What a Three-Year-Old Knows About Artificial Intelligence

What do we do when reality changes faster than the story we had written about it?

Today, AI is rapidly changing our world. But reality has always had a way of rewriting our stories.

Relationships change. Hearts break. People disappoint us. Health changes. Babies are born. People we love die. Careers end. New ones begin. Sometimes we learn something that changes our understanding of a person we thought we knew completely.

Life changes the story.

And perhaps the challenge has never really been preventing change. Perhaps it is learning how to see what is actually in front of us, rather than clinging to what we believed should be there.

For most transformative technologies, the future arrives almost imperceptibly.

A drip.

Another drip.

Something seems novel or optional. Then useful. Then increasingly normal. Eventually, the world reorganizes itself around it.

The printing press did this. The microscope did it. The telescope did it. Indoor plumbing did it. Electricity did it. Automobiles did it. Airplanes did it. The internet did it. Smartphones did it.

And I suspect AI will do it on a scale that is difficult for us to imagine precisely because we are standing inside the transition rather than looking backward at it.

When the World Changes Before We Are Ready

Smartphones changed our lives so thoroughly that most of us probably don’t remember exactly how we managed without them.

Mine seems to be getting a little too smart lately. 😊

Just yesterday, it decided to sting my heart a little by flooding me with memories from this week four years ago—a particularly happy and hopeful chapter of my life shared with someone I once cared for deeply.

I’ve gotten rather good at letting those memories stay where they belong: part of a life I once lived and can remember fondly without needing it back.

Then my phone serves me a photograph.

And for a moment, there it is:

the life I once thought I was going to have.

Technology didn’t create that feeling. It simply reminded me of something humans have always had to do:

update our understanding when reality changes.

Sometimes what changes is technological.

Sometimes it is medical.

Sometimes it is a relationship, a career, a dream, or our understanding of another person.

And sometimes new information arrives that changes the story entirely.

That can be profoundly disorienting because we aren’t always grieving the loss of a person, place, or possibility. Sometimes we are grieving the story we had constructed around it—the future we imagined, the meaning we assigned to it, or even our understanding of what had already happened.

And our brains don’t surrender stories easily.

People love saying that broken things become more beautiful.

Sometimes they do.

But sometimes broken glass is simply broken glass.

Sometimes new information changes our understanding of a person, institution, relationship, career, or dream so fundamentally that reconstructing our previous perception would require us to deny what we now know.

Perhaps wisdom isn’t always repairing.

Sometimes wisdom is being able to say:

I loved what I believed this was. I can honor that without pretending it is still true.

And then turning toward what actually exists.

AI and the Courage to Be a Beginner Again

AI fascinates me.

I first learned about ChatGPT in August 2024, and I was skeptical. But curiosity won.

I started asking questions.

Two years later, I use AI to research subjects, explore ideas, polish my writing, read a menu in France, challenge my assumptions, and sometimes even help me understand complicated human experiences.

I use several AI tools now, depending on what I am trying to learn, write, or research. I wouldn't want to imagine those tools disappearing.

For someone willing to remain curious, AI provides access to an extraordinary amount of knowledge.

But perhaps its most interesting contribution isn't simply access to information.

AI doesn't merely democratize knowledge. It can democratize the courage to learn.

Think about a three-year-old.

A three-year-old doesn't think:

I should probably already know this.

She asks:

Why?

You answer.

She asks:

Why?

You explain it differently.

She misunderstands something and isn't humiliated by being wrong. She simply incorporates the new information and keeps going.

Somewhere along the way, many of us learn that not knowing is embarrassing.

We stop raising our hands.

We become afraid of asking a stupid question.

We become reluctant to admit that we don't understand something everyone else seems to understand.

AI creates a private intellectual space where we can become beginners again.

We can say:

I don't understand.

Explain that differently.

Give me an example.

Why?

How do you know?

Could there be another explanation?

And no one rolls their eyes.

No one thinks we should already know.

No one rejects us.

We simply keep learning.

AI democratizes the time to learn, too.

You haven’t suddenly been given more hours in your life. You’ve been given a way to make the little pockets of intellectual curiosity within those hours extraordinarily productive.

Before Diplomas, There Were Scholars

That matters because scholars existed thousands of years before diplomas.

Ancient civilizations had scribes, philosophers, mathematicians, astronomers, physicians, priests, and teachers who devoted their lives to the pursuit of knowledge.

Ancient Greece had philosophical schools. China developed sophisticated scholarly traditions and eventually imperial examinations. India had major centers of learning. The Islamic Golden Age produced extraordinary scholarly institutions and libraries.

