I think my kids are pretty intelligent.

They're both A and B students. They're good at different things. They generally make good decisions.

Neither one spends Saturday night reading the encyclopedia for fun, but I suppose I'll survive that disappointment.

But lately I've been wondering what it even means to be intelligent.

To me, intelligence isn't simply being really good at one subject or being able to memorize a lot of information.

It's broader than that.

It's the ability to think clearly and rationally.

To connect ideas.

To recognize when something doesn't make sense.

To see nonsense for what it is.

And increasingly, I wonder what happens to those abilities when our kids grow up in a world where they don't really need to know very much at all.

Because almost anything they want to know is now seconds away.

And with AI, they don't even have to look for it anymore.

They can just ask.

I Used to Know Everyone's Phone Number

When I was younger, I knew people's phone numbers.

My parents.

My friends.

My girlfriend.

Probably the local pizza place.

Then I got a cell phone.

Suddenly, I didn't need to remember phone numbers anymore.

So I stopped.

Today, if something happened to my phone, I'm not entirely sure how many people in my life I could actually call.

Then GPS came along.

Before GPS, if I needed to get somewhere, I'd look at a map or directions. I'd figure out which roads connected. I'd pay attention to where I was going.

Eventually, I'd know how to get there.

Now I put an address into my phone and blindly obey a pleasant voice telling me where to turn.

And I've noticed something.

I can travel somewhere multiple times and still not really know how to get there.

Psychologists have a term for the broader phenomenon: cognitive offloading. We use something outside our minds—a phone, calendar, GPS, calculator, or even a handwritten note—to reduce the amount of mental work we have to do ourselves. That's not inherently bad. These tools can make us dramatically more capable. But when technology performs a cognitive task for us, we may have less reason to practice that task ourselves. 

Which makes me wonder:

What happens when we start doing that with knowledge itself?

AI Is an Amazing Tool

I should make something clear.

I'm not anti-AI.

Quite the opposite.

I use it. A lot.

And I think it's an incredible tool for my kids to have.

If they're trying to build something, solve a complicated problem, learn how something works, or figure out why something isn't working, having access to AI is amazing.

I want them to learn how to use it.

But I get much more uncomfortable when AI becomes the place they go for every basic answer.

Especially with schoolwork.

Because there's a difference between using AI to help you understand something and using AI so you don't have to understand it.

Those can look remarkably similar from the outside.

Both might produce the correct answer.

Only one necessarily produces learning.

I Started Looking at My Own Boys Differently

I have two teenage boys, and what I noticed wasn't identical in each of them.

My older son has a good group of friends and spends plenty of time with them.

My concerns with him were more about ideas.

Social media puts an incredible amount of information in front of teenagers every day. Opinions arrive fully formed, usually packaged into short videos and delivered by algorithms that get remarkably good at figuring out what keeps someone's attention.

I started worrying about whether he was really evaluating those ideas.

Was he asking whether something was true?

Was he considering another perspective?

Or was he simply absorbing whatever kept appearing in front of him?

My younger son concerned me in a different way.

He seemed more naturally inclined to think through ideas, but I watched him and his friends become increasingly content to socialize almost entirely through computers.

Getting a group together to actually go somewhere or do something could be surprisingly difficult.

Why go outside when everyone can sit at home and play together online?

And I found myself wondering what happens when that becomes the default version of friendship.

Not because playing video games with your friends is inherently bad.

I played video games too.

But because childhood only gives us so many opportunities to practice being human around other humans.

“Who Cares? It's Fixed.”

I learned a version of this lesson years ago while working in software.

When you're writing code, details matter.

Every step in your logic does something.

If you don't understand what you're building, you make assumptions. If enough of those assumptions are wrong, eventually the software lets you know.

Usually at a remarkably inconvenient time.

So you learn to ask questions.

A lot of questions.

What exactly are we trying to accomplish?

What happens if the user does this?

Where does this data come from?

What happens next?

What are we assuming here?

Over time, that way of thinking escaped from my computer and started infecting the rest of my life.

Things needed to make sense.

And if they didn't, I wanted to know why.

I remember times when I'd be trying to fix a software bug. I'd change a few things, test it again...

...and suddenly it worked.

Someone might say:

“Who cares? It's fixed.”

My response was always:

“If I don't know why it's fixed, it's not fixed.”

Because maybe I didn't actually solve the problem.

Maybe I changed something unrelated.

Maybe the conditions that caused the bug temporarily disappeared.

Maybe it was going to break again tomorrow.

A working result wasn't enough.

I needed to understand why it worked.

That's increasingly how I think about kids and AI.

Getting the answer isn't the same as understanding the answer.

And I'm worried we're entering a world that makes it incredibly easy to confuse the two.

Knowledge Still Matters

There's an argument you sometimes hear about modern technology that goes something like this:

Why memorize something you can just look up?

There's some truth to that.

I don't particularly care whether my kids memorize information simply for the sake of proving they can memorize it.

But there's a problem with taking that idea too far.

You need knowledge to think.

What you already know gives you context for understanding new information.

Research on reading comprehension has demonstrated this repeatedly. Background knowledge helps readers make sense of new material, connect ideas, and construct meaning from what they're reading. A review of decades of research found that background knowledge plays an important role in comprehension, although exactly how much it helps depends on the reader and the material.

That's intuitive if you think about it.

