Are Reading, Writing, and the Abacus Outdated? — The "3 + 2" Foundational Skills for the AI Era

Reading textbooks aloud, kanji drills, the after-school abacus class — these were everyday scenes on the streets of Showa-era Japan. Yomi-kaki-soroban — reading, writing, and the abacus — has long stood as the symbol of Japan's foundational education, and for many of today's parents it is woven into their own elementary school memories. Today, however, it is no longer unusual for a child to look up from their homework and say, "I asked ChatGPT."

In December 2024, Japan's Ministry of Education published its Guidelines on the Use of Generative AI in Elementary and Secondary Education (Ver. 2.0), and AI is beginning to enter the classroom as well. In an era when AI writes, calculates, and even handles research — has yomi-kaki-soroban become obsolete?

No — reading, writing, and the abacus do not become obsolete. Rather, they remain as the foundation that lets us understand AI's answers, question them, and connect them to our own thinking. What changes is what gets layered on top. The foundational skills of the AI era can be organized as five: three that remain as the base, and two that are newly added.

① "Reading" — A foundation that gains value in the AI era

According to mathematician Noriko Arai's Reading Skill Test research, many Japanese middle school students cannot accurately read their own textbooks. AI, meanwhile, has reached the level of solving University of Tokyo entrance-exam mathematics and English. Now that AI generates vast amounts of text in an instant, the value of being able to understand that information, judge its truthfulness, and rebuild it in your own words actually rises. Even to give AI clear instructions in the first place, you need to read accurately and put your own intent into words — that is the foundation.

② "Writing" — Even when AI helps, hold on to your own words

When AI can write a composition in seconds, writing may seem unnecessary. But if a child has AI write their book report, they can hand it in — and yet, when later asked, "Which part of the book moved you?" they cannot answer in their own words. Writing is the work of taking vague feelings and thoughts in your head, putting them into words, and giving them structure. Editing AI's output — judging "this part is wrong" or "I want to add this here" — is an extension of the same writing skill.

③ "Soroban" — Calculation moves to AI; number sense stays with us

The role of arithmetic shifts. Four-function arithmetic and complex calculations are now done by AI in an instant. What rises in importance instead is the ability to estimate "roughly this much" before getting the answer, and to notice when units or ratios are off. The more AI takes over calculation, the more humans are asked to read numbers.

④ "Asking" — Finding what is even worth asking

Once part of calculation and research is handed to AI, what do we do with the time and mental space that opens up? On top of reading, writing, and the abacus, two new skills are layered in.

The first is the ability to ask. AI answers the questions you give it, and it can even propose candidate questions. But choosing "what question really matters to me" requires your own experience and values. Vague questions return vague answers. The ability to make a question more specific shapes the answer AI can give you.

This is the skill known as "prompting" — how you pose a question to AI. Asking "Tell me a dinner recipe" returns generic results, but "Three Japanese dishes I can make in 20 minutes with the pork and cabbage in my fridge" returns something far more useful. The precision of the question directly determines the quality of the answer.

⑤ "Creating" — Using AI as material to shape your own ideas

The second is the ability to create. AI speeds up the work, but only you can decide "what I would make." Using AI as a draft to refine your own thinking, and working with others to give it form — that is what "creating" means here.

For children, "creating" does not have to mean a great invention. For a video introducing your neighborhood, you can have AI suggest a structure, then choose the scenes yourself, film them, and edit. When writing a story with AI, you can pick only the plot suggestions you like and rewrite the rest. The experience of treating AI's proposals as raw material to accept or reject — rather than handing the whole finished product to AI — is what grows the ability to create.

Encountering "asking" and "creating" from childhood

A January 2024 report from the International Monetary Fund (IMF) estimated that around 40% of jobs worldwide — and around 60% in advanced economies — will be affected by generative AI. The World Economic Forum's Future of Jobs Report 2025 likewise lists "analytical thinking" and "creative thinking" as the skills with rising demand. By the time today's children enter the workforce, AI will be a given in both work and daily life.

The way today's parents learned at school and at work may not be enough on its own to grow the abilities that the AI era will require. Schoolwork remains important, but skills like "forming a question," "questioning AI's answer," and "giving something form yourself" do not grow through memorizing knowledge alone. This is not a question of talent but of experience. A child who has been used, from a young age, to asking "why do I think so" and "is there another way" — trying things, then reworking them — learns to use AI not as a vending machine for answers, but as a partner that expands their own thinking.

Digital Kodomo BASE does not simply teach how to operate a computer or AI. We create a place to think alongside children about what we use these tools for, and what we want to ask and create with them. Rather than stopping at having AI generate an answer, we go on to ask "why does it come out this way," "is there another method," and "what would I make." We weave that experience of asking and creating into our PC, programming, and AI activities.