The Thousand-Year Usurpation

by Prof. Robert Karaszewski

There is a peculiar audacity in an institution that no longer remembers what it was built to do, yet continues to charge extraordinary prices for access to it.

The modern Western university presents itself as the heir of humanity's greatest intellectual traditions — the custodian of Socratic inquiry, Aristotelian empiricism, Humboldtian freedom, and the accumulated wisdom of civilizations stretching across three millennia. It is an impressive genealogy. It is also, in its most consequential respects, a selective appropriation disguised as inheritance.

What the modern university inherited from those traditions is their prestige. What it discarded — quietly, incrementally, and without ever quite acknowledging the loss — is their epistemological substance: the specific, hard-won understanding of what learning is for, what conditions produce it, and what institutional structures destroy it while appearing to support it.

The artificial intelligence systems now disrupting higher education are not the cause of this institution's crisis. They are its revealer. They have made visible what was always structurally latent: that the university's central claim — that it provides something irreplaceable, unavailable elsewhere, and worth the extraordinary cost of access — can no longer be defended on the institution's own terms.

Before the diagnosis can be made with precision, two clarifications are necessary. This manifesto does not argue that the university is worthless. It argues that the dominant model of mainstream higher education has optimized itself around the wrong problem, and that this misoptimization has become dangerous precisely because it is invisible to the institution performing it. Artificial intelligence has not eliminated the need for the university's highest function. It has eliminated the specific activity that most universities have mistaken for that function. The distinction is consequential, and the entire argument rests upon it.

I. What AI Has and Has Not Eliminated

The claim that artificial intelligence has abolished knowledge scarcity requires immediate qualification, because an imprecise version of that claim is both analytically wrong and strategically counterproductive.

What AI has eliminated — or is rapidly eliminating — is access scarcity: the condition under which organized, cross-referenced, expert-level knowledge could be reached only through institutional mediation. The large language model accessible today on a student's telephone contains more organized, retrievable, and synthesizable knowledge than any single university library in human history. It is available at zero marginal cost, without admission requirements, without tuition, and without the four-year waiting period that the credentialing apparatus requires.

What AI has not eliminated — and what no information system can eliminate, because it is not an information problem — is interpretive scarcity, validation scarcity, and contextual application scarcity.

Interpretive scarcity is the capacity to evaluate the reliability, relevance, and limits of applicability of information in conditions where no single authoritative source exists. Validation scarcity is the capacity to assess whether a chain of reasoning is sound, whether data is trustworthy, whether a model is adequate to the reality it claims to represent. Contextual application scarcity is the capacity to transfer knowledge from the conditions in which it was acquired to conditions that are structurally different, that involve genuine stakes, and that do not announce which frameworks are relevant.

These are not primarily information problems. They are problems of judgment. And they become more scarce, not less, as access to organized information expands. When everyone has access to the same knowledge base, the differentiating capability is not what one knows but what one can do with knowledge under conditions that AI cannot navigate on the human's behalf: genuine uncertainty, conflicting evidence, incomplete information, and real consequences.

This is the precise inversion of the dominant university's current operating assumption. Mainstream higher education has spent the better part of a century optimizing for the development of knowledge — curriculum design, content delivery, assessment of retention — while treating the development of judgment as an emergent property of sufficient exposure to knowledge. What AI reveals is that this assumption was always questionable and is now structurally untenable. Judgment does not emerge from knowledge alone. It emerges from the conditions under which knowledge is applied, tested, challenged, and revised in circumstances where error carries consequences.

The university's problem is not that AI knows more than its professors. The university's problem is that the dominant institutional model has organized itself to develop the one capability that AI has made abundant while systematically neglecting the capabilities that AI has made scarce.

II. The Traditions the University Claims to Inherit

To understand what has been neglected, one must first understand what existed before the modern institution consolidated its particular form of monopoly. These traditions are treated here as normative archetypes — deliberate simplifications intended to illuminate their epistemological cores rather than reconstruct their histories in full. Each was historically complex, internally contested, and institutionally imperfect. The point is not that they were golden ages. The point is that each understood something about the relationship between knowledge and judgment that the dominant model of mainstream higher education has largely forgotten.

