Preparing students for the AI age

On Teachers' Day, I came across a question on social media: What should be the greatest quality of a teacher? My response was: a teacher's greatest quality should be that he or she does not kill the student's curiosity. Students should be taught that knowledge begins with a question and, in some sense, ends with one too. In between, we find answers shaped by evidence, context, time and place - answers that must always remain open to scrutiny. In the age of AI, that reflection has acquired a new urgency. We need to train students to ask better questions more than feeding them answers.

For centuries, education was built around scarcity. Books were scarce. Experts were scarce. Information was difficult to access. The teacher therefore occupied a privileged position as the possessor and transmitter of knowledge. The good student was often the one who remembered most accurately what the teacher had said. AI is quietly dismantling that arrangement which, if not catered for with a better alternative, would make formal education completely redundant.

Today, a student can ask a chatbot to explain Keynesian economics, solve an equation, summarise a research paper, draft a business plan or compare competing philosophical arguments within seconds. The quality is uneven and errors remain real but the direction of travel is unmistakable. Answers, once scarce, are becoming abundant. And when answers become cheap, education built primarily around producing answers becomes increasingly difficult to defend. This does not make teachers irrelevant. It makes a certain type of teaching irrelevant.

The teacher who merely transfers information now competes with a machine that is available twenty-four hours a day, never becomes impatient and can explain the same idea repeatedly in ten different ways. Trying to preserve the old educational model by banning AI may therefore resemble protecting the horse-drawn carriage from the arrival of the automobile. The more meaningful question is not whether students use AI but what uniquely human intellectual capacities education should now cultivate.

The answer begins with questioning. AI can generate answers, but it does not absolve us from judging whether the question was worth asking in the first place. It can produce arguments, but students must still learn to examine assumptions, identify missing evidence, recognise bias and distinguish confidence from truth. It can summarise what humanity already knows far more quickly than any individual teacher, but education must still teach students when existing knowledge itself deserves to be challenged.

This changes the teacher's role fundamentally. The teacher of the future should be less an oracle and more an intellectual provocateur. Instead of asking, "Do you know the answer?", perhaps we should increasingly ask: "Why do you believe this answer? What evidence would make you change your mind? What has been left out? Who benefits if this interpretation is accepted? What would the opposite argument look like?" These are questions AI can help explore. But the habit of asking them must be cultivated.

The danger is that our education systems may respond to AI in exactly the wrong way. Worried about plagiarism and automated coursework, institutions may tighten surveillance, redesign assessments around control and become even more obsessed with proving that students produced work independently. If our response to AI is primarily policing, we may preserve the appearance of learning while missing the transformation of learning itself.

The teacher therefore has a double responsibility: to protect the student's right to question while simultaneously teaching the discipline required to evaluate answers. Perhaps the highest compliment a teacher can receive is not a student saying, "You taught me the right answers." It is a student saying, "You taught me not to stop asking questions." In the age of AI, that distinction may determine whether education remains transformative - or merely becomes an expensive way of delivering information that machines can efficiently provide.