Don’t Ask AI to Think for You

Artificial intelligence can make us stronger. But only on one condition: we must stop using it as a substitute for our own thinking and learn to talk to it in a way that helps us think better ourselves.

One day I caught myself having a thought that at first seemed almost insignificant.

I needed to write a simple business email. Nothing special. I had written emails like that dozens, maybe hundreds of times before. A few paragraphs. A calm tone. Precise wording.

Normally, the process was familiar. I would open a document, draft a first version, read it over, move a few words around, remove what was unnecessary, then read it again. Sometimes the right phrase appeared immediately. Sometimes I had to search for it for a few minutes. In the end, the email would become exactly the kind of message I was ready to send.

This time, everything happened differently.

I opened a chat with an artificial intelligence tool.

Briefly described the situation.

A few seconds later, I had a finished text.

It was genuinely good. Possibly even better than what I would have written myself on the first try.

I copied it almost without thinking and clicked “Send.”

That should have been the end of the story.

But later that evening, I found myself returning to that email.

Not because I doubted its quality.

Something else would not let go of me.

Why had I sent it without really thinking?

I had not chosen the words.

I had not looked for arguments.

I had not hesitated.

I had not rewritten anything.

I had simply accepted the first version offered to me.

And that was the moment I felt slightly uneasy.

Not because artificial intelligence had written a good text.

But because I had given up my own process of thinking far too easily.

For the past two years, we have been arguing endlessly about artificial intelligence.

Some people are convinced it will soon replace half of all professions.

Others dismiss it as an overhyped toy.

Some admire its new possibilities.

Others warn us about the uprising of machines.

But more and more often, I find myself thinking that the real question lies somewhere else entirely.

It is much simpler.

What happens to a person who gets used to receiving ready-made answers?

At first glance, the question seems almost ordinary.

But if you think about it, it concerns each of us.

Today, artificial intelligence can write emails.

Tomorrow, it can prepare presentations.

The day after tomorrow, it can draft contracts, create advertising copy, analyze documents, help with programming, explain difficult subjects, and even argue with us.

Each time, it saves us time.

That is wonderful.

But along with time, we sometimes begin to save something far more important without noticing it — our own mental effort.

And that effort is almost always what makes a person stronger.

When calculators appeared, many people stopped doing arithmetic in their heads.

When navigation apps appeared, we almost stopped remembering routes.

When the internet made almost any information available within seconds, the need to keep huge amounts of facts in memory really did become smaller.

There is nothing wrong with any of this.

Technology has always freed people from routine work.

But artificial intelligence is the first technology that does not merely free us from work.

It has begun to free us from thinking.

The difference may seem small.

In reality, it is enormous.

Asking a calculator to add two numbers is not the same as asking a machine what to write to a friend, how to answer a client, how to structure an article, or what conclusions to draw from a book you have read.

Little by little, we are no longer handing over calculations to a tool.

We are handing over our own thoughts.

And it is here, I think, that the most invisible boundary of this new era runs.

Not between human beings and machines.

But between a person who continues to think — and a person who gradually begins to entrust their thinking to ready-made answers.

The Very Expensive “Do It for Me” Button

After a few weeks, I began to notice something curious.

Almost all of my requests to artificial intelligence could be divided into two groups.

The first group sounded roughly like this:

Write this.

Come up with this.

Do this.

Explain this.

Make this.

These were the requests of someone who wanted a finished result.

The second group appeared much later.

And it sounded completely different.

Here is my text. What is weak in it?

I am almost sure of my conclusion. Try to prove me wrong.

What questions have I failed to ask myself?

Pretend you disagree with me. Object as convincingly as you can.

At first glance, the difference seems minor.

But it is precisely this difference that determines whether artificial intelligence becomes your assistant — or your substitute.

There is a precise observation that does not need an author or a date to be true.

If someone keeps carrying water for you, sooner or later you stop walking to the well yourself.

Not because you are lazy.

Simply because it is more convenient.

Our brains work in much the same way.

They are constantly looking for ways to conserve energy.

That is perfectly natural.

Whenever there is a shorter path, the brain almost always takes it.

Artificial intelligence has become the shortest path humanity has ever created.

Why struggle to solve a problem if an answer appears in five seconds?

Why wrestle with a headline if AI can generate one instantly?

Why argue with yourself if you can immediately receive a polished conclusion?

Each individual case seems completely harmless.

The danger is not in any single decision.

It lies in repetition.

Habits are rarely born overnight.

They are built from small choices repeated over and over again.

One day you ask AI to write a birthday greeting.

The next day, an email.

Then a report.

Then a presentation.

A speech.

A proposal.

And one morning, almost without noticing, you realize that you no longer begin with your own first draft.

The machine has started thinking first.

You merely edit what it gives you.

That, I believe, is the moment worth paying attention to.

Not because using artificial intelligence is somehow wrong.

Quite the opposite.

I find it difficult to imagine working without it now.

The problem is not the technology.

The problem is the role we assign to it.

Imagine someone joining a gym.

