The Five Words I Taught My 7-Year-Old That Will Compound for 70 Years

My 7-year-old was about to play soccer on the footpath. Cars on one side. Driveways on the other. A friend hyped up. Excitement off the charts.

I didn’t say “be careful.”

I didn’t say “no.”

I didn’t say anything.

I just watched him run the algorithm I’d installed in his head.

He paused. Scanned. Looked at the road. Looked at his friend. Looked at me. Then said:

Six words. No fight. No lecture. No drama.

That moment — that one tiny refusal — was three years of work paying off in real time.

Because two years ago, I stopped using the phrase every parent on earth defaults to. And I replaced it with something stranger, deeper, and 10x more powerful.

I started teaching Evaan how to think in algorithms.

“Be Careful” Is Broken Code

A 7-year-old’s brain is not stupid. It is just under-trained.

Telling a kid to “be careful” is like telling a junior engineer to “be senior.” There are no instructions in there. No states. No checks. No fallback path.

What kids actually need is an executable.

Something they can run when you’re not standing next to them. Something that survives the moment of excitement, the dopamine of the kick, the friend yelling “PASS IT!” right when a car turns the corner.

Vague rules fall apart at the exact moment they’re needed.

Algorithms don’t.

Notice → Decide → Act → Review → Refine

This is the loop I’m installing in my son’s head.

Five steps. Five verbs. Tiny enough to remember in real time. Strong enough to compound for a decade.

A micro-algorithm.

I am not raising a child. I am installing an operating system.

Every situation is just an input. The OS handles the rest.

Ray Dalio runs his $150B hedge fund on a nearly identical 5-step loop: identify the goal, identify the problems, diagnose, design, do. He didn’t invent it. He just labelled it. The pattern is older than money. It’s how every reliable system on earth — from immune systems to flight controllers to Toyota factories — actually works.

Notice. Decide. Act. Review. Refine.

The same loop runs your body’s heartbeat. The same loop runs Boeing’s quality control. The same loop runs the most successful hedge fund in history.

I just want it running my 7-year-old.

If I get this OS booted now, he runs it for 70 more years. That’s not parenting. That’s leverage.

(There’s a delicious irony here: this is the same compounding logic James Clear used to popularise atomic habits — 1% better every day stacks to 37x in a year. Clear was talking about habits. I’m talking about decision quality. They’re the same thing. A habit is just a decision algorithm running on autopilot.)

The Footpath Soccer Algorithm

Back to the soccer ball.

Most parents would have said “be careful” or “no, not here.” Both end in a fight. Both fail to install anything that lasts.

I gave him the algorithm instead:

1. Look left and right.
2. Find the road, the driveways, the bikes, the dogs.
3. Pick the safest direction to kick.
4. Kick with control, not just power.
5. After the kick, scan again.
6. If the situation is not safe enough — stop. Don’t play here. Wait for a real field.

That last branch is the one that did the work.

Because when he ran the algorithm — actually ran it, not theoretically — the output came back unsafe. Too many cars. Too many driveways. Too much movement.

So he refined. He chose the field.

That’s not safety. That’s judgement.

And judgement is the only thing in life that scales.

First, Second and Third-Order Consequences

A 7-year-old naturally sees one move ahead.

“I kick the ball hard. It goes far. Fun.”

That’s first-order thinking. It’s also how most adults run their lives.

The world doesn’t stop at first-order.

So we run the chain together:

• First-order: ball goes onto the road.
• Second-order: instinct kicks in, you sprint after it.
• Third-order: car brakes hard, somebody gets hurt, the game stops forever.

Farnam Street calls this “second-order thinking” — the discipline of always asking and then what? instead of stopping at the first outcome.

Charlie Munger went further: the people who only think first-order are the ones who lose money, get into accidents, and ruin marriages. Not because they’re stupid. Because they stop thinking too early.

I just call it the “and then what” game. Evaan and I play it constantly.

• “I’ll skip homework tonight.” And then what?
• “I’ll eat all the candy at once.” And then what?
• “I’ll wake up late and rush.” And then what? (Spoiler: we both know how that one ended.)

A child who runs this loop on themselves before the action is unstoppable.

A child who only runs it after is just learning the hard way.

Safety Is Not the Opposite of Greatness

This is the part I will die on a hill for.

Kids quietly believe safety is what stops them from being great.

It’s the opposite.

You cannot become great at anything if you are constantly injured, sidelined, or unavailable. You cannot get the reps if you keep blowing up the rocket on the launchpad. The unsafe player looks brave for one moment. The intelligent player keeps improving for fifteen years.

You can’t become the GOAT if you’re constantly broken. Stay in the game long enough to get great.

Messi, Kohli, Ronaldo, Federer — every all-time great is essentially a person who managed to stay in the game longer than everyone else.

Algorithms keep you in the game.

The Decision Tree (and the One Branch That Saves Lives)

I drew this with him on a piece of paper. Took 4 minutes. He laminated it inside his head.

That second-to-last branch is the one that saves lives.

Because the danger was never the kick. The danger is the moment a kid quietly decides “close enough is fine.” Most accidents don’t happen because kids don’t know the rules. They happen because kids override the rules in their head.

So the algorithm has to override the override.

Eight words. Burned in. That’s the path Evaan ran on the footpath that day.

That’s how you know the OS is loading.

The Moment Evaan Wrote His Own Algorithm

This is when I knew it had taken root.

