through an honest answer to the question the strategic plan did not ask: not what will make us competitive, but what do we provide that nothing else provides, and is anyone willing to pay for it?

He does not have the answer. He has the question, which is further than the strategic plan got.

This is the first essay in Arc 5 of The Reshaped World, examining education as civilizational system rather than as profession or institution. The arc traces what happens when the civilization’s self-reproduction mechanism, the system through which each generation develops the capacity to inhabit and extend the world it inherits, faces a transition faster than its structural lag allows it to process. This essay establishes the university’s six functions and asks which survive the unbundling. The essays that follow examine what education is actually for when knowledge is free (5-02), who gets thoughtful augmentation versus emergency content delivery (5-03), the credential that does not exist (5-04), and the civilizational transmission question that underlies all of them (5-05).

References

#

The University as Bundle

Christensen, Clayton M., and Henry J. Eyring. The Innovative University: Changing the DNA of Higher Education from the Inside Out. Jossey-Bass, 2011.

Zemsky, Robert. The College Stress Test: Tracking Institutional Futures across a Crowded Market. Johns Hopkins University Press, 2020.

Credentialing and Its Alternatives

Caplan, Bryan. The Case Against Education: Why the Education System Is a Waste of Time and Money. Princeton University Press, 2018.

Fuller, Joseph B., et al. Dismissed by Degrees: How Degree Inflation Is Undermining U.S. Competitiveness and Hurting America’s Middle Class. Accenture, Grads of Life, and Harvard Business School, 2017.

Formation, Community, and Coming-of-Age

Deresiewicz, William. Excellent Sheep: The Miseducation of the American Elite and the Way to a Meaningful Life. Free Press, 2014.

Chambliss, Daniel F., and Christopher G. Takacs. How College Works. Harvard University Press, 2014.

Higher Education Closures and Market Dynamics

Grawe, Nathan D. Demographics and the Demand for Higher Education. Johns Hopkins University Press, 2018.

Selingo, Jeffrey J. College (Un)Bound: The Future of Higher Education and What It Means for Students. New Harvest, 2013.

Social Sorting and Network Effects

Rivera, Lauren A. Pedigree: How Elite Students Get Elite Jobs. Princeton University Press, 2015.

Stevens, Mitchell L. Creating a Class: College Admissions and the Education of Elites. Harvard University Press, 2007.