Thinking Beats
Coding

Wisdom is not knowing more. It is knowing which question is worth asking before anything else begins.

AI can execute any task you define clearly. It cannot decide which task is worth defining. The child trained to frame problems precisely, asking the right question before reaching for tools, has a durable advantage that no amount of execution skill can replicate. That is the Stoic Wisdom angle.

My youngest daughter is seven. Last month she was watching me use an AI tool to draft something, and she leaned in and asked a question that stopped me.

Not “how does it work?” Not “can I try?” She asked: “But how do you know if it's giving you the right answer?”

Seven years old. I did not have a clean answer for her. Which is its own kind of answer.

Every parent I know is thinking about coding. Whether their child should learn it. Which app teaches it best. Whether to sign them up for a class.

I understand the instinct. Coding felt like the safe bet for twenty years. Learn to speak the language of machines and you will always have work.

That bet is off. Not because coding is useless. Because AI can now write most of the code most engineers need on most days. The syntax, the boilerplate, the standard patterns: handled. What AI cannot do is decide what to build. It cannot identify which problem is worth solving. It cannot ask the question that unlocks the whole project.

That is a human job. It has always been a human job. We just forgot it was, because for a long time the execution was hard enough to seem like the thing. The Stoics had a name for that: asking the right question at the right moment. They called it Wisdom.

Socrates never claimed to know anything. This was not false modesty. It was the method.

His entire practice was built on the observation that most people, especially confident and successful ones, were operating on assumptions they had never examined. The Socratic method did not deliver answers. It dissolved bad questions and replaced them with better ones.

AI is the greatest answer machine ever built. It has absorbed more information than any human could read in a thousand lifetimes. It cannot tell you which question is worth asking.^[2] That asymmetry is the entire opportunity.

A child who learns to ask precise, original questions has a lasting edge over a child trained only to execute.

Teaching a child to code without teaching them to think is like teaching them to type without teaching them to write. The keystrokes are not the thing. The judgment behind the keystrokes is the thing.

Research on expert performance consistently shows what separates top performers from the rest is not technical execution. Problem framing is what separates them. Experts spend more time analyzing a problem before solving it, and that investment consistently produces better outcomes.^[3] In the AI era, 69% of developers don't even plan to use AI for project planning,^[2] precisely because problem framing requires judgment AI cannot reliably provide.

Execution is being compressed toward zero cost. Problem framing remains expensive, rare, and entirely human. This is not a technical skill. It is a judgment skill. Which is exactly what the Stoics meant by Wisdom.

Stop rewarding answers. Start rewarding questions.

This is harder than it sounds. Our schools, our dinner tables, our family culture: all of it oriented around answers. We celebrate when children get it right. We feel relieved when they know things.

My youngest asked it better than most adults would: “But how do you know if it is giving you the right answer?” She was not asking how AI works. She was asking how to evaluate it. That is the discernment layer. Research backs this up: AI systems that ask questions instead of just giving answers produce better thinking in the humans using them.^[4] Questioning AI output is a skill. It can be taught.

I did not teach her that habit. But I can reinforce it. And so can you.

1. [1]

   Marcus Aurelius, Meditations, c. 161-180 AD. Public domain. classics.mit.edu

2. [2]

   Stack Overflow, "2025 Developer Survey — AI." 69% of developers don't plan to use AI for project planning; experienced developers most cautious at only 2.6% "highly trusting" AI output. survey.stackoverflow.co

3. [3]

   National Academies of Sciences, Engineering, and Medicine, "How People Learn: Brain, Mind, Experience, and School," National Academies Press, 2000. Chapter 2: How Experts Differ from Novices. nationalacademies.org

4. [4]

   Pataranutaporn, P., et al., "Don't Just Tell Me, Ask Me: AI Systems that Intelligently Frame Explanations as Questions Improve Human Logical Discernment Accuracy," ACM CHI Conference, 2023. media.mit.edu

The Better Question Game

Once a week at dinner, run one round. Someone names a topic, any topic. Then everyone has 60 seconds to write down the best question they can think of about it. Not an answer. A question.

A better question opens more doors than it closes. It reveals an assumption. It points to something nobody had thought to look at.

The Assumption Hunt

Take any task your child is doing: a school project, a decision, a plan. Ask them to write down three assumptions baked into how they are approaching it. Not facts. Assumptions. Things they are taking for granted that might not be true.

AI will execute whatever assumptions you give it without questioning them. The child who surfaces assumptions before execution will catch errors no AI will ever flag.