For most of a person's life, they are accompanied by the question: what do you want to be?

They start asking children this. Later, the teenager chooses a track, a university, a specialty. It is assumed that a suitable place exists somewhere for them: you just need to understand yourself well enough, get the necessary knowledge, find this place, and occupy it.

In this model, choosing a profession is one of the main decisions at the start of life. Mistakes are costly, which is why people are advised to approach the choice seriously.

But what if the very framing of the question is gradually becoming obsolete?

Not because professions will disappear. And not because education will stop being necessary.

But because the lifespan of an economic opportunity can become much shorter than the lifespan of a human career.

A good profession today promises nothing about tomorrow

In different periods, completely different occupations suddenly become profitable.

The growth of e-commerce creates a huge demand for couriers. Salaries rise, the barrier to entry costs almost nothing, and a person without years of training suddenly gets an income comparable to those who studied for many years.

Then the supply of workers grows, technologies change, the market saturates — and the same job ceases to be so attractive.

The arrival of foreign manufacturing creates another opportunity. In an industrial region, a young person can leave school, get a job at a modern plant, train within the company, and start earning earlier than a peer who spends the next five years at a university.

A few years later, a new wave arrives.

Today, practically a schoolchild can find a hundred operating companies on the internet with outdated websites, use AI to make a modern version in a few hours, and send proposals to the owners. Ninety-seven companies will ignore the email. Two or three will agree — and the experiment will prove profitable.

But this opportunity is not eternal either.

When thousands of people start using the same trick, the price will drop. When creating a new website becomes almost automatic, the advantage itself will disappear.

The problem is not that a person chose the wrong profession.

The world simply changed.

Therefore, perhaps we think too much about choice

Let's imagine two eighteen-year-olds and give each six years.

The first one chooses a profession.

They enter a university, go through a sequential program, pass exams, and by the age of twenty-four acquire a serious amount of specialized knowledge and a recognized diploma.

The second one doesn't choose anything definitively.

They take a job. Six months later, they realize it doesn't suit them. They try another. They try to sell something. They fail. They master a new tool. They do a small project. They find their first clients. The niche disappears. They move to another field. They learn something new again.

By the age of twenty-four, the first person's biography looks clear.

The biography of the second looks chaotic.

About the first, people say they invested six years in their future.

About the second, it's easy to say they couldn't make up their minds for six years.

But perhaps they were simply accumulating different types of capital.

The first was getting better and better prepared for a specific activity.

The second was getting better and better prepared for the necessity of changing activities.

We are good at measuring the first. For the second, we don't even have a universally accepted unit of measurement.

Mistakes also produce knowledge

If a person studied at a university for two years and realized they wanted to do something else, those two years are often perceived as a loss.

If they changed five jobs, it might look even worse.

But not every unsuccessful attempt is lost time.

A person does not know themselves in advance.

They might imagine they love programming until they try doing it eight hours every day. They might consider themselves a bad salesperson until they accidentally end up in sales. They might dream of a big corporation and realize a month later that they cannot stand such an environment.

Some knowledge cannot be obtained from a book.

You cannot read a detailed enough description of a profession and know if you will enjoy doing it for the next ten years.

For this, an experiment is needed.

In that case, a person who tried five occupations and abandoned four did not necessarily fail four times.

They eliminated four hypotheses about their own life.

And along the way, they learned something about markets, people, money, and their own capabilities.

The market can become a curriculum itself

A person trying to make money by building websites for companies does not have a pre-compiled curriculum.

But its functional analogue arises very quickly.

If they don't know how to find potential clients — they have to learn how to search.

If no one replies to emails — they have to learn how to formulate an offer.

If a client is interested but doesn't buy — they have to figure out sales.

If the website turns out poor — they have to master design and technology.

If the work takes three days, and a competitor does it in three hours — they have to master AI.

If there are many clients, but still no money — they have to understand pricing.

No one decided in advance that the person needed to study these specific topics.

The necessity arose before the knowledge.

This is the opposite sequence compared to most formal education.

The traditional model often looks like this:

first knowledge → then search for a place where it will be useful.

The experimental one:

first task → then discovery of missing knowledge → learning → immediate verification.

The second model does not cancel education. It changes the moment at which it appears.

Sometimes it's better to go to university later, not earlier

Imagine a person who spent several years trying different occupations and at twenty-one realized they wanted to design bridges.

Now they genuinely need serious mathematics, physics, engineering training, and formal qualifications.

They go to study.

But their situation is fundamentally different from that of a seventeen-year-old teenager who chose a specialty almost blindly.

They already know why they need this knowledge.

Education ceases to be a bet on an unknown future and becomes a tool to achieve a discovered goal.

This is not an argument against universities.

Doctors, engineers, researchers, and many other specialists cannot be trained through a few conversations with AI and a series of small projects.

But it does not follow that the exact same order suits every person:

first choose a profession → prepare for several years → only then test your choice against reality.

Sometimes the reverse might prove more rational:

try → discover a direction → understand limitations → learn what is truly necessary to go further.

AI dramatically lowers the cost of trying

Previously, experimenting was significantly more expensive.

If a person wanted to try building a software product, they needed to learn to code or find a programmer.

To do design — a designer.

To research the market — time and competencies.

To write a commercial proposal — another skill.