There are two possible ways to train.

In the first, you ask the instructor to lift the weights for you.

The result is excellent.

The weights are lifted.

The workout is complete.

Only one thing is missing.

You are no stronger than when you walked in.

The second approach is harder.

The trainer never lifts the weight for you.

Instead, they watch your technique.

Correct your mistakes.

Adjust the load.

Push you through one more repetition when you think you have reached your limit.

You leave exhausted.

But you also leave stronger.

Artificial intelligence can play either of those roles.

It can become the person who does everything for you.

In that case, results will come faster.

Or it can become a trainer for your thinking.

A conversation partner who refuses to replace your mind but constantly helps sharpen it.

The difference between these two approaches is enormous.

Yet surprisingly few people notice it.

Several months ago, I changed just one habit.

When I finished writing an article, I no longer asked:

"What do you think?"

Instead, I began asking something very different.

"Assume you completely disagree with this article. Tear it apart."

At first, I didn't enjoy the experience.

Arguments I thought were convincing turned out to be weaker than I had imagined.

Some paragraphs repeated the same idea.

Others introduced an important point too late.

Occasionally I discovered that I had become so pleased with a particular sentence that I had forgotten to ask whether it actually served the reader.

More than once, entire pages had to be rewritten.

But something unexpected happened.

I stopped seeing artificial intelligence as someone who wrote for me.

It became my first reader.

Attentive.

Patient.

Honest.

And completely unoffended when I disagreed with it.

That was the moment I understood something important.

A real conversation with artificial intelligence does not begin with asking it to write.

It begins with being willing to defend your own ideas.

Because a good conversation partner is not interested in replacing your thoughts.

It wants you to examine them.

Challenge them.

Strengthen them.

The quality of the conversation depends less on the intelligence of the machine than on the honesty of the person sitting in front of it.

And perhaps that is the greatest misunderstanding surrounding artificial intelligence today.

Most people judge it by the quality of its answers.

I have gradually come to judge it by the quality of the questions it encourages me to ask.

The Question Worth Asking Yourself

People sometimes ask me whether artificial intelligence worries me.

Honestly?

Not really.

What worries me is something else entirely.

One day, I might get used to letting someone else do my thinking for me.

Whether that "someone" is another person or a machine makes very little difference.

The ability to think independently is rarely lost all at once.

It fades quietly.

A little at a time.

Each time we accept the first answer without questioning it.

Each time we stop looking for our own arguments.

Each time convenience becomes more important than curiosity.

That is why, whenever I open an AI chat today, I try to remember one simple idea.

My goal is not to get an answer as quickly as possible.

My goal is to leave the conversation thinking a little better than I did ten minutes earlier.

If that hasn't happened, then I have reduced one of the most remarkable inventions of our time to little more than an expensive typewriter.

And that feels like a wasted opportunity.

Over the past few months, I have noticed something else.

The most rewarding conversations with artificial intelligence almost never begin with the words:

"Write this for me."

They begin very differently.

"This is what I think. Where am I wrong?"

"Here is my idea. Try to destroy it."

"I feel I'm overlooking something. Help me see what it is."

The moment you ask questions like these, something changes.

You stop being a customer placing an order.

You become a participant in a conversation.

It is no longer about getting a service.

It becomes a process of discovery.

Perhaps that is the greatest strength of artificial intelligence.

Not that it always has the right answers.

It doesn't.

Not even close.

Its greatest strength is that it is willing to stay in the conversation for as long as you need.

It doesn't grow impatient.

It doesn't become offended.

It doesn't roll its eyes because you have changed your mind for the third time.

It simply keeps asking you to think a little more carefully.

A little more honestly.

A little more deeply.

No, that doesn't make it human.

But it does make it an extraordinary tool for becoming more fully human yourself.

I sometimes wonder whether, a few years from now, we will stop dividing AI users into beginners and experts.

The more meaningful distinction may be something else entirely.

Some people will use artificial intelligence to collect answers.

Others will use it to ask better questions.

The first group will probably become faster.

The second may become wiser.

And I have a feeling that the future will belong to the second group.

If you remember nothing else from this article, remember just one small habit.

The next time you open ChatGPT, Claude, Gemini—or whatever comes after them—pause for a moment before you type your first prompt.

Instead of asking:

"What should I write?"

Try asking:

"Here is what I've written. What am I missing?"

Instead of:

"Solve this for me."

Try:

"Help me think this through."

The answers may be less convenient.

The conversation may take a little longer.

But you might discover something far more valuable than a polished response.

You might discover a better version of your own thinking.

Artificial intelligence did not make you wiser.

It doesn't know how.

But it helped you do, more often, the one thing that has always made a human being human.

Think.

And if that happened, then it has already done its most important job.

What do you think?

Have you noticed how your own conversations with artificial intelligence have changed over the past year?

Do you use it as a partner that sharpens your thinking — or do you increasingly catch yourself simply wanting a ready-made answer?

I would genuinely like to know your experience.

It might become the start of the next article.

After Login, everything is just beginning.