The AI Question Every Parent Is Asking
How our approach at Radnor maximises your child’s learning journey whilst preparing them for the world of tomorrow.
The hottest topic for parents I meet at school presently is AI. They want to know how we are building it into the curriculum, how we are leveraging its power to give children an advantage, and how teachers are using it to personalise learning. These are all valid ambitions and questions I am also drawn to.
As adults, many of us use AI regularly in our workplaces. Indeed, we recently launched an internal focus on AI for our staff to ensure we remain on the cutting edge of this new technology. However, what is less clear is whether the obvious productive advantages of AI for adults transfer safely to a learning environment for children whose neural pathways are still developing.
Adults possess a significant advantage when it comes to using AI. We typically have existing knowledge schemas and highly developed executive functions. We know that AI can boost productivity and we tend to treat it as a decision support tool rather than a decision replacement tool. We can often spot hallucinations and edit prompts in an iterative way to arrive at a better solution. In short, for adults, the potential reward of AI is high while the cognitive risk is low.
Most students, on the other hand, are still acquiring foundational knowledge. They are literally learning day by day how the world works. Their critical thinking skills are nascent. They are just starting to wake up as children become teenagers, but they are certainly not fully developed. When students are given any type of printed or screen based information, they simply lack the subject mastery to evaluate its quality.
The physiological changes in the developing brain are profound. As children turn 10, their limbic system, which processes rewards and emotions, develops very quickly. By 13, the brain’s processing speed is accelerating rapidly so we can cope with an increasingly complex world. However, the prefrontal cortex remains under construction. This area controls logic, self-regulation, and impulse control. Teenagers can cope with abstract problem solving, but they struggle to evaluate risks or spot flawed information. Around the age of 16, this part of the brain matures significantly. They can understand long term consequences better and acquire higher level critical thinking skills that only improve over the next decade.
These developmental stages coincide with the stages of our school. At the top end of Prep and into Lower School, children need lots of guidance on what they should be learning and how they should be behaving. They do not yet have the ability to discern between stronger and weaker evidence bases, and their metacognition skills are typically embryonic.
As children move into Middle School, they start on the GCSE pathway with more complex subject specific knowledge. They are rapidly starting to understand the world around them and acquire mental frameworks which can be applied in various situations. Those of us with teenage children will know they still struggle to make decisions for themselves and need our direction. As children reach the end of their GCSEs by Year 11, they become more logical, more self-governing, and more able to understand the long-term implications of their actions.
Assuming this basic biology of human development is unlikely to be altered by AI, we need to understand the cognitive risk and reward profile of when we introduce this technology in schools. Clearly, putting a five-year-old in front of an AI screen all day is not a good idea. However, completely banning AI for young people up to the age of 18 is also unrealistic. So, the question is at which age Radnor should encourage its use?
We are presently in the foothills of AI. Few of us could claim to be experts, and those who do often simply do not understand it enough. Anecdotal evidence from Radnor and discussions with other Headteachers tell us there is no doubt children as young as eight are using it regularly. This happens either with the tacit support of parents or simply through their own exploration.
Children are discovering that it can do all sorts of amazing things. If they are struggling with Maths, AI can solve the problem for them, so all they need to do is copy the answer in their own handwriting. If they are not quite sure how to structure an essay, AI will give them the framework. If they are not sure how to plan a debating speech, AI can print off PhD level arguments in seconds to make them sound like a genius.
The problem with all of this is that learning cannot be short circuited. It does not comply with the same productivity laws that govern most adults in the workplace. Children are growing and developing. If you take away the struggle and the challenge, you take away the growth and increase the risk of something going wrong.
Evidence is emerging that global intelligence is falling for the first time in human history as screens crowd out reading, learning, and thinking. These lagging indicators are the result of more screen time not less, more technology in the learning process not less, and new ways of convincing ourselves that just because something feels easier, it must be a good idea. AI will undoubtedly shape the future for all of us however the risk to children should not be underestimated. There is a growing consensus that phones are a net negative for children. We should all be alive to the real risk that AI super charges this and might, ironically, for the first time make our children’s generation less intelligent than our own.
