AI is driving down the price of knowledge – universities have to rethink what they offer

For a long time, universities worked off a simple idea: knowledge was scarce. You paid for tuition, showed up to lectures, completed assignments and eventually earned a credential.

That process did two things: it gave you access to knowledge that was hard to find elsewhere, and it signalled to employers you had invested time and effort to master that knowledge.

The model worked because the supply curve for high-quality information sat far to the left, meaning knowledge was scarce and the price – tuition and wage premiums – stayed high.

Now the curve has shifted right, as the graph below illustrates. When supply moves right – that is, something becomes more accessible – the new intersection with demand sits lower on the price axis. This is why tuition premiums and graduate wage advantages are now under pressure.

According to global consultancy McKinsey, generative AI could add between US$2.6 trillion and $4.4 trillion^[1] in annual global productivity. Why? Because AI drives the marginal cost of producing and organising information toward zero.

Large language models no longer just retrieve facts; they explain, translate, summarise and draft almost instantly. When supply explodes like that, basic economics says price falls. The “knowledge premium” universities have long sold is deflating as a result.

Employers have already made their move

Markets react faster than curriculums. Since ChatGPT launched, entry-level job listings in the United Kingdom have fallen by about a third^[2]. In the United States, several states are removing degree requirements from public-sector roles.

In Maryland, for instance, the share of state-government job ads requiring a degree slid from roughly 68% to 53%^[3] between 2022 and 2024.

In economic terms, employers are repricing labour because AI is now a substitute for many routine, codifiable tasks that graduates once performed. If a chatbot can complete the work at near-zero marginal cost, the wage premium paid to a junior analyst shrinks.

But the value of knowledge is not falling at the same speed everywhere. Economists such as David Autor^[4] and Daron Acemoglu^[5] point out that technology substitutes for some tasks while complementing others:

• codifiable knowledge – structured, rule-based material such as tax codes or contract templates – faces rapid substitution by AI

• tacit knowledge – contextual skills such as leading a team through conflict – acts as a complement, so its value can even rise.

Data backs this up. Labour market analytics company Lightcast^[6] notes that one-third of the skills employers want have changed between 2021 and 2024. The American Enterprise Institute^[7] warns that mid-level knowledge workers, whose jobs depend on repeatable expertise, are most at risk of wage pressure.

So yes, baseline knowledge still matters. You need it to prompt AI, judge its output and make good decisions. But the equilibrium wage premium – meaning the extra pay employers offer once supply and demand for that knowledge settle – is sliding down the demand curve fast.

What’s scarce now?

Herbert Simon, the Nobel Prize–winning economist and cognitive scientist^[8], put it neatly decades ago: “A wealth of information creates a poverty of attention.” When facts become cheap and plentiful, our limited capacity to filter, judge and apply them turns into the real bottleneck.

That is why scarce resources shift from information itself to what machines still struggle to copy: focused attention, sound judgement, strong ethics, creativity and collaboration.

I group these human complements under what I call the C.R.E.A.T.E.R. framework:

• critical thinking – asking smart questions and spotting weak arguments

• resilience and adaptability – staying steady when everything changes

• emotional intelligence – understanding people and leading with empathy

• accountability and ethics – taking responsibility for difficult calls

• teamwork and collaboration – working well with people who think differently

• entrepreneurial creativity – seeing gaps and building new solutions

• reflection and lifelong learning – staying curious and ready to grow.

These capabilities are the genuine scarcity in today’s market. They are complements to AI, not substitutes, which is why their wage returns hold or climb.

What universities can do right now

1. Audit courses: if ChatGPT can already score highly on an exam, the marginal value of teaching that content is near zero. Pivot the assessment toward judgement and synthesis.

2. Reinvest in the learning experience: push resources into coached projects, messy real-world simulations, and ethical decision labs where AI is a tool, not the performer.

3. Credential what matters: create micro-credentials for skills such as collaboration, initiative and ethical reasoning. These signal AI complements, not substitutes, and employers notice.

4. Work with industry but keep it collaborative: invite employers to co-design assessments, not dictate them. A good partnership works like a design studio rather than a boardroom order sheet. Academics bring teaching expertise and rigour, employers supply real-world use cases, and students help test and refine the ideas.

Universities can no longer rely on scarcity setting the price for the curated and credentialed form of information that used to be hard to obtain.

The comparative advantage now lies in cultivating human skills that act as complements to AI. If universities do not adapt, the market – students and employers alike – will move on without them.

The opportunity is clear. Shift the product from content delivery to judgement formation. Teach students how to think with, not against, intelligent machines. Because the old model, the one that priced knowledge as a scarce good, is already slipping below its economic break-even point.

References

1. ^{^} add between US$2.6 trillion and $4.4 trillion (www.mckinsey.com)
2. ^{^} fallen by about a third (www.theguardian.com)
3. ^{^} slid from roughly 68% to 53% (www.nga.org)
4. ^{^} David Autor (labourmarketresearch.springeropen.com)
5. ^{^} Daron Acemoglu (www.aeaweb.org)
6. ^{^} Lightcast (lightcast.io)
7. ^{^} American Enterprise Institute (www.aei.org)
8. ^{^} economist and cognitive scientist (www.nobelprize.org)