For most of human history, knowledge institutions were built around a simple and stubborn fact. Knowledge was hard to reach. Books were expensive to produce, fragile to preserve, and limited in number. Expertise traveled slowly and lived in particular bodies, places, and traditions. To learn meant to approach something that was already scarce, and to gain access often required proximity, permission, or affiliation.
Libraries emerged within this reality not merely as cultural symbols, but as practical responses to material limits. A library was a place where scarcity could be softened through collective ownership. A book that could only belong to one person at a time could, through institutional stewardship, belong to a community across generations. The public character of libraries grew naturally from this arrangement. Ownership enabled sharing, and sharing enabled learning.
Universities developed alongside this model. They gathered texts, scholars, and students into shared spaces where knowledge could be transmitted, debated, and extended. Teaching was inseparable from access. To attend a lecture was to gain entry to explanations that were otherwise unavailable. Authority followed this structure. Those who controlled access to texts and interpretation also shaped what counted as legitimate knowledge.
Publishing completed this structure. Commercial publishers operated within the same scarcity logic, transforming control over reproduction into revenue. This did not initially conflict with the public role of libraries. Both systems depended on the same material constraints, and both performed necessary functions. The tension between public good and commercial interest existed, but it remained stable.
The digital turn unsettled this balance without immediately overturning it. Texts became easier to copy, but ownership did not follow. Instead of dissolving scarcity, digital systems reintroduced it through licenses, platforms, and access controls. Libraries could no longer own what they acquired. Knowledge became available everywhere and nowhere at once. The old center of gravity began to wobble.
The First Fracture: Digital Knowledge and the Limits of Control
The promise of digital knowledge was straightforward. If information could be reproduced at negligible cost, then access should expand, preservation should improve, and learning should become more inclusive. In practice, the opposite often occurred. Libraries found themselves paying repeatedly for access they could not secure permanently. Ebooks expired. Journals vanished behind changing paywalls. The very formats that should have ensured continuity instead introduced fragility.
This was not simply a technical failure. It reflected a deeper structural choice. Digital knowledge was framed not as an object to be owned, but as a service to be licensed. This preserved scarcity artificially, even when material constraints had largely disappeared. Access could be granted or withdrawn. Use could be monitored. Authority could be enforced contractually.
At this stage, the conflict was usually described in economic terms. Libraries sought to serve the public with limited budgets. Publishers sought to protect revenue. Authors were caught somewhere in between. But this framing missed something important. Access control was never only about money. It was also about authority.
To decide who may read is also to decide who may teach, who may interpret, and who may build upon a body of knowledge. Control over distribution confers control over legitimacy. In this sense, digital licensing did not merely protect profits. It preserved institutional power structures that had grown around scarcity.
A distinction began to matter more clearly here. Some forms of knowledge behave like infrastructure. Academic research, educational materials, reference works, and methodological insights lose value when access is restricted. Their purpose is collective advancement. Other forms of knowledge function more like expression. Trade books, essays, and cultural works can legitimately operate within markets shaped by preference, identity, and taste.
Digital systems, however, treated both categories the same. Everything became licensable content. The mismatch created strain, but the system still held together. Then a deeper change arrived.
When Explanation Becomes Abundant
The emergence of generative AI altered the landscape in a way that neither digitization nor open access had fully accomplished. AI did not simply improve access to existing knowledge. It transformed the act of explanation itself.
Where search engines retrieve documents, AI reconstructs understanding. Where textbooks present fixed sequences, AI adapts explanations to the learner. Where instructors repeat clarifications across semesters, AI responds endlessly without fatigue. The marginal cost of explanation approaches zero.
This shift matters because education has always been constrained less by information than by interpretation. Knowing that a text exists is not the same as understanding it. Learning depends on guidance, pacing, and responsiveness. These were historically scarce and labor intensive. AI dissolves that scarcity.
At this point, the earlier conflict reframes itself. The tension is no longer primarily between libraries and publishers, or between public and commercial access. It becomes a conflict between AI mediated explanation and any institution whose authority depends on controlling content.
AI does not need to distribute a textbook to teach from it. It does not need to own a journal to explain its methods. It bypasses access without redistributing objects. This unsettles both commercial and public knowledge institutions, because it separates ownership from understanding.
The question that emerges is no longer about licensing models. It is more fundamental. If explanation is abundant and adaptive, what role remains for institutions built around access?
Why Libraries Still Matter When AI Explains Everything
The temptation at this point is to declare libraries obsolete. If AI can explain anything, guide anyone, and personalize learning at scale, why maintain institutions whose original function was to mediate access?
