Past, present and future: What is education for?  

In the first instalment of a two-part article, Darren Grech explores the future of education in the AI era. 

Every generation believes education is facing unprecedented change. Occasionally, it is right. We may be living through one of those occasions.  

For much of the last century, the great educational question was access. Who should be educated, for how long, and to what level? 

Education gradually ceased to be largely the preserve of those whose families could afford either the money or the time, and became a universal entitlement. Secondary education expanded, universities opened their doors more widely, and the idea that a child’s future should be determined by social background was, if never defeated, at least challenged.  

It was one of the great social achievements of the modern age. But universal education created another question. If everybody was to be educated, how could we ensure that everybody was educated well?  

And so came the age of standards.  

Targets, examinations, inspections, international comparisons and league tables assumed increasing importance. Governments wanted evidence that public investment was producing results. Parents wanted reassurance. Employers wanted young people equipped for the economy. None of this was unreasonable.  

Standards matter.  

A child who leaves school unable to read fluently, manipulate numbers confidently or express an argument coherently has been failed. There is sometimes a temptation to create a false opposition between knowledge and creativity, or between academic rigour and developing the whole person. In reality, creativity usually rests upon knowledge. Critical thinking requires something to think critically about.  

Yet measurement brings its own danger. What can easily be measured can gradually become what we value most.  

PISA (the Programme for International Student Assessment) has become perhaps the most influential global expression of this. Every few years, fifteen-year-olds across much of the world are tested in mathematics, reading and science. Governments wait anxiously. Countries rise and fall on the list. Ministers celebrate or explain.  

Consider Finland and Singapore.  

For years, Finland was regarded almost reverentially. It appeared to combine strong academic outcomes with teacher autonomy, equity, pupil welfare and a less pressurised educational culture. Yet its performance has declined steadily in recent years, a trend which has been picked up in the latest PISA results.  

Singapore offers a different model. In PISA 2025 (published 2026) it remained among the highest performing education systems across reading, mathematics and science. Its system places considerable emphasis on teacher quality, curriculum coherence, mastery and high expectations.  

It is tempting to turn the two into caricatures: relaxed Finland versus relentless Singapore. Reality is much more complicated. Both care deeply about teacher quality, and both continue to evolve.  

Perhaps the real lesson is that there is no educational Eden waiting to be copied. Culture matters. Families matter. Teacher recruitment matters. Expectations matter. What works in Helsinki or Singapore cannot simply be transplanted to Gibraltar.  

PISA should therefore be neither dismissed nor worshipped. If standards in reading, mathematics or science are falling, we should want to know. But no international test can adequately measure whether a school has helped produce a compassionate adult, an imaginative musician, a responsible citizen, or somebody with the resilience to recover when life goes wrong.  

Education has always dealt in things that can be counted and things that cannot.  

The same applies to another increasingly important question. Is education principally intended to serve the labour market, or does it possess a value independent of it?  

I have never believed we need to choose.  

Of course, education must prepare young people for employment. Schools, colleges, universities, employers and governments need to speak to one another. Economies need engineers, carers, scientists, builders, teachers, technicians, doctors and entrepreneurs.  

But a human being is not merely a future unit of economic production.  

There is value in knowing Shakespeare, even if nobody pays you to know Shakespeare. History matters because societies without historical memory are easily misled. Music and art enlarge lives. Physical education teaches lessons about effort, failure and co-operation.  

And perhaps we have become slightly uncomfortable with that word: failure.  

Inclusion matters enormously, but inclusion should not require us to pretend that everybody has won. There is dignity in taking part, in trying one’s hardest, and still discovering that somebody else was better on the day.  

Losing can teach resilience and humility. It can also teach the uncomfortable but valuable lesson that sometimes, if we want a different outcome, we need to improve. Protecting young people from every disappointment may leave them poorly prepared for a world that will not always award them first place. We can recognise effort without pretending there was no winner.  

Education should enable us to earn a living. It should also help us to live.  

Just as we have become increasingly determined to align education with tomorrow’s labour market, artificial intelligence has arrived to make that market extraordinarily difficult to predict. We have spent years talking about preparing children for the jobs of the future. Perhaps we should now have the humility to admit that we do not know what many of those jobs will be. AI can already write prose, translate languages, generate code, analyse information and solve problems that recently required considerable human expertise. Its capabilities will not stand still while today’s twelve-year- old completes secondary school.  

Does that make knowledge less important? I would argue the opposite.  

When information becomes virtually infinite, judgement becomes more important. You cannot recognise a historically absurd claim without knowing history, or faulty mathematics without understanding mathematics. You cannot effectively challenge an AI-generated answer unless you know enough to recognise when it may be wrong.  

Artificial intelligence will not eliminate the need for educated human beings. But it may change what being educated means. The premium may increasingly shift from producing answers to asking good questions; from retrieving information to evaluating it; from performing routine intellectual tasks to exercising judgement. Assessment will have to respond. If AI can produce a respectable essay in seconds, what exactly are we measuring when a student produces one unsupervised at home? Paradoxically, the most advanced technology in educational history might return us to some very old methods: conversation, oral examination, supervised writing, demonstration and debate.  

Not simply asking a student for an answer, but asking: Why do you think that? 

Darren Grech is a teacher by profession and held senior roles in the Department of Education prior to serving as Chief Secretary of the Government of Gibraltar between 2018 and 2024.