13.11.2026 · Learning and family

What should you study if professions change faster than curricula?

Picture a conversation at the kitchen table.

— You need to get a good profession.

— Which one?

— Well… one that's in demand.

At this point the adult wants to pull out a list. Preferably an official one, stamped and guaranteed until retirement. Meanwhile the teenager has already seen AI write code, put together a presentation and draw a picture. And asks a perfectly reasonable question: “While I'm studying, what will be left of this profession?”

Only yesterday the family was choosing between a few clear roads. Now it's discussing whether the road will still exist by the time university is over. And the adult being asked for an answer doesn't know either.

I take this question seriously. In Capybara, I'm interested in how to help a person develop their own thinking, learn and see the results of their actions. Now another task is added: building a working life when the very conditions of work are changing.

That's why I've prepared an open instruction for talking with AI about your professional path. Below I explain why it's built this way and give it in full. First, let's look at what's happening with this question in Europe.

This conversation is already happening in many European families

I looked at publications and research from ten countries: the UK, Ireland, France, Germany, Spain, Italy, Poland, the Czech Republic, Romania and Estonia. They cover different parts of Europe but don't represent every country or every family. You can't derive a single “level of European panic” from this selection. But you can see recurring questions and different ways of answering them.

In France, Le Monde in March 2026 examined creative career choices through conversations with parents and young people. Some families try to steer children away from illustration, games and film; others support creativity despite the uncertainty. This is reporting on specific families, not a measure of the whole country's mood.¹

In Germany, a June piece by tagesschau asks the question directly: should newcomers to the job market fear AI? It also shows the difference between expecting big market changes and fearing for one's own job. These states don't automatically coincide.²

In the UK, the family magazine MadeForMums wrote on 10 November about parents who find it hard to advise children on modern work and qualifications. The survey cited is linked to a Samsung campaign: a reason to look closely at the problem, not an independent census of parental opinion.³

In other countries the same uncertainty shows up in its own way:

Where	What's being discussed	What it lets us see
Poland	Radio PiK reports on a forum where young people discuss which studies and jobs to choose as AI develops.4	Young people ask the question themselves; alongside fears there's interest in the opportunities.
Romania	Știrile ProTV reports unfilled places in some IT programmes and young people choosing other engineering fields.5	Expectations may already be influencing decisions; the report doesn't prove a single cause of the change in enrolment.
Estonia	ERR discusses IT students' difficulties entering work and the change in entry-level tasks.6	Access to first experience is becoming a question in its own right.
Czech Republic	In an interview with Novinky, an Alma Career analyst talks about limiting part of junior hiring without mass replacement of workers.7	The market can change through jobs that are never created, not only through layoffs.
Spain	La Vanguardia writes about the difficulty of choosing further education, based on Stuveo data.8	The uncertainty is wider than AI: working out education and one's own choice was hard even without it.
Italy	Sky TG24 discusses how to choose a university when professions are changing.9	The question of speciality is joined by the question of how exactly a person will learn.
Ireland	Reuters reports early signs of worsening opportunities for young people in parts of the tech economy. The causal conclusion about AI remains limited.10	Even noticeable changes need their explanations checked.

I think an important shift is happening here. Parents have to rethink the very way they give advice. Past experience remains valuable but no longer lets them confidently describe a child's whole path twenty years ahead.

The anxiety is real. Its explanation is more complex than the headline

This confusion has a measurable background. In an OECD report published in 2025 based on PISA 2022 data, 39% of 15-year-olds on average across OECD countries are placed in the group with uncertain career expectations. In the same data, fewer than half of students had had contact with employers through a careers fair, a workplace visit or work experience.¹¹

The dates matter here: the data refer to 2022. They can't be used to prove that mass use of generative AI caused this uncertainty. Career guidance already had problems with access to experience, information and breadth of choice. AI adds one more fast-changing layer.

In a King's College London study in spring 2026, only 31% of surveyed parents of children aged 11–29 said they had discussed with their child how AI might affect their future career. The base for this part of the survey is 729 people.¹²

For me, that's a reason to start the conversation. A parent doesn't first need a degree in labour-market expertise. You can admit what's unknown and work out together where information is needed, where a trial is needed, and where help from someone in the profession is needed.

But it would be wrong to say young people everywhere are gripped by fear. In the German BARMER study 2025/26, a minority expresses strong concern about AI; other worries are more prominent.¹³ In a study by the Polish THINK! foundation, the share of young people oriented exclusively towards employment is growing: from 28% in 2022 to 35% in 2026. This is a study of preferences, not a forecast of the number of jobs.¹⁴

People can use AI, see opportunities in it and still want predictable work at the same time. That's an understandable position, and you also need to be able to build a route from it.

The most awkward question: where does a newcomer become a specialist?

Part of professional learning has traditionally happened on simple tasks. A newcomer gathered information, prepared a draft, fixed a spreadsheet, wrote a small piece of code. Gradually they understood the context and got more complex work.

When such tasks can be delegated to AI, the question arises: what will the path to experience look like?

