It will soon be normal for children to ask AI when they do not understand something in their homework.
Ask for a historical date and the answer appears immediately. Ask for an explanation of a science concept and the response may be more detailed than a textbook. That is useful, but it also forces us to rethink what memorisation is for.
Generative AI is already powerful at retrieving, organising, and reproducing information. Given that reality, the old assumption that education is mainly about how much a child can remember needs to be questioned from the ground up.
So what should children develop instead? The answer is not simply "the opposite of memorisation." It emerges from asking a deeper question: what was memorisation for — and what parts of that purpose can no longer be delegated to a machine?
Three Abilities AI Cannot Replace
AI is extraordinarily good at producing answers. What it cannot do is decide which questions are worth asking in the first place. The cycle of noticing a problem, framing a hypothesis, designing a way to test it, and revising based on what you find — that cycle remains irreducibly human. The abilities that hold their value in an AI-saturated world cluster around exactly this.
② Critical thinking: The ability to evaluate information — including AI outputs — for accuracy, bias, and underlying assumptions, rather than accepting it at face value.
③ Creative synthesis: The ability to combine existing knowledge and information in novel ways, producing new ideas, expressions, or solutions that did not previously exist.
OECD's Education 2030 project refers to these collectively as "student agency" — the capacity to set goals, reflect on progress, and take responsibility for one's own learning — and identifies it as the essential disposition for navigating the complex, unpredictable world of 2030 and beyond. The student who waits to be told what to think is poorly equipped for a world where the supply of ready-made answers is infinite but the judgement of which answers matter remains scarce.
Inquiry as the New Mode of Learning
Japan's Ministry of Education updated its generative AI guidelines (Ver.2.0, December 2024) with an explicit emphasis on using AI to support foundational knowledge acquisition while increasing opportunities for students to ask "how can I use what AI has given me to deepen my inquiry?" This framing captures something important: the value is not in obtaining the AI's answer, but in what the learner does next with it.
Children who have experience with inquiry-based learning develop a tolerance for open-ended problems — the kind that have no single correct answer. They have a habit of analysing what went wrong and adjusting their approach. This is structurally identical to the experience of debugging a programme: you formulate a hypothesis about the error, test it, observe the result, and revise. Digital creation environments are, in this sense, natural training grounds for inquiry.
Knowledge Still Matters — Its Role Has Changed
One important clarification: the declining relative value of memorisation does not mean knowledge has become irrelevant. To think critically, you need a body of knowledge against which to evaluate new claims. To ask good questions, you need to know enough to recognise what you do not know. Knowledge is shifting in role from "thing to be stored" to "scaffolding for thought" — but the scaffold is still necessary.
What children need is not "stop learning facts" but rather "learn facts in the service of asking better questions." That is an experience that must be lived, not just described. It lives in the act of building something with a computer and encountering a problem you did not anticipate. It lives in using AI to research a topic and then asking whether the AI got it right. It lives in the moment of creating something — a programme, a presentation, a piece of digital art — and realising you need to understand more in order to make it better.
At Digital Kodomo BASE, we do not hand children answers. We give them a computer, access to AI and programming tools, and the freedom to try. The ability to ask the next question — that is what we are here to help grow.