What Should My Children Do?

Parents want to prepare their children for the future. That sounds obvious – until we try to answer what exactly they should prepare for.

Which occupations will exist in 2040? Which tasks will artificial intelligence take over? Which skills will remain valuable? And what should schools teach today when the goalposts move faster than curricula?

Daniel Susskind places this uncertainty at the heart of his new book What Should My Children Do? How to Flourish in the Age of AI. His answer is refreshingly measured: We do not need to pretend we can predict the future precisely. We need to prepare people to deal competently with change.

1. “Future-proof” is probably the wrong idea

Many debates about the future follow a familiar pattern: when machines take over certain activities, we search for skills they supposedly will never master. A few years ago, programming was still considered particularly safe advice. Today, AI systems generate complex code themselves.

Susskind therefore warns against the idea that we merely need to find the one human niche permanently protected from technological progress. The more capable AI becomes, the less certain that particular bet becomes.

This does not mean skills become unimportant. Quite the opposite. But the question shifts from “Which skill will AI never master?” to “Which skills help a person continue learning and acting when tasks and tools change?”

For Future Skills, this distinction matters. A list of competencies is useful only if we do not treat it as a way to simply insure ourselves against an uncertain future.

2. AI makes foundations more important

At this point, Susskind's recommendations seem surprisingly traditional: reading, writing and mathematics remain central.

Why, when AI writes texts, calculates and summarises information in seconds? Because a powerful tool does not make our own understanding redundant. Anyone who wants to assess results, ask good questions, identify contradictions and take responsibility needs a foundation from which to exercise judgement.

This is more than nostalgia for an analogue school. It is about cognitive independence. Delegating every initial attempt at thinking to a system may produce an answer faster – but may also leave us less confident in assessing its quality and plausibility ourselves.

The challenge is therefore not “AI or foundations?”, but: How can we strengthen foundations while also using the new opportunities AI offers?

3. Learn with AI – and remain capable without it

Susskind proposes a simple principle: “Teach both, test both.”

He compares this to the calculator. Schools did not abandon mathematics when calculators became faster at arithmetic. Pupils now learn both fundamental procedures and how to use technological aids sensibly.

For AI, the implications would be broader: children should learn to achieve good results with AI – while also being able to demonstrate what they understand, can express and can solve without AI .

This could apply to many subjects: develop an argument independently, then use AI to critique it; structure a problem yourself first, then generate alternative solutions; acquire foundations and later use AI as a personal tutor.

Susskind sees personalised AI tutors in particular as a major opportunity. At the same time, it remains uncertain how reliable, educationally effective and universally accessible such systems will actually be. Technology can support individual learning – it does not automatically replace good learning.

4. Prepare for problems – not job titles

One of the book's strongest ideas concerns career guidance. Susskind recommends directing young people less towards prestigious job titles and more towards meaningful problems .

This makes sense because occupations are bundles of tasks – and technology can reconfigure those bundles. The doctor, teacher, programmer or lawyer of 2040 could work very differently from today.

The underlying problems, however, often remain: people want to become healthy. Children need to learn. Conflicts must be resolved. Organisations need decisions. Societies need trust, infrastructure and direction.

For our JobRadar 2040 , this is central. Rather than simply searching for new job titles, we ask: Which tasks and responsibilities are emerging – and where are people needed despite, or precisely because of, powerful technology?

5. Readiness for the future extends beyond employability

At the Future Skills Institut, we would broaden Susskind's opening question further.

The question “What should my child do one day?” quickly directs attention towards education, careers and the labour market. This matters. But a person needs more than employability. They need to be able to shape a life.

If AI changes substantial parts of today's paid work, skills that support life beyond the workplace will therefore become more important: self-regulation, building relationships, learning, judgement, handling uncertainty, technological competence and the ability to give one's own life direction.

This connects our research on Parents, Schools & Future Skills, the Future Skills Framework, JobRadar 2040 and Lebensfähig 2040.

Perhaps the crucial question about the future is therefore not simply:

“What should my child do one day?”

But:

“What should my child develop today to continue shaping their own life when the world turns out differently from what we expected?”

What we take from this

What Should My Children Do? is not an attempt to predict the occupations of 2040. Its strength lies precisely in taking uncertainty seriously.

This yields an approach that we consider highly relevant to Future Skills: Strengthen foundations. Actively learn new technology. Do not relinquish independent thinking. Stay flexible. And orient yourself towards relevant problems rather than supposedly safe labels.

This does not guarantee a successful future. But perhaps that is precisely the wrong expectation to place on education. Education cannot make an uncertain future certain. It can, however, better prepare people to retain their agency within it.

If we cannot know today which occupation a ten-year-old will pursue in 2040, which skills would we nevertheless consider essential to give them?

Context: This 3-Minute Insight summarises selected arguments from Daniel Susskind's What Should My Children Do? in our own words and connects them to the research questions of the Future Skills Institut. The book deliberately takes an open-minded approach to technology and offers practical recommendations for education and work. How quickly AI capabilities will grow, how strongly individual occupations will be affected and what educational benefits AI tutors will deliver in everyday practice remain uncertain. Connections to the Future Skills Framework, the parent study, JobRadar 2040 and Lebensfähig 2040 are FSI interpretations.

Source: Daniel Susskind (2026): What Should My Children Do? How to Flourish in the Age of AI. Allen Lane, 240 pages; UK edition published on 10 September 2026. Further information: Penguin UK · Official book page ↗ · Daniel Susskind on education and AI ↗ · Penguin Random House · Book description ↗