Many educated people learned through apprenticeship, observation, debate, mentorship, and relentless curiosity. Then they taught the next generation, which expanded what had come before.

Modern university credentials came much later.

Medieval European universities such as Bologna, Paris, and Oxford eventually formalized degrees. Over time, credentials became society's way of saying:

This institution certifies that this person has mastered a particular body of knowledge.

And that was enormously useful.

It still is.

But AI introduces a fascinating disruption because access to knowledge is becoming radically democratized.

That does not make expertise obsolete.

Quite the opposite.

I think genuine expertise may become more valuable.

A brilliant physician isn't valuable simply because they can remember thousands of medical facts. Their value lies in recognizing patterns, understanding uncertainty, knowing which facts matter, exercising judgment, communicating with another human being, and accepting responsibility for decisions.

What may lose some of its value is knowledge as status.

Knowing something simply because you had privileged access to the place where it was taught becomes less remarkable when an extraordinary tutor can sit beside almost anyone at 11:00 at night and essentially ask:

What are you curious about? Let's explore it.

That is extraordinary.

Medicine Has Been Here Before

Healthcare may be one of the most interesting places to watch this transformation collide with thousands of years of history.

There was no single moment when the "physician" neatly separated from the healer.

In the ancient world, healing was often learned by watching an experienced practitioner.

The Hippocratic tradition (5th–4th century BCE) helped move medicine toward observation and natural explanations of disease rather than exclusively supernatural ones.

There were no modern licensing boards, residency programs, or board examinations.

Quality depended heavily upon reputation, mentorship, and circumstance.

Over centuries, different cultures developed increasingly sophisticated medical systems. Ayurveda, traditional Chinese medicine, and the medieval Islamic medical world developed complex traditions. European universities eventually formalized medical education.

Then something enormously important happened.

Medicine became standardized.

Medical schools, licensing exams, residency, board certification, clinical trials, professional standards, and evidence-based medicine collectively reduced the role of luck.

The bar was raised.

And we absolutely do not want to lose that.

Modern medicine became astonishingly good at defeating discrete problems.

Antibiotics.

Manufactured insulin.

Joint replacement.

Organ transplantation.

Cardiac imaging.

Neonatal intensive care.

Tumor sequencing.

Robotic surgery.

Medicine naturally developed a remarkably effective workflow:

symptom → diagnosis → intervention

For many problems, that model is spectacular.

But many of today's most consequential health conditions don't fit neatly into it.

Cardiovascular disease, metabolic disease, frailty, cognitive decline, and many cancers emerge from complicated interactions among genetics, environment, nutrition, sleep, movement, stress, exposures, medications, and behavior.

Lifestyle does not replace indicated medical treatment.

Food isn't a substitute for chemotherapy when chemotherapy is indicated.

Exercise isn't a substitute for insulin when insulin is needed.

Meditation doesn't replace antibiotics in bacterial sepsis.

But food, movement, sleep, prevention, and human connection can still be extraordinarily powerful components of medicine.

Those ideas do not have to compete.

From Keeper of Knowledge to Expert Navigator

This is where I think the future becomes fascinating.

The traditional model looked something like this:

Patient → Physician → Medical Knowledge

The physician was the gatekeeper because the patient had limited access to medical information.

Then came the internet:

Patient → Google → enormous undifferentiated pile of information

We gained access without necessarily gaining judgment.

Thus, the era of:

"Don't believe everything you read on Dr. Google." 😄

AI potentially creates something different:

Patient + AI → questions and understanding → Physician + clinical expertise → shared decision making

Imagine the patient who walks into an appointment understanding what LDL is, having read about several treatment options, knowing her family history, carrying years of wearable data, and arriving with thoughtful questions.

That doesn't diminish the physician.

The physician brings things AI and internet research cannot replace: examination, clinical context, experience, diagnostic reasoning, access to testing, understanding of probabilities, recognition of unusual presentations, and—critically—responsibility for medical decisions.

Neither needs to diminish the other.

They can make one another better.

Perhaps the physician gradually moves from being primarily the keeper of knowledge to becoming the expert navigator of knowledge.

When information itself is abundant, the scarce and valuable skill is no longer merely possessing facts.

It is contextualizing them.

Correcting misunderstandings.

Recognizing what matters.

Understanding what doesn't.

And helping one particular human being decide what to do next.

Because a Patient Is Not a Population

Evidence-based medicine necessarily studies populations.

Essentially:

In a population of people who resemble you, intervention X produced outcome Y this often.

That is enormously valuable.

But the person sitting in the exam room isn't a population.

She's one person.

She is nuanced.

