The Last Word: Before we start deploying AI in schools, perhaps we should learn more about what, exactly, we want to achieve by doing so.
The most dangerous phrase in education today may be ‘AI-ready’. It sounds practical, modern, reassuring. It flatters governments, schools and universities into believing that readiness is a procurement issue: buy the tools, train the staff, update the policy, reassure the parents, move on.
I am not sure I am convinced though. We are speaking with far too much confidence about a technology we do not yet fully understand. That does not mean AI should be kept out of education. It means education should be much clearer about what it is protecting before it decides what it is adopting.
The World Economic Forum’s new report, Shaping the Future of Learning: Education Readiness for the Age of AI, is valuable because it does not treat technology as destiny. It frames readiness around whether the conditions are in place for responsible and effective AI adoption, and whether AI strengthens learning, equity and trust rather than undermining them. That is the right frame. But my concern is simpler and sharper: do we know enough about human intelligence to redesign education around artificial intelligence?
At the moment, I do not think we do. We are in the fog. We do not yet know what routine AI use will do to attention, memory, curiosity, confidence, originality or the slow discomfort of working something out for oneself. We can see the convenience, the efficiency, the sales decks. What we cannot yet see clearly is the long-term effect on the learner.
Education has always been tempted by shortcuts: teaching to the test, rewarding neat answers over messy reasoning, confusing memorisation with understanding and presentation with substance. AI gives that old temptation a superb new interface. Why wrestle with a blank page when a machine can draft? Why sit with uncertainty when a system can respond instantly? Why form an argument when one can be assembled, polished and submitted before the lesson ends?
The problem is not that pupils might cheat. That is the small scandal. The larger scandal would be teaching a generation to confuse fluency with thought. A good answer is not the same as a good mind. A neat paragraph is not the same as judgement. A correct summary is not the same as understanding.
AI cannot be brave
This is where we need to be precise about the difference between AI and HI, artificial intelligence and human intelligence. AI can process, predict, imitate, generate and optimise. Human intelligence can care, doubt, notice, choose, regret, take responsibility and change its mind for a reason. AI can produce a lesson plan on empathy. It cannot see the child who has gone quiet. It can write about courage. It cannot be brave.
The Cambridge Advance paper on three in-demand skills in the age of AI makes the same point from another angle. It argues that as AI and data skills rise, communication, problem-solving and creative thinking become more vital, not less. Its most useful warning is that we cannot rely on AI for problem-solving: outputs may look plausible or impressive while still being inaccurate or inappropriate, which means humans must frame the problem, judge the answer and take responsibility for the result.
That should unsettle education leaders. If those are the skills we need most, are we teaching them with enough seriousness? Too often, creativity is still treated as decoration, communication as presentation polish and problem-solving as an exam technique. They are not extras. They are the core curriculum for a world in which machines will increasingly supply the first draft.
So the future-ready school is not the most automated one. It is the one with the discipline to say: here AI helps; here it harms; here it must wait. It teaches pupils to use the tool, but also to challenge it, ignore it, and sometimes begin without it.
The fog is not a reason to stop. It is a reason to move more intelligently. So, before education becomes AI-ready, it must become more human-ready.
Photo: Dreamstime.