Two news stories caught my eye this week.  Wayve, a British company developing an AI system for autonomous driving, is raising $1bn and poised to disrupt the emerging driverless car market with its machine learning algorithms.  The other quotes a Professor at Trinity College Dublin who laments the risk of AI for the undergraduate masses as they increasingly harness its large language models to ‘pen’ essays in topics they know very little about. 

This complex relationship between risk and reward for AI is increasingly seen in our schools today. 

For years, education has remained the final slice of society broadly untouched by technology.    We’ve seen agriculture move from brute force to mechanisation, and in the process shed jobs and grow yields.  Manufacturing, too, has witnessed huge increases in productivity as robots create cheap products for the global market, and health care has seen a revolution in diagnosis and treatments as big data is leveraged to spot problems early and treat them accordingly. 

Yet, education has remained broadly static.  I can remember completing my first Masters in Education on the topic of how schools use technology and raising an eyebrow when my tutor proclaimed that IT is about to ‘revolutionise’ the classroom.  That was in 2005 and I think it’s safe to say that almost 20 years later not much has changed.   

Could AI be the missing piece of the puzzle with the potential to finally disrupt our schools? 

For the teacher, AI is already creating lesson plans in seconds, marking quizzes in a flash and helping to personalise learning.  These are significant productivity improvements which mean teachers gain time with children and should not be overlooked. 

For the child, AI is an altogether different prospect.  The idea of learning being simplified is hugely problematic - learning should be a challenge, it should make our brains ache with the effort and we should feel proud of our accomplishments when we are finished.  Anything which bypasses this struggle is not positive for child development and is likely to exacerbate the problems seen in young people today.  The output of learning is not the words on a page but the understanding. 

Universities are concerned.  Professor Laurent Muzellec at Trinity College Dublin, said AI is ‘de-skilling’ learners and is hugely problematic for their progress on undergraduate courses.  As he points out, students today, ‘literally don’t need to know anything to use the technology’, it has the power to produce an entire thesis in minutes and plagiarism tools are increasingly bypassed due to the unique nature of the output. 

So what next for AI in schools?  We need to stay alert; in time, AI may be able to supplement learning in the same way that videos on the Khan Academy supplement a pupil’s understanding of quadratic equations.  If we get to the point where AI is able to seamlessly create content based on the individual learning needs of every pupil, it could be huge. But it’s unlikely to ever completely change the relationship between a teacher and child. We all need a human guide, a guru to help us grow, and for the time being I think it is difficult to see ChatGPT or Microsoft’s Copilot providing that relationship.  But you never know….   

David Paton is Head of Radnor House Sevenoaks