Why AI Cannot Fix an Education System Built for 1920

The challenge facing the education system

The debate about AI in education misses the point entirely. The issue isn’t the technology. The issue is the education system itself, a system that has remained largely unchanged for over a century.

A System Frozen in Time

Our education system hasn’t changed much since the 1920s really. We’ve just added different subjects or incorporated slightly different techniques, tweaking around the edges while the fundamental structure remains intact. Research highlights that students learn more through workshops and activities, through the actual experience of learning, but instead we teach in a didactic way. Not because teachers want to, but because the sector wants everything standardised, measured, and controlled.

AI could give teachers back so much time, freeing them from administrative burdens to focus on what actually matters in education. Instead, schools are aiming to make the system more homogeneous, more efficient at doing exactly what we’ve always done. This isn’t all schools, you have some great teachers and great schools doing brilliant work, but this is the system at large, the dominant pattern that shapes educational practice across the country.

The Deeper Problem

Honestly, education at all levels has faced these challenges for decades before AI arrived on the scene. Over-commodification has diminished learning, reducing teaching to a mere shell of what it should be. Standardisation has prioritised bureaucracy over creativity, leading many teachers to leave not for lower pay but for less favourable conditions, for environments where they can no longer actually teach in ways they know are effective.

This mirrors a broader trend in how we approach complex problems. We favour quick fixes over substantial investments that yield long-term benefits, choosing the path that looks efficient on a spreadsheet over the one that actually works. Industry shapes learning not for individual benefit but to reduce their own training costs, treating education as a pipeline for pre-trained workers rather than a space for human development. CPD often functions more as a marketing tool than an educational one, box-ticking rather than genuine professional growth.

Schools focus on cost-cutting rather than improving student experience or learning outcomes. The shift from student-focused to accountancy-focused education shows up everywhere, in every teaching method, every policy decision, every strategic priority. We’ve moved from asking what helps students learn to asking what costs less to deliver.

We talk about the next generation critically, pointing to their lack of attention and declining standards, but the system they’re being taught in just doesn’t prepare them adequately for the future as it stands. We blame students for failing to thrive in conditions designed to produce compliance rather than understanding.

What Real Change Looks Like

To change it we need fundamental shifts in how we assess learning. Rather than looking for recall and comprehension, testing whether students can repeat back what we’ve told them, we need to evaluate thoughtfulness and genuine understanding. What questions did they ask in approaching a problem? What was their method for investigating it? How did they evaluate their approach? What did they recognise as missing from their analysis?

We should assess core skills through competency exercises that demonstrate actual ability rather than memorised knowledge. Embed project and group activities more thoroughly throughout curricula, making collaborative learning the norm rather than the exception. See thoughtfulness, those building blocks of philosophical inquiry, and critical thinking, the practices of evaluation and reflection, as two halves of a whole rather than separate skills to be tested independently.

The textbook as a resource should still be valid, but the method of teaching it should have died a death a long time ago. We need to move toward a more facilitative model of education that guides rather than instructs, that informs students why the resources are valuable and doesn’t expect everything to be memorised and regurgitated on command.

Because unfortunately many people weren’t able to memorise information before, but they could understand it when given the chance to engage with it properly. We haven’t really considered understanding in this process, only comprehension, treating these as interchangeable when they represent fundamentally different ways of knowing. That distinction matters enormously for how we think about what education should accomplish.

The Real Question

We need the system to evolve to really incorporate AI in meaningful ways. And honestly, we needed this evolution before AI even arrived.

The question isn’t whether students should use ChatGPT for their essays. The question is whether we’ll finally build an education system that prepares humans for meaningful lives, for genuine participation in a complex world, or whether we’ll just automate the inadequacies of 1920 with the efficiency of 2025, doing the wrong thing faster and more consistently than ever before.