The Indian tradition — Takshashila and Nalanda. Takshashila, flourishing from at least the sixth century BCE in what is now Pakistan, predates the universities of Alexandria, Athens, and Constantinople by several centuries. It served simultaneously the Śramaṇa and Vedic traditions, drew students from China, Greece, and Central Asia, and offered instruction across linguistics, law, astronomy, medicine, mathematics, and reasoning — not as disciplines in sealed departmental containers but as integrated domains of a unified inquiry into how the world works and how one ought to act within it.

Nalanda, established in the fifth century CE and operating for nearly a thousand years, pursued a model that the mainstream university has only rarely approached: the integration of rigorous intellectual formation with explicit moral development, organized around the proposition that education is the transformation of a person rather than the transmission of content to a person. This distinction is not a rejection of content. Nalanda's curriculum was vast and demanding. The point is that content was understood as the medium of transformation, not its product. The question the institution asked of every learning activity was not what does this deliver but what does this develop — a question that mainstream higher education has largely stopped asking at the level of institutional design.

These were not perfect institutions. Nalanda was organized around Buddhist orthodoxy in ways that shaped what inquiry was legitimate. Takshashila's curriculum reflected the priorities of the Brahmanical and later Buddhist traditions that sustained it. But their founding epistemological intuition — that acquisition is the beginning of education, not its end — is the intuition that the modern credential system has most systematically suppressed.

When Bakhtiyar Khilji burned Nalanda's libraries in 1193 CE, the fires burned for months. The institution destroyed in those months was, at minimum, epistemologically more ambitious than the dominant forms of higher learning Western Europe would produce for the following three centuries. This fact appears in almost no standard history of the university.

The Confucian tradition. Confucius is routinely caricatured in Western educational discourse as the patron of rote memorization and hierarchical deference. The caricature inverts the historical record. The Analects preserves a model of teaching in which knowledge was not transmitted from teacher to student but scaffolded according to the student's prior understanding and individual constitution — the yin cai shi jiao principle, teaching in accordance with individual capacity. When different disciples asked Confucius about the same virtue, he gave different answers, calibrated to their specific developmental needs.

The Confucian educational ideal — the junzi, the person cultivated in wisdom, ethical sensitivity, cultural literacy, and practical competence — was not a scholar in the modern academic sense. It was a person whose learning had produced a transformation of character, whose knowledge and judgment were inseparable, and whose education was understood as a lifelong practice of xiushen — self-cultivation — rather than a completed phase of institutional exposure.

The complications are real and must be named. Confucian pedagogy historically reinforced hierarchical social structures and prescribed gender roles that its own epistemological premises, consistently applied, would have challenged. The institutionalization of Confucian ideals through the keju examination system converted a project of ethical formation into machinery of state selection — optimizing the production of administrators rather than the development of judgment. The keju's abolition in 1905, not because it failed to select capable people but because it selected for capabilities the world no longer rewarded, is the historical warning that the mainstream university has not heard.

The Islamic tradition. Bayt al-Hikma in Baghdad, established around 830 CE, operated on an assumption that neither the medieval European university nor its mainstream successor has fully internalized: that all knowledge, regardless of cultural origin, is worth translating, aggregating, and subjecting to disciplined inquiry. Al-Qarawiyyin and Al-Azhar represented, in their founding spirit, institutions whose legitimating principle was amplification rather than control — places where the boundary between the known and the unknown was pursued rather than policed.

The historical record is more complicated. Al-Azhar became, over centuries, one of the most powerful gatekeeping institutions of Islamic orthodoxy. The Islamic Golden Age's extraordinary intellectual openness coexisted with internal debates about the legitimacy of rational inquiry that were not always resolved in inquiry's favor. But the founding epistemological impulse — that knowledge aggregation is an obligation, that the institution's value lies in what it activates rather than what it withholds — is precisely what the dominant credential-based model has inverted.

The African tradition. The Ubuntu philosophy — umuntu ngumuntu ngabantu, a person is a person through other persons — represents an epistemological tradition that mainstream Western higher education has not merely ignored but, through colonial export, actively suppressed. Ubuntu is not primarily a pedagogy. It is an ontology of knowing: the proposition that knowledge is inherently relational, that understanding is always co-constituted, that the knowing subject is not an autonomous individual processing information but a node in a network of relations that constitute simultaneously his identity and his comprehension.