A French study presented by HEC Paris in October 2026 and carried out with the Banque de France shows that between November 2022 and March 2026 the monthly hiring trend for 21–29-year-olds on permanent contracts fell by 28.3% in the occupations most exposed to AI, compared with 7.5% in the least exposed. This concerns hiring flows in the private sector studied, not an equal reduction of everyone employed. The authors note separately that AI causality is not established by this comparison.¹⁵

At the same time, a Eurofound analysis of the 27 EU countries, published in September 2026 on data up to 2025, does not confirm a separate youth employment shock caused by generative AI. It looks at the wider market picture and actual use of the technology.¹⁶

These results can't be mechanically added to or subtracted from each other. They study different groups, periods and indicators. But together they keep us from two hasty conclusions: both from the certainty that nothing significant is happening and from declaring the end of the professional future.

Even the Irish story changes on closer reading. In a November speech, the Governor of the Central Bank of Ireland stresses that the fall in young people's ICT employment coincided with a general correction in tech hiring, and the data don't yet clearly confirm the story of AI destroying junior work in particular.¹⁷

So I would check each direction closer to the ground: how people get into it now, who trains newcomers, which tasks they're trusted with and how readiness for work is demonstrated.

What future I suggest taking into account

My working scenario is a deep restructuring of the professional world. I think a significant share of familiar expensive services will become cheaper, small teams will gain the capabilities of large organisations, and more people will start creating their own products and services. In some segments people will especially value an author's view, beauty, trust and a recognisable attitude to the work.

I want to prepare myself and my children for such a future. But this is my perspective, and it needs checking and refining. The material I found doesn't show that employment is disappearing everywhere or that running your own business will be the right solution for everyone.

The ILO–NASK 2025 study estimates potential exposure to generative AI at roughly a quarter of global employment. The authors consider transformation of work a more likely overall outcome than full replacement and warn that exposure does not equal actual job losses.¹⁸

What matters to me is preparing for several scenarios. If automation speeds up, a person knows how to use the tools and create useful results. If change is uneven, they know how to find their place in a specific environment. If employment remains a large part of life, they can work in a team, learn from colleagues and show quality.

Entrepreneurship also has an everyday side: finding clients, agreements, deadlines, bookkeeping, rejections. It suits some people. Others prefer a partnership or a team. This choice should be explored without romanticising it.

So how should you learn — and what for?

I would start with a more detailed look at what lies behind a profession's name.

Say someone wants to work in design. This wish may combine love of visual form, interest in space, a desire to tell stories, pleasure in working with materials or a wish to help people find their way. These are different foundations for development. They're worth seeing before buying a course.

Then come specific questions: which actions you really enjoy, what result others need, where serious subject knowledge is required, what AI changes and what it costs to enter the field.

When choosing a study programme, I suggest looking at the following:

• Foundations of knowledge. What will let you understand the task and check the result even when the tools change?
• Practice. What real work, mistakes and corrections will the person go through during their studies?
• Feedback. Who will be able to explain why a result is good or bad?
• Use of AI. What may a tool help with, and what abilities must the student demonstrate on their own?
• Entry into work. How are internships, mentoring and first projects organised?
• The ability to change direction. What will remain useful if you change specialisation?

A degree and access to qualifications are necessary for a number of fields. Your own project doesn't replace professional requirements. But an attractive programme name says little on its own about the quality of the future experience.

The universal advice “just learn to use AI” is also too short. To judge its work, you need to understand what exactly you're doing. For myself I keep one simple question: after this work, do I understand the task better — or have I just got a finished file?

The Capybara instruction: exploring your path with AI

For exactly this kind of conversation I've put together a portable instruction. You can attach it to a chat with AI and go through the interview in small portions.

First the conversation partner finds out your situation. Age or age range is clarified along the way; gender and biography aren't assumed in advance. Then the conversation goes through concrete experience: what you were into, what you did, what results you got, what conditions suit your life and work.

From this, several professional hypotheses gradually take shape. Each is checked from different angles: interest in the actions, existing skills, demand, the role of AI, an accessible way to earn, and the time and money needed to start.

Then a route emerges:

Horizon	What you want to gain
Next few weeks	One or two small trials and new knowledge about yourself and the field
Next few months	Practice, the necessary knowledge, feedback, a portfolio or team experience
Several years	An area of mastery, a life that suits you and the possibility to revise the direction

If income is needed urgently, this is taken into account right away: accessible earnings and developing a new direction can run in parallel. You can't demand months of self-exploration from someone who has no money for everyday life.

This is an author's tool for structuring a conversation. I don't present it as a validated psychological test or a researched method with proven effectiveness. Its usefulness is to be judged by whether it helps you think more clearly and take feasible action.

What a small trial looks like

Let's take a hypothetical example — this isn't a participant's story or a promise of income.

A person likes drawing but doesn't know what activity to connect it with. Instead of buying a course straight away, they can pick a small task: make a clear illustrated guide for a familiar audience. First find out what people find hard to understand, then prepare the material, show it to a few readers and see whether they manage to use it.

Such a trial reveals several things at once. Do you like finding out someone else's problem? Can you explain through images? Do you have the energy to redo the work after comments? Which parts does AI speed up, and where is your own checking needed?

