What Should Education Teach When AI Can Do the Thinking?

I was reading a recent interview with Howard Gardner in Education Week, and one of his observations stayed with me long after I closed the tab.

Consider a student who can use AI to break down a difficult problem, process information from hundreds of books, weigh different viewpoints, explore possible solutions, draft an essay, or even create an original piece of work, all in a matter of seconds. If AI eventually becomes exceptionally capable at these tasks, do humans still need to spend most of their educational lives learning to perform them in exactly the same way? That is the question at the centre of Gardner’s argument.

For generations, education has been built around developing a student’s ability to acquire knowledge, analyse information, solve problems, communicate ideas, and produce intellectual work. These abilities mattered because human beings had to do these things themselves. AI is quietly changing that equation, and plenty of observers are already asking whether AI is actually improving education or simply layering automation on top of it.

To be clear, Gardner is not suggesting that students should stop learning. They will still need solid foundations in reading, writing, mathematics, science, and other core areas. Without that foundational knowledge, a student may not even be able to use AI intelligently, let alone judge whether its output is any good. Sound study habits, from taking better lecture notes to reading with a critical eye, still have a genuine place in the classroom.

But Gardner’s point is subtler: perhaps education should gradually place greater emphasis on the abilities that machines cannot easily replace, or the qualities we value precisely because they are human.

A machine can hand you an answer, but it cannot decide what matters to you. It can generate ten possible solutions to a problem, but a human still has to choose which one is desirable, ethical, or worth pursuing. It can write a story, but the deeper question remains: why tell this story in the first place?

The qualities that survive this shift are the ones we have always valued most, but rarely made the centre of the curriculum:

• 🔹 Judgment, deciding which of many possible answers is the right one to act on.
• 🔹 Curiosity, wanting to ask questions in the first place, not just receive answers.
• 🔹 Ethics, recognising what is fair, responsible, and worth doing.
• 🔹 Empathy, understanding what another person feels and needs.
• 🔹 Relationships, building trust and meaning with real people.
• 🔹 Purpose, deciding what is worth contributing to the world.

These are not soft extras. In a world where machines can think faster than we can, they may become the main thing worth teaching.

Another part of Gardner’s thinking that struck me is his vision of a more individualised education system. Today, students generally move through a predetermined sequence of grades and subjects for many years, with everyone following roughly the same structure regardless of their individual interests, strengths, or ambitions.

But imagine a different model, guided by research from organisations like Edutopia on student-centred learning. Once students have acquired the essential foundations, they could spend far more time exploring the subjects they genuinely care about. AI could become a powerful personal learning assistant, helping them investigate questions, find resources, practise skills, and explore increasingly complex ideas.

One student might go deep into medicine. Another into history. Someone else might want to build businesses, create films, write novels, study philosophy, or design technology. Instead of education being primarily about getting everyone through the same curriculum, it could become much more about helping each person discover what they want to understand and what they want to contribute.

Teachers would remain essential, but their role would evolve. They could spend less time delivering information and more time mentoring, challenging students, developing judgment, encouraging curiosity, and helping young people navigate the human side of learning.

There is a warning hidden inside this idea, and it matters enormously. If AI does the thinking for students, they may gradually lose the ability, or the willingness, to think for themselves. This is where the conversation connects to the broader hidden risks of AI in education.

There is a huge difference between using AI to extend your thinking and using AI to avoid thinking:

If a student asks AI to solve every problem, write every essay, and generate every idea, school could become easier while the student becomes intellectually weaker. We have already seen early examples of exactly this, including a student who never studied for their degree and simply leaned on AI agents instead.

So the challenge is not simply deciding whether to use AI. The deeper challenge is deciding what we should keep asking humans to learn and practise, even when AI can do it faster.

The future of education may not be about preparing students to compete with machines at everything machines can do. It may be about helping students develop the judgment to know when to use machines, the wisdom to question their answers, and the human qualities that make knowledge meaningful in the first place.

After reading Gardner’s ideas, I was left with a single question: if AI can eventually do much of the intellectual work that schools have traditionally prepared us to do, should education continue in essentially the same way, or is it time to rethink what we are educating humans for?

What do you think? Is Gardner right?