Youthful Futures · Essay 4 of 12
Personalised Everywhere, Standardised at School
Young people live in systems that constantly adapt to them, until they enter one of the most consequential systems in their lives.
A teenager can open Spotify and hear music selected from millions of songs based partly on what they listened to yesterday.
YouTube reconstructs itself around what they watch.
Games continuously adjust difficulty.
Social feeds learn what holds their attention.
An AI tutor can explain the same mathematical idea through football, cooking, diagrams or a conversation and try again when the first explanation fails.
Then the teenager goes to school.
Thirty people sit in a room.
Chapter six begins today.
The examination is on Friday.
Everyone gets forty-five minutes.
There are good reasons schools work this way.
A society needs common knowledge.
Teachers cannot construct thirty entirely different schools inside one classroom.
Shared experiences matter.
Standardised assessments can expose inequalities that would be easy to hide if every learner were evaluated differently.
A common qualification can allow a student from an unknown school to demonstrate competence to a university or employer that has never met them.
Standardisation is not simply evidence that education failed to modernise.
But something around it has changed.
Young people are growing up in environments where personalisation increasingly feels like the normal relationship between a person and a system.
In education, perhaps the most consequential part of their lives, adaptation often stops precisely when the stakes are highest.
The tension is becoming sharper as generative AI enters learning.
A student can now ask for an explanation at exactly the level they understand, request another example, change language, generate practice questions and continue asking without embarrassment.
In principle, that approaches something educators have wanted for centuries: responsive individual support at very low marginal cost.
But technology's history in education should make us cautious.
UNESCO's global review of education technology argues that technology improves some kinds of learning in some contexts, but that benefits can disappear when technology is used excessively or without qualified teachers. Its core recommendation is not to maximise technology, but to decide whether a technology is appropriate, equitable, evidence-based and sustainable for the educational problem at hand.1
The world is already experiencing the backlash.
By March 2026, UNESCO counted 114 education systems with national restrictions on mobile phones in schools, representing 58 per cent of countries worldwide. That was up from 24 per cent in June 2023.2
There is evidence behind some of this concern.
In PISA 2022, 30 per cent of students across OECD countries reported becoming distracted by digital devices in mathematics lessons.3
So the future is unlikely to be every child sitting alone with an infinitely personalised machine.
That would solve one problem by creating another.
Education is not merely information entering an individual brain.
It is also the experience of encountering other people.
Waiting while someone else speaks.
Explaining your thinking.
Discovering that another person interpreted the same story differently.
Negotiating disagreement.
Being bored occasionally.
Performing.
Failing publicly.
Helping somebody who understands less than you.
Asking for help from somebody who understands more.
UNESCO makes this tension explicit. Technology can support learning, but should not supplant the human connection on which teaching and learning depend.1
The harder question is therefore not personalised or standardised.
It is deciding what should be which.
Perhaps every learner does need some common intellectual ground.
Perhaps multiplication should not become optional because a child prefers dinosaurs.
Perhaps citizenship requires encountering histories and ideas you did not personally select.
But why must time be standardised?
Why must evidence of understanding always be standardised?
Why does one child have to stop because the lesson is over while another is expected to move on despite not understanding it?
And if two students can demonstrate the same mathematical reasoning through different contexts, how much should the context matter?
AI makes these questions harder to avoid because it exposes how much variation was always possible.
The constraint was partly economic.
One teacher could not continuously generate thirty explanations, thirty practice sets and thirty sequences of feedback.
A machine increasingly can.
That does not mean the machine should decide what learning is.
It means some of our old explanations for why education must look a certain way are becoming weaker.
Assessment may eventually become the most uncomfortable part of this conversation.
We increasingly tell young people that learning can follow their curiosity, pace and interests.
Then, at the moment society decides whether the learning "counts", we often compress everybody back into identical conditions.
There are reasons for that too.
Personalising assessments could introduce bias, make comparison difficult and allow wealthy families to negotiate easier pathways.
Standardisation sometimes protects fairness precisely by refusing to personalise.
But there is a difference between common standards and identical demonstrations.
We should probably spend more time thinking about it.
The future of education may not be a personalised curriculum for every child.
It may instead be a system with a common intellectual core but far more flexibility in the paths, pace, practice, support and evidence through which students reach it.
That requires something harder than buying software.
It requires answering a philosophical question that technology has allowed us to postpone.
Which parts of education should belong to the individual, and which should belong to all of us?