Universities need to adapt to a world in which knowledge is no longer scarce - LSE Impact

The organising principle of higher education has traditionally been the delivery of knowledge under conditions of scarcity. If as Anupama Saini argues AI tools are the latest instance in a long run trend for growing access to an abundance of knowledge, how might universities adapt, when one of their fundamental purposes has changed?

For universities ubiquitous AI has created a crisis in confidence. Around the world, institutions are rethinking how students learn, how learning should be assessed and what role artificial intelligence should play in higher education. International organisations such as the OECD and UNESCO have encouraged universities to respond quickly to these changes.

These responses are necessary. But they also share a largely unquestioned assumption: that artificial intelligence is the disruption universities must adapt to. What if that is only part of the story? Perhaps AI is not changing the way universities teach and assess. Perhaps it is exposing something that has been changing for much longer.

Long before ChatGPT, there were other things that transformed how knowledge is created, accessed and shared - digitisation, search engines, open educational resources and online learning. Generative AI has accelerated this transition dramatically, but it did not begin it. It has compressed years of gradual change into a remarkably short period of time.

That distinction matters. If AI is accelerating a transformation already underway, then the most important question facing universities may not be how they should respond to artificial intelligence. It may be whether institutions that were designed for a world in which access to knowledge and expertise was the defining educational constraint are still organised around the 'right' problem.

Universities as a response to the scarcity of knowledge

Universities did not develop their defining practices by accident. They evolved to solve a particular problem: reliable knowledge and recognised expertise were difficult to access. Books were expensive to produce, scholarly communities were geographically dispersed, and opportunities to learn from experts were limited. Preserving, verifying and transmitting knowledge required institutions capable of concentrating expertise and passing it from one generation to the next.

Many of the structures that continue to define higher education emerged as effective responses to these conditions. For example, libraries assembled collections that individual scholars could never hope to own, lectures enabled one expert to teach many students simultaneously, academic journals created trusted mechanisms for evaluating and disseminating new discoveries. Degrees provided a widely recognised signal of expertise. Each emerged as a response to the defining scarcity of its age.

Over the past three decades, however, that scarcity has been steadily transformed. Generative AI represents the latest stage in this longer trajectory, making it increasingly easy not only to retrieve information but also to summarise, compare, analyse and explain it in real time. As UNESCO has argued, generative AI should be understood not simply as another educational technology but as part of a broader transformation that requires institutions to rethink pedagogy, policy and the capabilities learners need to thrive.

Universities have adapted repeatedly throughout their history as knowledge has been produced and shared in new ways. The current moment may be different for one reason: it is not simply changing how knowledge is accessed - it is changing the defining scarcity around which universities have organised teaching and learning for centuries.

Why current AI debates feel incomplete

Universities today find themselves navigating a series of challenges. They seek to discourage inappropriate uses of generative AI while encouraging students to develop AI literacy. They redesign assessments while simultaneously questioning what graduates should still be expected to know. They invest in increasingly capable technologies while worrying that students may become overly dependent on them. These are legitimate concerns. Yet they all begin from the same assumption that artificial intelligence is the primary disruption to which higher education must respond.

A different interpretation is possible.

History suggests that technological revolutions rarely eliminate the fundamental challenges societies face. Instead, they change where those challenges are encountered. The printing press dramatically increased access to books, but in doing so made literacy the next educational frontier. As literacy spread and printed material multiplied, finding trustworthy information became increasingly difficult, giving rise to libraries, catalogues, indexes and, eventually, search engines. Each breakthrough solved one educational bottleneck while exposing another that had previously remained in the background.

This recurring pattern suggests what might be called a 'conservation of scarcity.' Technological progress rarely abolishes scarcity altogether. It transfers it. Every innovation resolves one constraint while revealing another, shifting the frontier of human effort rather than eliminating it.

Generative AI appears to be accelerating this same transition. By making information retrieval, explanation and synthesis increasingly effortless, it reduces the scarcity that shaped universities for centuries while making other capabilities comparatively more valuable.

Seen through this lens, the challenge facing universities is not primarily technological. It is institutional.

Universities were organised around the defining educational scarcity of their age. If that scarcity is changing, then the most important question is not how universities should respond to artificial intelligence, but whether they remain organised around the educational problem they were originally designed to solve.

What changes when the defining scarcity changes?

If universities were organised around the defining scarcity of their age, then a shift in that scarcity has implications that extend far beyond AI policy. It requires institutions to reconsider not simply the technologies they adopt, but the assumptions embedded within curriculum, assessment, academic support and leadership. The question is no longer how universities should respond to AI, but what educational problem they now exist to solve.

Assessment is perhaps the clearest example. In a world where information can be retrieved, summarised and explained within seconds, the reproduction of knowledge becomes a less reliable indicator of learning. This does not diminish the importance of disciplinary expertise; it changes how that expertise is demonstrated. Educational value increasingly lies in students' ability to interpret evidence, justify decisions, synthesise ideas, recognise uncertainty and apply knowledge in unfamiliar contexts. The question shifts from What do students know? to What are they able to do with what they know?

The same logic extends to curriculum. Many of the defining challenges of the twenty-first century, from climate change and public health to cybersecurity and artificial intelligence itself, resist disciplinary boundaries. Universities have long acknowledged the importance of interdisciplinary learning, yet it often remains peripheral to institutional design. If the conservation of scarcity is a useful lens, then this is no coincidence. As access to information becomes easier, the ability to connect knowledge across domains becomes comparatively more valuable. Interdisciplinary learning is therefore no longer simply desirable; it becomes a response to a changing educational constraint.

The evolution of the academic library illustrates this transition particularly clearly. For centuries, libraries addressed the scarcity of access to knowledge. Today, their distinctive contribution increasingly lies in helping students navigate abundance. Developing AI literacy, research literacy, critical evaluation and responsible information practices becomes as important as providing access to collections. The library evolves from a repository of knowledge into an institution that helps learners make sense of knowledge.

The implications also extend to academic staff and institutional leadership. Faculty expertise remains indispensable, but teaching increasingly involves cultivating intellectual habits that cannot be outsourced to intelligent systems: evaluating competing claims, exercising judgement, working across disciplinary boundaries and engaging responsibly with uncertainty. Institutional leaders face a parallel challenge. Rather than asking how universities should integrate AI into existing structures, they must ask whether those structures remain aligned with the educational problem universities now need to solve.

Seen in this light, artificial intelligence is not the destination of institutional change. It is the catalyst that reveals a deeper transformation already underway. Universities are not losing their purpose. They are losing the scarcity around which that purpose was organised.

📨Enjoying this blogpost? Sign up to our mailing list and receive all the latest LSE Impact Blog news direct to your inbox 📨

This article gives the views of the author, not the position of LSE Impact or the London School of Economics. You are agreeing with our comment policy when you leave a comment.

Image credit: AnilD on Shutterstock.