AI in the classroom: friend or threat?

Few topics provoke as much disagreement among teachers as the arrival of generative AI in education. On one side, enthusiastic adopters describe it as the most powerful tutoring tool ever invented: patient, available at all hours, capable of explaining the same concept fifteen different ways until the student understands. On the other side, sceptics worry that it will damage the basic skill of independent thought. Both positions contain something important, and both, in their pure form, are probably wrong. The case for AI is strongest in the area of personalisation. A single teacher with thirty students simply cannot give each one custom attention every day. A well-prompted AI assistant can. Used carefully, it can also help students who are too shy to ask their teacher questions in front of the class. The case against AI is strongest in the area of struggle. Learning, especially difficult learning, requires productive struggle: the kind of slow, frustrating thinking that builds genuine understanding. If students hand every challenging problem to an AI and copy the answer, they get the marks but not the learning. The schools that handle this transition well are not banning the technology, but they are not embracing it uncritically either. They are teaching students when AI helps and when it harms. Used as a tutor, AI extends education. Used as a substitute for thinking, it hollows it out. The line between the two is sometimes blurry, but it is too important to ignore.