The complications here are significant. Ubuntu is not a unified philosophical system but a family of related traditions across Sub-Saharan Africa, with internal variations that resist easy synthesis. Its application to formal institutional design is contested among African scholars themselves. But its core epistemological claim — that the isolation of the individual knowing subject from the relational context in which knowing actually occurs is not a methodological refinement but an epistemological distortion — poses a challenge to mainstream educational design that the dominant model has not taken seriously.

The Mesoamerican tradition. The Aztec educational system institutionalized a distinction that mainstream Western pedagogy would not theorize systematically until the cognitive revolution of the twentieth century: the distinction between declarative and procedural knowledge, between knowing that and knowing how, between knowledge as proposition and knowledge as performance under conditions of genuine consequence.

The Calmecac trained priests and rulers in organized knowledge — astronomy, law, theology, statecraft. The Telpochcalli trained warriors through physical discipline, strategic reasoning, and performance under actual conflict conditions. That these two institutions coexisted as parallel and complementary tracks reflects a sophistication of educational design that the mainstream university model has only marginally approached. Contemporary higher education offers the Calmecac. The Telpochcalli — the systematic development of procedural capability under genuine pressure — has been largely forgotten as a design principle.

The Greco-Roman synthesis. Socrates had no institution, charged no fees, and was executed by democratic vote for the epistemological crime of demonstrating that his fellow citizens did not know what they believed they knew. The elenctic method was not a teaching technique. It was a form of cognitive intervention: an assault on the epistemic complacency that Socrates regarded as more dangerous than ignorance, because it foreclosed the recognition of ignorance that genuine inquiry requires.

Quintilian's Institutio Oratoria — the first comprehensive educational program in Western history — understood something that neither Plato nor Aristotle fully articulated: that capability is formed only under conditions of genuine pressure, genuine consequence, and genuine stakes. Roman rhetorical education was the training of reasoning under adversarial conditions. It understood that the examination room is, at best, a rehearsal of the conditions that actually develop judgment — not a substitute for them.

III. The Pipeline Problem

The traditions surveyed above did not reject transmission. They understood it as the necessary first stage of a sequence that transmission alone cannot complete.

Knowledge acquisition — the transmission of organized content, conceptual frameworks, and disciplinary method — is the irreducible foundation of education. Without something to work with, there is nothing to develop. Nalanda transmitted vast bodies of philosophy, mathematics, and medicine. The Confucian tradition transmitted the classical canon. Bayt al-Hikma made transmission its founding act. The Telpochcalli warriors could not function without the declarative knowledge of strategy and terrain that preceded their procedural training.

The pipeline that education must complete runs from acquisition through structuring through application to judgment. The failure of the dominant university model is not that it performs acquisition. It is that it has confused acquisition with the terminus of the pipeline — treating the delivery and retention of organized content as the educational product rather than as the raw material from which education, properly understood, must produce something fundamentally different.

The difference is not between transmission and transformation as mutually exclusive alternatives. It is between an institution that understands transmission as the beginning of a developmental sequence and one that treats it as the sequence itself. The former asks, at every stage: what is this knowledge being developed into? What conditions will test, extend, and complicate it? What pressure will reveal whether it has actually changed the person who holds it? The latter asks only: has the content been delivered and retained?

Mainstream higher education has optimized, with considerable technical sophistication, for the delivery-and-retention end of the pipeline. At the level of systematic institutional design, it has done remarkably little to engineer the conditions under which delivery and retention become application and judgment. This is not because the latter is impossible. Certain institutions — among them Minerva University, Olin College of Engineering, Deep Springs College, and the clinical education models of the best medical schools — have demonstrated that it is possible and have built pedagogical architectures around it. But these remain exceptions operating at the margins of a system whose mainstream has not moved.

The point is not that innovation is absent. The point is that it has not scaled, has not been measured rigorously, and has not displaced the dominant model that continues to organize the experience of the overwhelming majority of students in higher education globally.

IV. The Medieval Compact and Its Inversion

The University of Bologna, in its origins, was not a papal instrument of doctrinal control. It was a guild — a corporation of students who hired professors, could dismiss them for incompetence, fined them for arriving late, and organized collectively to protect their interests. The universitas was not a community of scholars. It was an organized body of clients, and the professors were their contractors.

This founding structure was subsequently inverted by the very institutional development that made the European university powerful. As universities acquired papal charters, royal patronage, and the authority to grant degrees recognized across Christendom, the students who had originally held sovereignty became the institution's objects. The licentia docendi became the mechanism by which the institution controlled not merely who could teach but what knowledge was legitimate and what conclusions were orthodox.

