The Roundup - Edition 113

A message from 
Rich Jones, Head.

Last week I did Morning Talk in the Junior School. To my shame, it had been a long time, and Katie quite rightly asked me to make amends. I have always liked Junior and prep schools, but now my bunch have gravitated towards the Senior School, I have lost some of those everyday interactions with younger students. Something I am determined to rectify.

Since then, I have had lunch with different groups of Junior School students a couple of times. I have to say, rarely have I laughed quite as much as I did eating with them.

What struck me most was just how confident, energetic, opinionated and engaged they are. We discussed swimming, the subject of my talk last week, the importance of singing together, the pros and cons of rugby and netball, why tomato ketchup is the greatest culinary invention in the history of humanity, ChatGPT and the Romans.

An eclectic series of discussions, I grant you.

But I was struck by their passion and engagement with almost every topic. All had views, opinions, and ideas and there was zero shyness. They were relatively scathing about AI and very effusive about ketchup, and it got me thinking about the importance of discovery and exploration without always rushing to find the answer.

I had been reading an interview with Ying Xu, an Assistant Professor at Harvard, about AI and children. It would be easy to turn this into another piece about the dangers of AI, but I don’t think that is particularly helpful, nor actually what she argues. AI can be an extraordinary tool for learning. It can open doors, answer questions, provide explanations and allow children access to knowledge in ways we could barely have imagined even a few years ago.

Indeed, Xu’s own research suggests children can learn effectively from AI when it is designed with learning in mind. AI that asks children questions, gives them hints and encourages them to think can support learning rather than diminish it.

But she also uses a phrase I rather like: “productive struggling”. Sometimes there is value in not knowing.
In being stuck. In trying something and getting it wrong. In arguing with the person sitting opposite you. In following a thought down a completely pointless avenue and discovering, twenty minutes later, that it wasn't pointless at all. Something I am highly experienced in!

There is actually some interesting research behind this. The educational researcher Manu Kapur has explored something he calls “Productive Failure”. In one study, students were given problems they had not yet been taught how to solve. Unsurprisingly, they struggled. Some failed altogether. Another group were taught the method first and then practised it.

You might assume the second group learned more. In fact, while both groups learned the procedure, those who had wrestled with the problem first developed a better conceptual understanding and were better able to apply their learning to unfamiliar problems later.

The struggle had done something. And perhaps there is something important for schools in that.

For a long time, education has understandably placed enormous value on knowing things. Knowledge matters, of course it does. Children need facts. You cannot think critically about something if you know absolutely nothing about it.

But we are entering a world in which simply obtaining an answer is becoming astonishingly easy.

I can ask a device in my pocket who the fourth Roman Emperor was, how photosynthesis works, the population of Peru or how to calculate the hypotenuse of a triangle and have an answer within seconds. AI takes that another step further. It can explain it to me, simplify it, test me on it and then explain it differently when I still don't understand.

That is remarkable. I don't think we should fear it. I think we should teach children how to use it brilliantly.

But it does make me wonder whether one of the jobs of a school is increasingly to protect some of the space before the answer.

The wondering. The wondering is so very very important. 

The “I don’t know.” The “what happens if...?” The half-built thing that collapses. The experiment that doesn't work. The story that takes an unexpected turn. The experient and picture that look wrong. 

And so we all must acknowledge something that schools can sometimes find difficult: play is messy.

Children take risks. They make mistakes. They change direction. They become frustrated. They don't always produce something easily measurable at the end of it.

There is a wonderful tension here. Schools, understandably, like things to be purposeful. We have timetables, curricula, objectives, assessments, and outcomes. Play doesn't always behave itself quite so neatly.

But perhaps that is precisely why we need some of it. And I don't just mean play in the Junior School.

There is a temptation to think that play belongs to childhood and that, as children grow older, education should gradually become more serious. I am not convinced. A Sixth Former experimenting in an art studio, musicians improvising together, pupils arguing about the motives of a character in a novel, somebody trying to make something work in DT, a group devising a piece of drama, a scientist wondering why an experiment produced a completely unexpected result. These are all forms of play too.

They are all moments when the outcome isn't entirely predetermined.

Because the danger of an education system overly focused on outcomes is that children become very good at asking, “What do I need to do to get this right?” rather than, “I wonder what would happen if I did this?”
Those are quite different questions.

The first has its place. Of course it does. Sometimes there is a right answer, sometimes things simply need to be learned and sometimes success requires discipline, repetition and hard work.

But I don't ever want us to educate the second question out of children. Perhaps that was what I was noticing over lunch with our Junior School pupils.

Nobody was waiting for me to tell them what to think. Nobody was particularly concerned whether their contribution was the “correct” one. They wanted to tell me things. Challenge things. Ask things. Make connections. Disagree with one another and, occasionally, disagree quite enthusiastically with me.

There was a joy in finding things out. And that, I think, is something St Chris should protect fiercely.

Our children need knowledge and they need technology. They need to understand AI and become confident, thoughtful and critical users of it. Pretending it isn't there would be rather like schools in the 1990s deciding the internet probably wasn't going to catch on.

But alongside that, they need mud, paint, music, laboratories, workshops, stages, animals, sports pitches, books and other human beings.

They need teachers who sometimes answer a question with another question. They need to discover things nobody told them to look for and they certainly need time occasionally to follow the interesting tangent.

One of the things that struck me most in Xu's research was that children were perfectly happy talking to AI, and could learn from doing so. But when they talked to another human being, they were more likely to steer the conversation themselves, ask follow-up questions and volunteer their own thoughts.

That seems important. Because education is not simply the exchange of information. If it were, AI might eventually become extraordinarily good at providing much of it.

It is, occasionally, a Head sitting down for lunch with a bunch of Junior School children and finding himself involved in a passionate debate about Romans, rugby, artificial intelligence and condiments.

Perhaps one of the increasingly important jobs of a school isn't simply to help children find answers. It is to make sure they never lose the desire to find things out.

And, for the record, they were also completely wrong about ketchup. Horrid stuff.