One result doesn't yet prove professional suitability. But it gives grounds for the next attempt. If you want to explore earning, that's a separate check: who chooses such a result and what they're willing to pay for. A kind review and a paid order mean different things.

Even the conclusion “I love drawing, but I don't want to work with clients yet” is useful. It helps you choose conditions rather than blame yourself for lacking a calling.

What parents can do

The parent still has a big role: giving access to experience, helping to see limitations, discussing the cost of education and supporting the testing of interests. You don't need to know the one profession of the future for that.

You can look together at what a specialist's working week consists of. Find someone who'll show both the interesting part of the job and the routine. Help organise a small project. Discuss how much the family can spend on education without risky commitments.

At the same time, it matters that the participant goes through the interview themselves. Especially if they're already an adult. Parental anxiety is understandable, but it shouldn't become the model's hidden task: to persuade them to choose the 'right' direction.

If you're going through the instruction for yourself, you can honestly answer “I don't know”, skip questions and correct the AI's conclusions. It may spot a pattern too quickly or offer an attractive explanation. It helps to bring it back to the grounds: “Which of my examples did you base that on?”

You have the right to keep a hobby outside of earning. You have the right to choose employment. You have the right to first secure stability for yourself and then widen the space for experiments.

How to start

1. Open a new chat with your chosen AI.
2. Attach the file or paste the full text from the box below.
3. Write: “Start the interview using this instruction”.
4. Answer at your own pace. To stop and continue there are the commands “Pause”, “Show the map” and “Save context”.

If the model has no internet access, you can still do the interview. Current information about professions, salaries and training will need to be checked separately. Don't share unnecessary personal data in the conversation, and compare important decisions with real experience and competent people.

I can't give a family a map of the world for the rest of their lives. I can offer a way to refine it as you go. You can start with one question about your own experience and one small trial, after which it becomes a little clearer where to go next.

Sources and limits of the research

Materials checked on 13 November 2026. This is a qualitative review of publications and research, not a systematic review of the whole European press. Surveys describe the answers of specific groups; reports describe cases and opinions; hiring data describe observed changes. These kinds of evidence are not interchangeable. The author's suggestions above and the instruction below have not been tested by the listed studies as a single intervention.

Sources (18)

1. Le Monde — Parcoursup : chez les jeunes, l’IA bouscule les rêves de métiers créatifs, 12.03.2026, updated 13.03. The article is marked as subscriber-only; the accessible text was used.
2. tagesschau / ARD — Müssen Berufsanfänger Künstliche Intelligenz fürchten?, 18.06.2026.
3. MadeForMums — on parents' difficulties discussing modern careers, 10.11.2026. Censuswide survey, 1,000 parents; the publication is linked to a Samsung campaign, representativeness has not been independently verified.
4. Radio PiK — Jakie zawody może wykonywać sztuczna inteligencja? Jakie studia wybierać w czasach AI?, 21.09.2026.
5. Știrile ProTV — Facultățile spre care au început să se îndrepte tinerii, 01.09.2026.
6. ERR — Coding school founder: People who do not use AI universally disadvantaged, November 2026. Interview and expert opinions, not a study of the whole market.
7. Novinky — interview with an Alma Career analyst, 29.09.2026.
8. La Vanguardia — on uncertainty in choosing education, based on Stuveo data, 07.09.2026, updated 08.09. Sample of about 20,000 people; its national representativeness could not be established from the publication read.
9. Sky TG24 — AI e Università, come scegliere i corsi di laurea per i lavori del futuro, 19.08.2026. The survey of American students mentioned in it is not used here as statistics for Italy.
10. Reuters — AI adoption already hitting Irish graduate jobs, finance department says, 18.02.2026.
11. OECD — The State of Global Teenage Career Preparation, 2025; official summary of results. Main data: PISA 2022.
12. King's College London — AI and the Future of Work, full report, May 2026, printed p. 9 / ninth page of the PDF; fieldwork for the parent group 16–22.04.2026.
13. BARMER — Youth study 2025–2026: young people's expectations of AI.
14. Fundacja THINK! — Aspiracje zawodowe ludzi młodych w kontekście nowych technologii i sztucznej inteligencji, 2026, pp. 4–5, 10–12. Ariadna online panel, 1,057 people aged 18–30, 29.04–04.05.2026.
15. HEC Paris — L’embauche des jeunes recule davantage dans les métiers exposés à l’IA, 09.10.2026. Institutional summary of joint work with the Banque de France; the limits of causal inference are given by the authors.
16. Eurofound — AI is already in use in Europe’s workplaces, but not at the expense of young workers, 16.09.2026, EU27, data 2011–2025.
17. Central Bank of Ireland — Making Sense of a Changing Economy, speech by Gabriel Makhlouf, 11.11.2026.
18. ILO–NASK — Generative AI and Jobs: A Refined Global Index of Occupational Exposure, 20.05.2025; ILO explanation of the results.

Read also

Learning and familyHow to Help Children Succeed in a Post-AI Transformation WorldReadLearning and familyHow teaching children to children allows elders to record their knowledgeReadLearning and familyThe worked example makes sense, but solving it alone doesn't workRead