Almost no one in education will say this sentence out loud. Schools do not know what specific jobs today's children will do.
That has, in some sense, always been true. But the gap between what a school can predict and what a working life will look like is wider now than at any point in the last three decades. Since ChatGPT launched in late 2022, AI systems have gone from interesting toys to passing professional exams, generating working software from a single sentence, and operating routine office workflows end to end without a human typing a key.
So how should a parent, or a school, prepare a child for working life when working life itself is moving this fast? The answer is not to chase the latest tool or to guess the next set of jobs. It is to be honest about the shape of work that is emerging, which can be described with some confidence, and to make sure children leave school equipped for that shape.
The shape of work is changing
Three numbers help to mark how quickly. The first is how fast the underlying technology has moved into ordinary life.
The second is how seriously the global economy has taken it. Stanford's 2025 AI Index recorded global private investment in AI reaching $252 billion in 2024 alone, a jump of more than 40% on the year before, and many multiples of the levels seen even three years earlier. McKinsey's 2024 State of AI survey, run across thousands of organisations, found that 72% now use AI in at least one business function, up from 55% the year before.
The third is the impact economists now expect on jobs themselves. The most-cited estimate comes from the International Monetary Fund: in early 2024 the IMF's managing director, Kristalina Georgieva, set out that around 40% of jobs globally are exposed to AI, meaning the work inside those jobs will be meaningfully reshaped. In advanced economies, the IMF puts the figure closer to 60%.
The honest reading of these numbers is not that work is ending. It is that work is being reshaped. The cognitive parts of almost every white-collar job are being handed off to AI, and the human role is moving up the stack. Less typing the report. More deciding what report needs writing, briefing the AI to draft it, judging whether the draft is correct, and shipping the work.
A reasonable summary of the next decade of work, for most professions, is that today's children will not be replaced by AI. They will lead a team of AIs.
This is not an inspirational poster. It is a description of what graduates are already doing in fields from law to design to software. What used to need a team of people, engineers to build, designers to draft, marketers to write copy, accountants to handle the numbers, one person can increasingly do, by directing AI agents in plain English. The team has not gone. It has changed shape. The team is becoming a conversation.
What an effective AI team leader needs
If that is the shape of work, the question for parents and schools sharpens. What does it take to be the human at the centre of that conversation? Four things, on the evidence so far.
The first is the ability to set a clear goal. AI does whatever it is told. The bottleneck is no longer execution; it is deciding what is worth executing in the first place. A child who knows what "done" looks like, and why "done" matters, can direct an AI team to a real outcome. A child who does not will produce a great deal of polished, plausible, ultimately pointless work. This is taught through giving children real problems where the answer is not in the back of the book, where they have to define what success even is. Open-ended projects, design briefs, debate motions, science fair questions are not extras; they are training for the bottleneck of the next working life.
The second is the ability to communicate precisely. Vague briefs produce vague work. This was true of human teams; it is even more true of AI ones, because AI cannot tell when it is being asked the wrong question. A pupil who has been taught to write clearly, to mean what they say and say what they mean, will give the AI usable instructions. A pupil whose writing is woolly will get woolly output, look at it, conclude the AI is not very good, and miss that the limit was their own brief. The English curriculum, taught well, is one of the most economically valuable preparations a child can receive for a career spent directing AI.
The third is the habit of asking the right questions. Better questions get better answers. The AI will answer almost any question; the value lies in the question itself. This is harder to teach than it looks, because most schooling, by accident rather than design, rewards children who give the right answer to questions adults set. The skill of generating a good question, of noticing what is interesting, of pulling at a loose thread, is different. It is the skill that separates someone who uses AI to crank out adequate work from someone who uses AI to do extraordinary work.
The fourth is critical judgement of the answers. Confident output is not the same as correct output. AI is uncannily good at producing things that sound right, are well-structured, look like the genuine article, and are wrong. A child without the underlying knowledge to spot when the model has hallucinated a fact, or made a subtle error of reasoning, will sail past the mistake and ship it. Subject knowledge therefore matters more, not less, in an age of AI. A pupil cannot judge an answer in a field they do not understand. Reading, mathematics, history, science are not nostalgia. They are the substrate that makes critical judgement possible.
What this asks of a school
These four skills are not a radical syllabus. They are most of what a serious British education has always tried to do: teach children to read carefully, write precisely, ask hard questions, and trust their own judgement against the loudest voice in the room. What changes, in an age of AI, is the seriousness of the why.
When the bottleneck of execution is dissolving, what a child can plausibly attempt grows enormously. A pupil with a good idea, strong fundamentals, and an AI team to direct can build, write, design, and ship work that would have taken a team of professionals five years ago. Schools have started to see this in classrooms, and what it reveals is consistent with the larger picture: AI rewards the children who arrive prepared, and pulls further ahead the ones who do not.
Almost every working life ahead will involve directing AI, judging what it produces, and being responsible for the result. ISJ has written separately about how schools can prepare pupils for this kind of judgement, and about why teaching quality remains the single biggest school-based influence on what a child becomes.