Galileo's eighteen years at Padua — protected by Venetian pragmatism from Roman authority — and his subsequent trial under papal jurisdiction illustrate the structural logic. The institution free enough to harbor heresy as a method produced the most consequential scientific revolution of the early modern period. The institution that reasserted orthodoxy destroyed its most important mind.

The lesson is not that ecclesiastical authority is uniquely dangerous to inquiry. The lesson is that any institution organized around the certification of legitimate knowledge will, over time, prioritize the stability of its certification system over the vitality of the inquiry that certification was meant to represent. Modern accreditation regimes, in softer and more bureaucratic form, often reproduce the same logic.

V. The Humboldtian Moment and Its Systematic Betrayal

The University of Berlin, established in 1810 on Humboldtian principles, represents the most serious attempt in the history of the Western university to constitute an institution around the production of knowledge rather than its management. Humboldt's proposition — that teaching and research are inseparable, that Lehrfreiheit and Lernfreiheit are constitutional principles, that the university's purpose is free inquiry rather than authorized transmission — was a genuine epistemological revolution.

It did not survive its own success.

The Humboldtian university was colonized by two forces structurally incompatible with its founding logic. The nation-state required graduates for its apparatus and redesigned curricula accordingly. The market required credentials for its hiring processes, and the degree became a signal of completed exposure to approved content rather than evidence of intellectual formation.

The result was an institution that retained the vocabulary of Humboldtian ideals while organizing itself around the efficient production of credentialed graduates. The research reward structure — which elevates publication above pedagogical development — did not emerge from bad intentions. It emerged from the rational operation of an incentive system that measured what was measurable and rewarded what could be standardized. Path dependency accumulated across decades: hiring criteria, promotion standards, accreditation requirements, and funding structures that each made rational sense in isolation and that together produced an institution systematically misaligned with its own stated mission.

Path dependency is not the same as choice. The dean who understands precisely what needs to change does not face a simple decision. She faces a regulatory framework that evaluates her institution against criteria designed for the model she is trying to replace, a faculty whose members have built careers within the existing incentive structure, a board that measures success in rankings rewarding inputs rather than outcomes, and an accreditation body that has not yet developed instruments to certify what genuine educational transformation would produce.

This is why the accusation this manifesto levels is structural rather than moral. An institution can operate exactly as its incentive structures prescribe and still produce outcomes that betray its founding purpose. The absence of a villain does not imply the absence of a problem. It implies a harder problem: one that cannot be solved by identifying and replacing responsible parties, but only by redesigning the system that makes their behavior rational.

VI. The Functions the University Actually Performs

Any honest account of the university's crisis must acknowledge the functions it performs that are not reducible to knowledge transmission and that no alternative institution currently replaces.

Network formation is one of the university's least discussed and most economically significant outputs. The years a student spends at a selective institution produce a dense, durable network of relationships with peers who will occupy positions of authority across professional, civic, and economic domains for the following four decades. This network has real and substantial value that is independent of the quality of the education and that cannot be replicated by any online platform, AI tutoring system, or alternative credentialing mechanism. The fact that elite institutions command extraordinary premiums is not primarily a testimony to their pedagogy. It is a testimony to the value of the networks they provide access to. Any account of disruption that ignores this function is analytically incomplete.

Research infrastructure represents a second non-substitutable function. Capital-intensive science — particle accelerators, long-term longitudinal studies, biobanks, oceanographic programs — requires institutional infrastructure that no market mechanism provides at the required scale. The research university's claim to social and economic value rests significantly on this function — one that has been progressively decoupled from its pedagogical mission but that constitutes a genuine and irreplaceable contribution that disruption will not touch.

The credential coordination function is more complex. The degree currently performs labor market signal coordination: it tells employers something about candidates that employers cannot easily observe directly. This function is real, but its reliability is not yet being replaced — it is being progressively degraded. The degree increasingly predicts content exposure rather than performance capability, and the correlation between the two is weaker than four decades of institutional marketing suggested. A degraded coordination mechanism that the system continues to rely upon is more dangerous than one that has been clearly superseded, because its failure mode is invisible until it is sudden.

