I have spent a large part of my working life in and around universities, and I remain rather attached to the idea of the university. That probably explains why I have become so irritated by what we have allowed many of them to become.
I despise the corporate model of education.
That is deliberately stronger than saying that I dislike managerialism. Universities obviously need managing. They employ thousands of people, occupy expensive buildings, handle public money and student fees, and operate within regulatory frameworks. Somebody has to make budgets add up.
My objection begins when management ceases to support the purpose of the university and starts redefining it.
Listen to the language
One only has to listen to the language.
Universities now have stakeholders, portfolios, strategic priorities, transformation programmes, business partners, customer journeys, performance indicators, engagement strategies and market positions. Senior job titles increasingly sound as if they have been borrowed from a corporate restructuring handbook (one suspects the next edition will rebrand the rest of us as Customer Experience Associates).
Perhaps this seems trivial. I don't think it is.
Language has a habit of revealing what an organisation thinks it is for. If universities spend long enough talking like businesses, it should hardly surprise us when they start behaving like them.
The word "stakeholder" particularly bothers me.
Of course students have an interest in universities. So do employers, government, taxpayers, professional bodies and the communities in which universities operate. There is nothing objectionable about listening to any of them.
The difficulty comes when stakeholder expectations begin to determine the intellectual purpose of education.
Existing stakeholders represent existing arrangements. Employers naturally describe the skills they currently need. Governments have current economic priorities. Regulators measure things that can realistically be measured. Students themselves quite reasonably want qualifications that will help them earn a living.
A university has to listen to all of this without becoming captive to it.
There should still be room for someone to say: perhaps the prevailing assumption is wrong.
That seems to me one of the defining privileges of a university.
Newman's nineteenth-century The Idea of a University belongs to a very different social world, and I would certainly not wish to resurrect the institution he knew. Yet his insistence that knowledge had value beyond immediate utility now feels surprisingly radical. The Humboldtian tradition, for all its own historical baggage, similarly placed intellectual independence close to the centre of university life.
Compare that with the contemporary preoccupation with graduate outcomes and employability.
I am not dismissing employability. For many students, particularly those taking on substantial financial commitments to study, getting a decent job afterwards matters enormously. Universities have a responsibility to take that seriously.
I object when employability becomes our definition of education.
There is a difference between educating someone who subsequently becomes employable and designing an education around the immediate requirements of employers.
Metrics change behaviour
The distinction has become especially important in the UK because universities now compete remarkably aggressively with one another.
League tables, the National Student Survey, graduate outcomes, continuation rates, research measures and assorted institutional benchmarks all feed into reputation. Some provide genuinely useful information. I would be very wary of an argument that allowed poor teaching to hide behind the language of academic freedom.
Yet metrics change behaviour.
This is particularly visible among universities outside the elite end of the rankings. Oxford or Cambridge can survive a disappointing movement in a table that most prospective students have never heard of. A university struggling for recruitment cannot treat reputation quite so casually.
The temptation to manage the metric becomes considerable.
Courses are marketed against competitors. Student experience is monitored. Recruitment is segmented. Retention becomes strategically important. Resources follow the things an institution believes will improve its competitive position.
At some point it becomes reasonable to ask whether universities are competing to provide better education or competing to produce better evidence that they provide better education.
Those are not always the same activity.
AI and the disappearing employer
There is another reason I think this model is becoming increasingly difficult to defend.
Artificial intelligence has arrived at precisely the point at which universities have become most enthusiastic about preparing students for large organisations.
For much of the post-war period, the logic was understandable. A graduate entered a profession, joined a substantial employer and developed a career within an organisation. Universities supplied educated labour to an economy that needed large numbers of people to perform increasingly specialised forms of work.
I am not convinced that this will remain the dominant model.
I cannot prove that AI will result in fewer major employers. Nobody can. Predictions about the future of work have an impressive history of being wrong.
Nevertheless, I think the possibility deserves considerably more attention than it currently receives.
AI lowers the cost of expertise.
Work that previously required several people can increasingly be undertaken by one person working with software. A small company can have access to programming, research, statistical analysis, design, translation and administrative capability that would once have required employees or external specialists.
The technology remains imperfect. Sometimes spectacularly so. That does not alter the direction of travel.
