Nobody can tell you with any certainty what jobs your children will be doing in fifteen or twenty years. Not the futurists, not the consulting firms, not the people writing the reports. AI is progressing at a pace where the honest answer to “what should kids be learning” starts with an admission: we are guessing. The serious reports are guesses built on more data than mine, but they are still guesses. Saying that out loud before any list of skills is, I think, the first useful thing.
So what do the people who study this actually predict?
The cleanest signal I have come across is the World Economic Forum’s Future of Jobs Report 2025, published in January last year. It draws on a survey of more than 1,000 of the largest employers around the world, representing 14 million workers across 22 industry clusters and 55 economies. That is not the last word on the future. It is, however, the largest dataset I am aware of pointed at exactly the question parents are asking.
The report’s headline number is that employers expect 39% of workers’ core skills to change by 2030. That is down from 44% in the 2023 Worth flagging: the report itself opens its skills chapter by acknowledging “current uncertainty around the long-term impact of generative AI.” The people whose job it is to forecast this are not pretending to be sure. That is a useful tone to borrow when you read the rest of it.
The skills that keep showing up
Beneath the percentages, the same handful of skills keep appearing across the report’s lists — core today, rising fastest, and expected to be core in 2030. If your children will need ten years of practice to be any good at these, “now” may not be too early.
The report names analytical thinking as the top core skill today, with seven in ten companies considering it essential. Then resilience, flexibility and agility. Then leadership and social influence. Creative thinking. Curiosity and lifelong learning. Technological literacy. AI and big data.
The pattern, more than the list, is the part I find useful. The skills that keep showing up split roughly in two. One half is technical fluency — AI, data, cybersecurity, technological literacy. The other half is some version of “able to keep adapting when the technical fluency stops being enough” — curiosity, resilience, creative thinking. The report is hedging in its own way. It is saying: here is the toolset we know how to use now, and here is the toolset you will need when this one stops being useful.
The detail I keep coming back to is what the report identifies as the skills least likely to be replaced by current AI. After grading more than 2,800 individual workplace skills, the authors note that skills rooted in human interaction — empathy, active listening, and working with your hands — “currently show no substitution potential due to their physical and deeply human components.” Not low. Zero. That is worth sitting with.
How might you actually nurture any of this?
I am not a parent. I should say that early. Anything I am about to suggest comes from reading widely about behavior change, watching adults in my life who handle children well, and thinking about my nieces. If you are a parent of small children, you already know more than any report writer about your specific child’s wiring.
What I will say is that none of the skills mentioned are ones you sit a child down and teach in a thirty-minute block. They are environments. The job of a parent — or an aunt, or anyone in a kid’s orbit — is closer to setting up the room than running the lesson.
With that caveat — the moves the report points at are mostly ones child development researchers have been making for decades.
Curiosity is often fed by letting a child sit with boredom long enough to go looking for something interesting. Experts have found that unstructured time pushes children toward self-directed activity in ways that structured schedules simply do not.
Creative thinking follows a similar pattern. A study found that children who played freely with objects named around three times as many creative uses for those objects as children who had been shown how to use them.
Resilience works differently: Stanford psychologist Carol Dweck, whose decades of research on failure and learning shaped the growth mindset movement, argues that the best thing parents can do is teach children to “love challenges, be intrigued by mistakes.” Allowing small, recoverable failures — rather than pre-clearing the path — is often how that disposition gets built.
When I think about my nieces, I do not really worry about the specific jobs they will hold. The specific jobs are the part the report cannot tell you and the part I cannot guess at. What I think about is whether the adults around them are giving them the room to keep asking, to keep failing safely, and to keep finding new things interesting. Those are the inputs the report keeps circling back to without quite saying out loud.
The list of named skills might be wrong in five years. The list itself is one of the things that has to keep updating. The meta-skill — being someone who keeps learning, who is comfortable not knowing yet, who can pick up the next tool when the last one stops working — that one looks safe. It is also the one you can start nurturing at any age. Including yours.