VII. The Unaddressed Problem: Social Selection

The university does not only teach. It filters, sorts, and legitimizes. This function — the conversion of institutional proximity into social credential, of academic performance into labor market position, and of degree into class passage — is arguably the most consequential thing the dominant university model does, and it is almost never named honestly in discussions of educational reform.

Any manifesto that proposes the radical reconstruction of the university without addressing what replaces this function is incomplete. The economy does not merely need people who can think. It needs a mechanism for identifying, at scale, which people can think — one sufficiently standardized to function as a coordination device and sufficiently reliable to be trusted by the parties using it.

The direction worth pursuing is a shift from completion-based credentialing toward performance-based documentation: systematic evidence of how a graduate's reasoning quality, decision calibration, and adaptive capability under complexity has actually developed across the duration of their education. Not a certificate of proximity to organized knowledge, but a documented trajectory of demonstrated capability under defined conditions.

The details of how such documentation would be constructed — how reasoning performance is operationalized, how measurement resists gaming, how cultural bias is identified and controlled, how the instrument is externally validated, and crucially how it avoids reproducing the keju's pathology of selecting for performance within an orthodoxy rather than capability beyond it — are questions of institutional architecture rather than manifesto. They require empirical answers, not rhetorical ones. This manifesto identifies the direction. The architecture is a separate and harder project.

What can be said here is this: the risk of creating a new keju in AI form — a system that selects with great technical precision for the capabilities its own instruments can measure while systematically suppressing the capabilities it cannot — is real, is serious, and must be named as the central design constraint of any alternative to the current credential system. The failure mode of the solution must be understood before the solution is built.

VIII. The Accusation, Precisely Stated

Mainstream universities have understood, for at least two decades, that something fundamental was changing. They have responded with a pattern of institutional behavior that is, at this point, sufficiently consistent across institutions, systems, and national contexts to constitute a structural tendency rather than a collection of individual failures.

The pattern is: manage the appearance of transformation while preserving the structures that make transformation impossible.

The online campus that reproduces the lecture at digital scale. The innovation center that occupies a corner of the building while core processes remain unchanged. The AI strategy document that addresses how the university will leverage artificial intelligence to enhance content delivery — as though content delivery were the problem rather than the premise. These are not failures of imagination, and they are not, in most cases, expressions of bad faith. They are the rational outputs of an incentive structure that measures and rewards what can be measured and rewarded, and that has not developed the instruments to measure what it claims — but cannot demonstrate — that it produces.

The asymmetry of this disruption matters and is not adequately captured by global generalizations. In the United States, elite research universities continue to generate genuine value through network formation, research infrastructure, and reputational positioning that partial credential erosion has not touched. The crisis is asymmetric: it threatens mid-tier institutions that offer credential without network and teaching without research far more urgently than it threatens institutions producing all three. In the UAE, the system is explicitly instrumental — organized around workforce development and state-building rather than epistemological aspiration — which means the relevant argument is not that it has betrayed its intellectual traditions but that it risks failing on its own terms: producing credential without the capability that a rapidly developing labor market is beginning to distinguish. In Central and Eastern Europe, institutional inertia is high and external pressure currently modest — creating the conditions for the most dangerous failure mode: the one that becomes visible too late, when accumulated pressures suddenly exceed the regulatory protection that has been suppressing their effects.

The structural accusation is this: the dominant model of mainstream higher education is producing outcomes inconsistent with both the intellectual traditions it claims to represent and the capabilities the economy is beginning to demand — and has not developed the internal instruments to perceive this inconsistency clearly enough to act on it. The absence of malicious intent does not reduce the severity of this indictment. A system that produces harm through structural rational behavior is not less harmful than one that produces it through deliberate choice. It is simply harder to fix.

IX. What Reconstruction Requires

The reconstruction this manifesto proposes is not the abolition of transmission. It is the completion of the pipeline that transmission begins.

The lecture remains. Content acquisition is the pipeline's necessary first stage. The question is what the institution builds downstream of it: conditions under which acquired knowledge is structured into frameworks, applied under uncertainty, and tested against real pressure until something that can legitimately be called judgment has developed.

The credit hour is replaced, or substantially supplemented, by measures of reasoning performance that are longitudinal, multi-dimensional, and capable of distinguishing genuine cognitive development from surface-level optimization for assessment. Elements of this are technically feasible with current AI capabilities, though the governance, validation, and fairness problems are at least as significant as the technical ones.