She has her own genome, medications, lab values, imaging studies, microbiome, medical history, sleep patterns, exercise habits, preferences, risk tolerance, family circumstances, financial situation, and goals.

Increasingly, she also has continuous streams of information from glucose monitors, smartwatches, wearables, and other sensors.

Medicine therefore begins moving from asking only:

What generally works?

toward:

What generally works—and what appears to work for you?

That is a very different relationship.

But patient empowerment should never become patient certainty.

AI can help us formulate better questions. It should not convince us that ten minutes of research makes us equivalent to someone who spent a decade learning medicine.

The intellectually healthy patient should be able to say:

I found something interesting. Help me understand whether this applies to me.

And the intellectually healthy physician should be able to say:

That's an interesting question. Let's look at it.

Neither person needs to feel threatened.

Curiosity isn't a challenge to authority.

And expertise doesn't require unquestioning obedience.

Perhaps the Future Is Strangely Circular

Thousands of years ago, healing often involved observation, food, plants, movement, environment, and the healer's accumulated wisdom.

Medicine then became increasingly scientific, specialized, standardized, and technological—and produced astonishing advances.

Perhaps the next era doesn't reject either world.

Perhaps it combines them.

We keep antibiotics.

Randomized trials.

Robotic surgery.

Insulin.

CT scanners.

Molecular diagnostics.

Highly trained physicians.

And we rediscover walking.

Strength.

Sleep.

Food.

Sunlight and nature.

Human connection.

Prevention.

Knowing the person rather than merely knowing the diagnosis.

Except now we can study those things scientifically rather than assuming something is beneficial simply because it is "natural."

And sitting beside both patient and physician may be an intelligence capable of helping sift through more information than either human could possibly read in a lifetime.

That isn't really going backward.

It is carrying something valuable from the past into something entirely new.

The People Who Keep Asking Why

Perhaps one of the most important skills for living in an unimaginably technologically advanced world will be something we possessed naturally when we were three years old.

The willingness to say, without embarrassment:

I don't understand.

Why?

Explain that again.

What if?

How do you know?

Could there be another answer?

And then—

Why?

Again.

And again.

Until we understand.

Maybe AI's greatest gift won't ultimately be that it gives humanity more answers.

Maybe it will be that it gives more of us the courage to keep asking questions.

So perhaps the interesting question isn't:

Do I like the fact that AI is changing the world?

It is:

Who do I want to become in a world where AI exists?

That question feels much more empowering.

Because I don't think the greatest divide will necessarily be between people who are technologically sophisticated and people who aren't.

I suspect one of the most consequential divides may be between people who remain curious and people who stop being curious.

A curious 70-year-old with AI may have an astonishing intellectual life.

She can study astrophysics because she wondered about a star on her evening walk.

She can learn Italian before Florence.

She can understand a medical paper rather than simply accepting its conclusion.

She can explore philosophy, history, nutrition, architecture, music, economics, or genetics at whatever depth she chooses.

She can ask the embarrassing question she would never raise her hand to ask in a classroom.

She can say:

I still don't understand. Explain it differently.

And she can keep going until she does.

For most of human history, that kind of individualized access to knowledge belonged disproportionately to people with money, proximity, education or privilege.

We may be approaching a world where curiosity itself becomes one of the primary barriers.

I find that hopeful.

Not because AI will make humans unnecessary.

Perhaps the most interesting possibility is almost the opposite.

When information becomes abundant, maybe the things that are distinctly human become even more valuable:

Judgment. Wisdom. Discernment. Imagination. Courage. Empathy. Integrity. Curiosity. Love.

And perhaps most importantly:

the willingness to change our minds.

Learning to Swim

The faucet is turning.

I can hear the water now.

I don't know exactly what the world will look like five years from today. I suspect anything I can imagine is only a tiny fraction of what is coming.

But I don't want to spend these years trying to stop the water.

I want to learn to swim.

I want to remain curious enough to question what I know, brave enough to release what is no longer true, humble enough to learn from people who know more than I do, and confident enough to remember that no degree, title, rejection, relationship or institution gets to determine the value of my one extraordinary life.

We get one.

And perhaps the people who thrive in whatever comes next won't necessarily be the people who knew the most when the faucet began to turn.

Perhaps they'll simply be the ones who never stopped learning how to swim.

And perhaps that has always been humanity's most important form of intelligence:

Learning to let go of the world—or person, identity, credential, dream, or relationship—we thought should exist, seeing clearly what actually exists, and having enough curiosity and courage to move forward anyway.

Stay curious. Keep asking why. And never become too grown-up to change your mind—

🍯 Honey 🐝

The future doesn’t require us to know everything. It requires us to remain willing to learn.