A few weeks later, mid-game on the actual field, he’s standing with the ball at his feet and he says — completely unprompted — “Dad, when I have the ball, I do this:”

1. Look up.
2. Check the defender.
3. Check teammate.
4. Check space.
5. Decide: dribble, pass, or shoot.
6. Then move into space again.

I almost dropped my coffee.

He didn’t memorise that. He generated it.

He took the safety algorithm and rewrote it for offense. Same framework. New problem. Different scale.

That’s the whole game.

The goal isn’t obedience. The goal is transfer.

A rule only works when you’re watching. A model works when you’re not.

Lev Vygotsky, the Russian psychologist, called this the Zone of Proximal Development — the sweet spot just above what a child can do alone but within reach with a small scaffold. Frameworks are scaffolds. Once the scaffold becomes internal, the child climbs into territory the scaffold never even pointed to.

That’s exactly what happened. I gave him a safety scaffold. He built his own offense.

Good Decision ≠ Good Outcome

This might be the most underrated lesson of all of this.

Annie Duke spent twenty years as a poker player and now writes about decision quality. Her core idea: a good decision can still produce a bad result, because of luck, randomness, or missing information. And a bad decision can occasionally win.

So the question to ask after anything goes wrong is not “did it work?”

The question is: “Was it a good decision with the information you had?”

This protects kids from two awful default modes — over-celebrating reckless wins and over-punishing thoughtful failures.

Evaan once smashed a piece of home decor doing a “perfect” indoor power kick. (We’ve now lost about 6–7 home decor items in two years. Six-Seven 😉.) The kick technique was actually flawless. The bug was the decision to do it inside.

We didn’t yell. We diffed the algorithm.

What did you notice? What did you decide? What happened next? What would you change?

Process over outcome. Every time.

System 1, System 2, and How Algorithms Become Instinct

This is the part that should make every parent reading this stop and re-read.

Daniel Kahneman won a Nobel Prize for showing that humans run on two systems.

System 1 is fast, automatic, instinctive. It’s what makes you swerve before you consciously see the obstacle. It’s where your habits, biases, reflexes, and “gut” live.

System 2 is slow, deliberate, effortful. It’s what runs when you’re solving a problem, weighing options, doing maths in your head, or resisting an impulse.

Most life-changing decisions are made by System 1. We just think they’re made by System 2.

Here’s the move that nobody talks about:

Anything you train through System 2 long enough eventually moves to System 1.

That is the entire game.

When I teach Evaan a micro-algorithm, I’m using his System 2 — slow, conscious, effortful. He has to think through every step. It feels heavy. He hates it the first three times.

But by rep fifty, the algorithm has migrated. It’s no longer something he thinks. It’s something he is.

That’s why he could refuse the footpath without struggling. He wasn’t fighting an impulse. He was running a different impulse.

The dirty secret of high-performing adults is that their “instincts” are mostly just frameworks they ran consciously for ten thousand reps until those frameworks went invisible.

You don’t raise a wise child by lecturing them.

You raise a wise child by giving their System 2 the right scripts and letting time bake those scripts into System 1.

That is the most important sentence I’ve written in three years on this blog.

This Is How I Build Software, Too

The wild part is — same loop runs everything I touch.

EVAAN — the multi-LLM workspace I’m building — has a routing layer called The Conductor. It runs Notice → Decide → Act → Review → Refine on every prompt. Notice the request. Decide which model fits. Act by routing. Review the response. Refine the routing for next time.

Vitality — my personal wellness dashboard, built on top of my Apple Health data — runs the same loop on my own body. Notice the sleep, the recovery, the strain. Decide what the day demands. Act on it. Review the data. Refine tomorrow’s choices.

PANTRI — the household OS I’m building for busy families — runs it on logistics. Notice the pantry. Decide the meal. Act on the shop. Review the spend. Refine the cycle.

Same five verbs. Different scale.

A kid is a tiny AI. A parent is the alignment team. A health stack is a feedback loop. A product is a refinement engine.

You don’t program facts into any of them.

You program loops.

Models Beat Rules. Always.

Children don’t need more lectures. They need better models.

A rule says: don’t run on the road.

A model says: here’s how to think about the road, and now you can handle every road for the rest of your life.

Carol Dweck’s whole growth-mindset thesis is built on this — kids who believe their brain can grow take feedback as data, not damage. Kids who believe ability is fixed take feedback as identity.

A micro-algorithm is the most natural growth mindset machine ever invented.

It says: “your process can improve.”

It doesn’t say: “you are good,” or “you are bad.”

It says: “let’s debug.”

That language alone changes everything.

How to Install This OS in Your Own Kid (in 5 Steps)

You don’t need a PhD. You need 10 minutes and a willingness to slow the moment down.

Do this for two weeks on one situation.

Then watch them apply the same loop to a second situation on their own.

That’s the moment. That’s transfer. That’s the OS booting.

The Compounding Effect

One safe kick doesn’t change a life.

But a thousand reps of Notice → Decide → Act → Review → Refine — that does.

Over time, the kid becomes:

• Less impulsive
• More strategic
• More aware of consequences
• More creative inside the rules
• More trustworthy with freedom

And the parent’s job changes too.

Less policeman. More coach.

Less “stop doing that.” More “what’s the algorithm?”

That is the real compounding curve. Tiny frameworks, repeated often, become instinct. Instinct becomes character. Character becomes a life.

Greatness is not one big decision.

It’s a thousand tiny algorithms, refined over years.

Teach your kid to think in loops. Then watch them build their own.

MKGA. Make Kids Great Again.

— Dhawal