The faculty role is reconstituted — not diminished, reconstituted. The professor who can design conditions under which reasoning develops, interpret behavioral evidence of cognitive growth, and calibrate developmental pressure across learners is performing a pedagogical task more demanding than the delivery of even an excellent lecture. The institution that recognizes this in its hiring criteria, promotion standards, and professional development investments is the institution capable of delivering what it has always promised.

The credential is rebuilt from certifying completion toward documenting development. This requires building systematic observation of reasoning performance into core operations, publishing what those observations show, and staking institutional credibility on the results rather than on inherited reputation. This is the hardest requirement because it exposes the institution to the discovery that its processes produce less development than it claimed. That discovery is not a failure. It is the precondition of genuine improvement. The institution that forecloses this possibility has chosen its reputation over its integrity. That choice is survivable in the short term. It is not survivable indefinitely.

X. The Stakes

The university is not the first institution to claim custodianship of civilization's intellectual inheritance while organizing itself primarily around the preservation of its own authority. Churches, legal systems, and medical establishments have all faced versions of the same crisis: the moment when external conditions sustaining their monopoly changed, and the question became whether they would reconstitute themselves around their actual function or defend their incumbent position until an alternative answered the question for them.

The university has genuine advantages. It possesses intellectual resources whose depth no alternative institution replicates. It has research infrastructure that capital markets will not fund at the required scale. It has social relations and sustained intellectual community that no technology eliminates. These are real, irreplaceable sources of value.

What it does not have is unlimited time. Demographic pressures on traditional enrollment are structural and not reversing. Employer signals on credential reliability are weakening in one direction. Alternative certification ecosystems are developing. AI systems that have eliminated access scarcity are improving. And the interpretive, validation, and contextual application scarcities that remain — the capabilities the mainstream university could develop but largely does not — are becoming more economically valuable with every increment of AI capability, not less.

The institutions that understand this — and build the measurement infrastructure, redesign the pedagogical environment, and develop the credentialing systems to demonstrate that they develop what they claim to develop — will define what higher education becomes. Not by announcing a vision. By producing evidence.

XI. The Return

What this manifesto ultimately proposes is not the end of the university. It is its return — to the epistemological substance of the traditions it has spent five centuries claiming to represent while discarding their most demanding implications.

A return to the understanding — present in Nalanda, in the Confucian tradition, in Quintilian, in Ubuntu, in the Telpochcalli — that acquisition is the beginning of education, not its terminus; that content is the medium of transformation, not its product; and that the institution's design must reflect this distinction at every level of its operation.

A return to the recognition that judgment cannot be produced by a system that separates knowledge from the conditions of its application and certifies exposure as though it were formation.

A return to the insistence that the knowing subject is constituted through relations of inquiry with others, and that an institution built around the systematic isolation of individual performance has organized itself against one of the deepest epistemological truths available to it.

A return to the understanding that capability exists only in its exercise under conditions of genuine consequence — and that the examination room is, at best, a rehearsal of those conditions, not a substitute for them.

A return, finally, to the Socratic recognition that the most dangerous epistemic condition is not ignorance but its confident misrecognition as knowledge — and that the institution which certifies what it cannot measure, measures what it cannot justify, and sells access scarcity in a world of access abundance is engaged in exactly this misrecognition, at industrial scale, with consequences that extend well beyond its own survival.

The thousand-year usurpation is ending. Not because the university has been exposed as fraudulent. Its faculties contain some of the most serious intellectual work of our civilization. Its research infrastructure supports discoveries that no alternative institution would fund. Its networks form relationships that no platform replicates.

What is ending is the specific claim — the one that justified the structure, the tuition, the monopoly, and the assumption of irreplaceability — that the dominant model of mainstream higher education is doing the work of developing judgment while it is actually doing the work of distributing knowledge.

The traditions this university claims to represent understood this distinction. They built institutions around it — imperfect institutions, historically constrained institutions, institutions that contained their own contradictions and their own betrayals. But institutions that asked the right question: not what should we deliver, but what should we develop, and under what conditions does development actually occur.

AI has made that question unavoidable again.

The mainstream now faces two options. It can treat this question as a threat to be managed, or as the first serious opportunity in a century to reconstruct itself around development rather than delivery.

The choice it makes — or fails to make — will be the most consequential educational decision of the next generation. Not because the manifesto demands it. Because the conditions that made the alternative possible have already arrived, and conditions do not wait.