What jobs will AI replace, and what should my child learn instead?

Short answer: nobody can tell you which jobs AI will replace, and anyone who gives you a confident list is guessing. What the evidence does tell us is which kinds of work hold their value when the tools get better, and it is not the kind of work most parents are steering their children towards. It is the deeply human kind, paired with a comfort using the tools. Both of those can be built at home, and neither needs a coding class.

This is the question parents ask me quietly, at the end of a talk, when the others have gone. Here is the honest version of my answer.

The number everyone quotes, and what it is worth

You have probably seen the claim that 85 percent of the jobs that will exist in 2030 had not been invented yet. It comes from a 2017 estimate by Dell Technologies and the Institute for the Future, and it is an expert guess, not a measurement. Treat it as a conversation-starter. But the direction it points in is real. Job titles that did not exist ten years ago include AI ethicist, prompt engineer, human-AI interaction designer and climate intelligence analyst. A decade from now there will be equivalents we cannot name.

That is the uncomfortable truth behind the question. If you prepare your child for a specific job, you are betting on a title that may not survive. If you prepare the person who will hold whatever job turns up, you cannot lose that bet.

Picks and shovels

In a gold rush the reliable money was made by the people selling picks and shovels, not by the people digging. I have spent years watching AI reshape real companies, and the picks and shovels of this economy are not technical. They are the five skills I built the book around: resilience, curiosity, adaptability, systems thinking and empathy. Creativity, critical thinking and AI fluency are the things those five produce.

Look at who gets the raise. PwC's research on the AI economy found that workers with real domain expertise who are also comfortable with AI tools command a wage premium of up to 25 percent, and rising. Notice the order. Expertise first, AI fluency second. The person who cares deeply about something, and who can use the tools to do more of it, is the one the market wants.

The work that holds its value

Some patterns are already clear, and they are more hopeful than the headlines.

Care. Nursing, teaching, social work, mental health. The care economy is among the most AI-resistant and fastest-growing sectors there is. A child who finds purpose in helping people is not choosing idealism over practicality; they are choosing excellent positioning.

Judgement. Deciding what should be done, not just what can be done. Every AI tool produces options; someone has to weigh them. That is systems thinking and ethics, and it starts with dinner-table questions like "who else is affected by that?"

Making things that did not exist. The creative collaborator path from Chapter 1, the child who uses AI to build something bigger than they could alone, is a preview of a great deal of future work.

Starting things. Entrepreneurial mindset, which is just the try-learn-adjust loop applied to working life, predicts career adaptability in disrupted fields better than any specific technical training does.

What I actually tell my own children

I do not tell them what to become. I open doors and watch which ones they walk through, and I ask about the ones they linger at. The research on this is unusually consistent: curiosity about a child's interests produces better career outcomes than directing their choices.

Three things I do say.

Go deep on something, anything. Depth in a domain plus comfort with AI is the combination that pays. The domain matters less than most parents think. The depth matters more.

Broad before narrow. Early specialisation looks efficient and is a bet on a future nobody can see. Wide exploration builds the adaptability that survives a change of plan, and there will be a change of plan.

Learn to start, fail and adjust. The child who runs a car-wash for a weekend, builds a game nobody plays, sells three cupcakes and eats the rest, is learning the loop that every future job will run on.

Where this goes next

Every one of the five skills has a chapter in the book with things to do at home this week, at every age from five to eighteen. Chapter 22 is the one about careers, and it is the most optimistic chapter I wrote, because once you stop trying to predict the job and start building the person, the future gets a lot less frightening.

Questions parents ask

Which jobs are safest from AI?

Nobody can promise safety, but the pattern is clear. Work that depends on caring for people, exercising judgement, creating things that did not exist, and starting new things holds its value best. Nursing, teaching, social work and mental health are among the most AI-resistant and fastest-growing fields.

Should my child learn to code?

If they enjoy it, yes, as one door among many. As insurance, no. Coding is increasingly something AI does well, and comfort with AI tools matters more than writing them from scratch. Depth in a subject they care about, plus the ability to use the tools, is the combination that pays.

Should my child specialise early to get ahead?

Broad exploration beats early specialisation for a future nobody can predict. Adaptability, the ability to try, learn and adjust, is what survives a change of plan, and there will be one. Let them go deep on what they love while keeping the doors open.

How do I talk about this without frightening my teenager?

Do not ask "what do you want to be?" Ask what problem they would like to help fix and what they lose track of time doing. Then look together at the tools people who do that work use today. Curiosity about their interests, rather than direction of their choices, is what the research says produces the best outcomes.

Sources

• Dell Technologies and the Institute for the Future, 2017 estimate on future job creation, discussed in Chapter 22 of Raising AI-Ready Children.
• PwC research on the wage premium for AI-fluent workers with domain expertise, cited in Chapter 22.
• World Economic Forum, Future of Jobs Report, cited in Chapter 13.