The answer becomes clearer when explanation is distinguished from memory. AI excels at synthesis, interpretation, and generation. It does not preserve texts in stable forms. It does not maintain provenance. It does not guarantee continuity across time.
Libraries are not simply warehouses of information. They are institutions of temporal care. They preserve versions of knowledge as they were articulated, contested, and revised. They hold records steady even as interpretation changes. They make it possible to ask not only what we know, but how we came to know it.
In a world where knowledge is increasingly generated on demand, this anchoring function becomes more important, not less. Without it, societies risk replacing shared memory with endlessly plausible reconstructions. Understanding becomes fluent but ungrounded. Coherent but detached from history.
This does not mean libraries should remain unchanged. Physical circulation desks and static catalogs are not the essence of their role. What matters is stewardship. Libraries may evolve into custodians of open corpora, auditors of AI training sources, and guarantors of public memory. Their legitimacy shifts from access provision to trust maintenance.
AI accelerates understanding. Libraries slow knowledge down enough to be accountable. These roles are not redundant. They are complementary.
The AI Native University: When Teaching Stops Being Delivery
The university faces a more radical transformation. Traditional academic structures were shaped by the same scarcities as libraries. Lectures existed because expert explanation was rare. Textbooks existed because synthesis was costly. Curricula existed because pacing had to be standardized.
An AI native university begins from a different assumption. Explanation is abundant. Content generation is cheap. Personalization is scalable. Once this is accepted, teaching as delivery dissolves.
This does not eliminate education. It changes its center. Professors no longer serve primarily as transmitters of knowledge. They become mentors, critics, and guides. Their value lies in judgment, not repetition. In framing problems, not reciting answers. In shaping intellectual character, not covering material.
The shift unfolds differently across disciplines. In the natural sciences, research already revolves around experimentation, instrumentation, and validation. Writing papers has long followed standardized formats. AI can draft manuscripts, translate findings, and synthesize literature without undermining the epistemic core of scientific work. Responsibility remains with the researcher who designs the experiment and stands behind the claim.
In the humanities and social sciences, the change is deeper. Writing has traditionally been inseparable from thinking. AI does not replace this process. It forces it into dialogue. Scholars test interpretations against AI generated counterarguments. They explore alternative genealogies of ideas. They write with and against a system that mirrors the archive of human discourse.
Refusing this collaboration does not preserve purity. It narrows perspective. Just as the availability of books reshaped scholarship in earlier centuries, the availability of generative synthesis reshapes it now.
Teaching in this environment focuses on formation. Students learn how to ask meaningful questions, how to evaluate sources, how to justify claims, and how to use AI responsibly. Assessment shifts from recall to process. The path of reasoning matters more than the final answer. Learning becomes visible, iterative, and reflective.
The Quiet Deauthorization of Academism
As AI integrates open knowledge into its explanatory capacities, another shift becomes visible. Authorization loses its monopoly. Researchers increasingly publish freely on blogs, preprint servers, and open platforms. Ideas circulate before formal validation. Influence emerges through resonance rather than placement.
This does not abolish standards. It relocates them. Peer review remains valuable, but its function changes. It validates rather than permits. It evaluates methods and reasoning rather than granting access to visibility.
Authority moves from gatekeeping to transparency. Trust is earned through clarity of argument, reproducibility of claims, and sustained contribution over time. Institutions certify processes rather than controlling content.
Commercial publishing adapts along a different path. Much of it already operates as a culture of affiliation rather than a repository of knowledge. Readers pay not primarily for information, but for voice, presence, and identification. This model coexists with AI. It is not threatened by generative explanation because its value lies elsewhere.
Academic knowledge, by contrast, increasingly behaves like infrastructure. When it is openly shared, it gains influence. When it is locked away, it becomes peripheral. Many scholars recognize this intuitively. Being read, engaged, and integrated into collective understanding often matters more than royalties or exclusivity.
After Scarcity: Knowledge as a Shared Condition
The world that emerges after scarcity is not one where institutions vanish. It is one where their justification changes. Libraries no longer exist primarily to provide access, but to preserve memory and trust. Universities no longer exist to deliver content, but to cultivate judgment and responsibility.
AI becomes part of the environment in which thinking happens, not a replacement for thinking itself. Explanation is abundant. Meaning remains fragile. Stewardship becomes the central task.
The deeper question is not whether AI will reshape knowledge institutions. That has already occurred. The question is whether societies can redesign these institutions to preserve continuity, accountability, and care when access is no longer the organizing principle.
Knowledge stops being something we acquire and becomes something we inhabit. The task is no longer to control it, but to live with it well.
Image: StockCake
Tom’s Blog 