Suppose someone tells me Jacksonville is about 100 miles south of Miami.

I don't have to Google that.

I know enough about Florida to immediately think:

Wait. That doesn't sound right.

But imagine I know absolutely nothing about Florida.

Now I'm entirely dependent on whoever—or whatever—is giving me the answer.

I don't have enough knowledge to recognize the mistake.

That's why I don't think AI makes knowledge less important.


​In some ways, I think it makes knowledge more important.

You Need to Know Enough to Know When Something Is Wrong

AI can produce incredibly convincing answers.

That's part of what makes it useful.

It's also part of what makes critical thinking around it so important.

A well-written answer feels authoritative.

But confidence and correctness aren't the same thing.

Software taught me that years ago.

You can sit in a meeting with someone who sounds extremely confident about how something works.

Start asking questions.

Then another.

Then another.

Eventually, you discover one of two things.

They really understand the subject.

Or they understood it right up until question number three.

I don't mean that cruelly. I've been the person who discovered at question number three that I didn't understand something nearly as well as I thought I did.

That's the value of questions.

They expose the difference between familiarity and understanding.

Our kids need that ability when dealing with people.

They need it when watching social media.

And they're absolutely going to need it when working with AI.

Something I've Noticed With My Boys

Every now and then I'll be talking with one of my boys and use a word that seems completely ordinary to me.

They'll stop me.

“I don't know what that means.”

And my immediate internal reaction is usually:

Really? You don't know that word?

Then I have to remind myself that I'm a lifelong reader.

I've had decades more exposure to books, work, conversations, and unfamiliar words than they have.

Maybe I'm simply expecting too much.

But it has made me curious.

Kids today consume information very differently than I did.

How much does that change the vocabulary and background knowledge they accumulate along the way?

I don't know the answer.

And I'm deliberately not going to turn one dad's observations about his two kids into a sweeping conclusion about an entire generation.

But we do know that reading exposure is associated with vocabulary, comprehension, spelling, and broader language development across childhood and into adolescence. A large meta-analysis found those relationships become stronger through the school years.

So I do wonder what happens as the ways kids consume information continue to change.

Not because books are magical and screens are evil.

But because how we encounter information influences what we have the opportunity to learn along the way.

Intelligence and Rational Thinking Aren't Exactly the Same Thing

There's another distinction I've found fascinating as I've learned more about critical thinking.

Psychologist Keith Stanovich has spent decades studying the relationship between intelligence and rationality.

One of the ideas in his work is that the cognitive abilities measured by traditional intelligence tests don't capture everything involved in rational thinking. Smart people can still reason poorly, fall victim to biases, or make irrational judgments. Stanovich even coined the term “dysrationalia” to describe the difficulty of thinking and behaving rationally despite having adequate intelligence.

That distinction matters to me as a parent.

I don't simply want my boys to be smart.

I want them to recognize when they don't know something.

I want them to question assumptions.

I want them to notice contradictions.

I want them to ask for evidence.

I want them to change their minds when better information comes along.

I want them to be skeptical without becoming cynical.

And I want them to understand something deeply enough that they can explain why they think what they think.

Those aren't things you develop simply by having access to more information.

They require practice.

The Smartest Kid in the Room May Not Know the Most Answers

For most of my life, knowing something gave you an advantage.

If you knew how to solve the problem and I didn't, you had something I didn't have.

That's changing.

Our kids are entering a world where nearly everyone will be able to pull an extraordinarily powerful AI out of their pocket and ask:

What's the answer?

Explain this.

Write this.

Solve this.

Summarize this.

Tell me what to think about this.

That's incredible.

But if everyone has access to the answers, the advantage begins moving somewhere else.

Can you ask the right question?

Can you recognize a weak answer?

Can you spot the assumption hiding underneath it?

Can you distinguish evidence from opinion?

AI Is a Terrible Reason to Stop Learning Things

This is where I think parents and educators have to be careful.

If AI knows everything, it's tempting to conclude that our kids don't need to.

I think that's backwards.

Our kids need knowledge because AI knows so much.

They need enough information stored in their own heads to evaluate the information coming from everywhere else.

They need vocabulary so they can understand complicated ideas.

They need history so they can recognize when someone makes a questionable historical claim.

They need science so they can evaluate scientific claims.

They need math so they can recognize when numbers don't add up.

They need experience so they can recognize when something that sounds good in theory doesn't make sense in reality.

And they need critical-thinking skills to connect all of it.

The goal isn't to compete with AI at remembering facts.

We're going to lose that competition.

The goal is to develop minds capable of doing something useful with the information AI provides.

Grades Still Matter to Me

I'm also not someone who believes grades don't matter anymore because Google and AI exist.

Grades matter in our house.

My wife and I both went to college, and we'd like our boys to have that opportunity too.

Good grades help make that possible.

But there's another reason I care about them.

In my experience with my own boys, when a grade starts dropping, it's often not because they've suddenly become incapable of understanding the material.

It's effort.

An assignment wasn't taken seriously.

Studying didn't happen.

Something got rushed.

And that bothers me.

Because intelligence without effort isn't particularly useful either.

There are going to be plenty of things in life my kids don't enjoy doing.

I certainly have them.

Being able to apply yourself when something isn't fascinating is part of becoming capable.

AI doesn't change that.

If anything, having an effortless shortcut available makes developing that discipline more important.