If these systems continue to improve, I suspect we will eventually see fewer very large employers of knowledge workers, or at least considerably leaner versions of the organisations we know today. Some work will disappear, some will change, and entirely new occupations will emerge. More interestingly, small organisations may become capable of doing things that previously required considerable scale.
That should concern universities, because we continue to talk about students becoming "work ready" as though the nature of work were reasonably settled.
What does work ready mean for somebody graduating in 2035?
Which employer are we preparing them for?
Will that employer still require several thousand graduates?
Will the profession exist in anything resembling its present form?
Perhaps the more useful graduate will be somebody capable of operating without the security of a large employer at all.
This is where I think the corporate university has got itself into a peculiar position. It has increasingly modelled its own culture on large organisations while educating students to enter large organisations, just as technology is beginning to give individuals and small groups capabilities that previously belonged almost exclusively to large organisations.
We may have copied yesterday's workplace just as yesterday's workplace begins to disappear.
What we should be teaching now
This also changes my view of what we should be teaching.
There has been a lot of discussion about integrating AI into curricula. Much of it concerns AI literacy, appropriate use, assessment and the skills employers might expect. All sensible enough.
I think the more difficult question is what education is for when access to knowledge is no longer the scarce resource it once was.
A student can already ask an AI system to explain a theory, summarise a paper, write code, suggest an experimental design or criticise an argument. The answers vary enormously in quality, but the capability exists.
Teaching students simply to reproduce information therefore looks increasingly inadequate.
The graduate I would want to educate is someone who can look at a confident AI answer and think, "I'm not persuaded."
Someone who spots the assumption that everybody else missed.
Someone prepared to ask an awkward question in a meeting.
Someone capable of following an argument into places that have no obvious commercial application.
Someone, occasionally, who becomes a complete intellectual nuisance.
I mean that as a compliment.
Universities need iconoclasts. Society needs them too.
The history of scholarship is full of people who became interesting precisely because they did not behave as obedient representatives of the intellectual establishment around them. That does not mean every contrarian is a genius. Universities are quite capable of producing cranks (a category I suspect this essay may nudge me toward, though I prefer the term “constructively obstinate”). Critical thinking requires evidence as well as independence.
But excessive institutional conformity carries its own danger.
If students learn that success means reproducing approved language, satisfying rubrics, meeting learning outcomes, pleasing stakeholders and preparing themselves to slot neatly into an organisational hierarchy, we should not be surprised if they become extremely competent at complying.
I would like higher education to aim considerably higher than that.
Freedom of thought is not an ornamental feature of university life. It is one of the reasons universities deserve to exist.
Students should encounter ideas they disagree with. Academics should be able to pursue questions whose commercial value is obscure. People should occasionally leave seminars having changed their minds. Institutional assumptions should be open to challenge from people far below the top of the organisational chart.
This will become harder if universities increasingly resemble the corporations with which they seek to engage.
Corporate organisations have understandable reasons to dislike unpredictability. Universities ought to have a rather higher tolerance for it.
And AI makes the issue urgent.
If routine intellectual work becomes easier to automate, the premium on obedience falls. There is little point educating a human being to behave like a predictable information-processing system when we can manufacture predictable information-processing systems by the million.
Human advantage may lie somewhere messier: judgement, originality, doubt, argument, curiosity and the willingness to abandon a perfectly respectable answer when the evidence points elsewhere.
Those qualities do not sit comfortably on a dashboard.
Perhaps they never did.
Conclusion
So when I say that I despise the corporate model of education, I am not asking for universities without budgets, accountability or management. Nor am I suggesting some romantic return to a mythical golden age when academics wandered around cloisters thinking great thoughts while somebody else quietly paid the bills.
I want us to remember that management is the machinery of the university. It isn't the purpose.
The purpose is education and scholarship.
And if AI changes employment in anything like the way I suspect it will, our students will need considerably more from us than training in how to become satisfactory employees.
Some of them will have to create their own work. Some will establish businesses with very few employees. Others will move repeatedly between occupations as technologies alter what can be automated. They will have to decide when machines are useful and when they are talking nonsense.
Above all, they will need the confidence to think for themselves.
If a university cannot teach that, I am increasingly unsure what claim it has left to the name.