Therefore, between the thought "I wonder if it will work?" and actual verification, there could be months of work and significant money.

AI is beginning to shorten this distance.

A single person is already capable of using a machine as a programmer, researcher, translator, designer, analyst, and sales assistant — even if still imperfectly.

And this is beginning to be reflected in the real economy.

In 2026, Stripe reports that the share of solo founders among companies created via Stripe Atlas has reached an all-time high. At the same time, the number of solo entrepreneurs with unusually high revenue is growing. AI does not prove that everyone can build a successful business alone, but it reduces the number of different specialists a person needs just to test an idea.

This is a fundamental change.

If previously an unsuccessful experiment could cost a year and a lot of money, tomorrow some experiments could cost a week.

Which means a person can afford many more mistakes.

But an opportunity is easy to confuse with a profession

Another trap arises here.

Suppose today building websites with the help of AI brings a person several thousand dollars a month.

One might conclude:

"I found a profession."

But precisely this could turn out to be a mistake.

If entering a niche became cheap for one person, it became cheap for a thousand of their competitors.

Today's high income guarantees nothing.

Therefore, the value of a specific opportunity lies not only in how much money was made from it.

It may leave something else behind:

the person learned how to find clients;

understood how to test demand;

learned how to sell;

mastered several tools;

saw how quickly an advantage is copied;

learned how to close a failing project;

and most importantly — gained experience in finding the next opportunity.

Then the disappearance of a niche ceases to be a catastrophe.

It wasn't supposed to last a lifetime anyway.

Perhaps the main asset becomes the ability to rebuild oneself

We are used to describing human capital through accumulation.

A person knows mathematics.

Knows how to program.

Has ten years of experience in logistics.

Got a medical education.

But in a rapidly changing environment, another type of ability emerges:

see → understand → learn → try → get feedback → adapt.

You can know a lot and be bad at going through this cycle.

Or you can have relatively little formal education, but be able to enter a new field deeply enough within a few months to start creating value in it.

This ability can be called adaptation capital.

This is not a stock of specific knowledge.

It is the ability to recreate the necessary stock of knowledge after the environment changes.

And here an uncomfortable question arises for education

Labor market research already shows that a diploma and the ability to do a job are not the same thing.

Employers spent decades adding bachelor's degree requirements even to vacancies whose content did not substantially change. Later, the reverse movement arose — skills-based hiring, where some companies began removing formal diploma requirements and describing necessary abilities more specifically.

Yet even this change is slow: announcing the abandonment of degree requirements is much easier than actually restructuring hiring.

In parallel, the OECD in 2026 draws separate attention to informal learning — knowledge and skills that a person acquires through work, everyday activities, and self-study. This form of learning is particularly suited to an environment where skill requirements change rapidly under the influence of digitalization and AI.

A strange situation arises.

The economy increasingly demands the ability to learn constantly.

Technologies increasingly allow learning directly at the moment of necessity.

Yet the social system remains very good at recognizing only pre-standardized educational trajectories.

Perhaps the problem is not that formal education is useless.

The problem is that we know how to measure accumulated education much better than a person's ability to keep educating themselves.

A ready-made trajectory both helps and deprives you of practice

Formal education has a huge advantage.

It relieves a person from the need to constantly decide what to do next.

Here is a subject.

Here is a teacher.

Here is the material.

Here is the deadline.

Here is the success criterion.

Here is the exam.

Here is the next course.

For a person who is not yet capable of building a trajectory on their own, this is an extremely valuable system.

But this advantage may have a downside.

While someone else spends years answering the question "what should you do next?", a person doesn't have to learn to answer it themselves.

An experimental life forces you to do this constantly:

What is happening right now?

Where did an opportunity appear?

Why didn't the previous attempt work?

What don't I know how to do?

Is it worth learning?

Can AI close this gap?

Should I continue or quit?

What should I try next?

This is education too.

It just doesn't come with a diploma.

AI can make independence less elitist

The experimental path has always had a serious problem.

It requires independence.

Not every eighteen-year-old is capable of researching the market, choosing a direction, compiling a study program, finding clients, evaluating the result, and figuring out what to do after a failure.

Therefore, ready-made trajectories were not just a restriction. They solved a real problem.

But AI is capable of changing this too.

A person no longer needs to know independently how to build every next step. They can discuss possible directions with a machine, explore an unfamiliar field, set up cheap experiments, analyze mistakes, and quickly get missing explanations.

AI does not create curiosity and does not force a person to act.

But it can lower the intellectual cost of independence.

Then the ability to build one's own trajectory ceases to be the privilege of a small number of people who happened to learn it on their own.

It can be developed too.

Perhaps we are asking children the wrong question

"What do you want to be?" is a good question for a world in which the answer remains useful long enough.

But if a person lives to be eighty, and technologies, markets, and ways of creating value change radically several times over their career, a single answer may not be enough.

Then the question should sound different:

Will you be able to figure out who to become next?

Not necessarily an entrepreneur.

Not necessarily a programmer.

Not necessarily changing professions every year.

The point is not constant movement for the sake of movement.

The point is the absence of the assumption that a place found once must remain yours.

Perhaps the person of the future will have to discover several times where their abilities are currently needed by the world, build what is missing, and become a specialist again.

Then the final result of education is not a person prepared for a single profession.

The final result is a person capable of making a specialist out of themselves again — as many times as needed.