According to the World Economic Forum, 65% of children entering primary school today will end up working jobs that don’t exist yet. The WEF Future of Jobs Report 2025 projects 170 million new roles created by 2030, alongside 92 million displaced. The most valuable work of the next decade hasn’t been invented.
That should change how you plan your career.
“AI prompt engineer” is a six-figure job. So is “head of remote work.” So is “AI ethicist,” “MLOps engineer,” “TikTok strategist,” “creator partnerships manager.” None of those titles existed a decade ago. Most didn’t exist five years ago. The direction is undeniable.
So how do you prepare for jobs that don’t exist? You don’t, not exactly. You prepare yourself. Different problem, very different answer.
The trap of training for a snapshot
Most people approach career planning the way you’d approach a road trip. Pick the destination, plan the route, drive.
That worked when destinations were stable. Lawyer. Accountant. Manager at a big company. The route was well-marked and the destination would still be there when you arrived.
That’s not how it works anymore. The new roles aren’t replacements for the old ones. They’re different categories of work entirely, growing out of technologies and business models that barely exist now.
If you’re studying for the destination, you’re training for a snapshot. Snapshots go out of date. The skill of arriving somewhere doesn’t.
I built my first company on the idea that I could plan my way to success. The plan looked great on paper. Reality looked nothing like it. The company I sold at twenty-seven was barely recognisable from the one I pitched investors at twenty-three. The version that worked emerged from constant adaptation, not from sticking to the original map.
That experience is the best preparation I had for the world we’re in now. The map keeps changing. The people who do well got comfortable redrawing it.
Build a stack of transferable building blocks
Here’s the reframe that’s served me best.
Stop thinking about jobs. Start thinking about capability stacks.
A capability stack is the set of underlying skills you bring to any role. They don’t change when the job title does. They get reassembled into whatever the moment demands.
Some examples from my own stack: writing clearly, structuring an argument, analysing data, building things from scratch, understanding how businesses make money, communicating with technical and non-technical people, learning new tools quickly. None of those are jobs. All of them combine into dozens of jobs that exist now and hundreds that don’t yet.
The professionals who keep landing in great roles aren’t the ones with the most impressive job titles. They’re the ones with deep, transferable capability stacks. When a new role emerges, they can plausibly do it because the underlying components are already there. They just learn the specifics.
Pick three or four capabilities valuable across many possible futures and invest deeply. Communication. Judgment. Technical fluency in whatever’s relevant to your field. Emotional intelligence. The ability to build something from nothing. These are bricks you’ll use to build a dozen different houses over your career.
Get good at the thing nobody can teach you
There’s a specific skill underrated in conversations about future-proofing. It’s the one that actually matters most.
The ability to figure things out without instructions.
Almost every emerging role I’ve seen up close was built by someone who decided to do the work before there was a job description for it. They saw a gap, filled it, and the title got invented around what they were already doing.
The first social media managers weren’t trained as social media managers. They were marketers who started experimenting with Twitter and Instagram and figured it out. The first prompt engineers weren’t trained as prompt engineers. They were curious people who spent hours playing with early language models and noticed they were getting better outputs than other people. The first growth hackers weren’t trained as growth hackers. They were generalists who looked at conversion funnels and ran experiments until something worked.
My failed startup taught me that the bias toward action becomes a problem when you act before understanding. The opposite trap is worse. If you wait until someone hands you a curriculum for the future, you’ll always be late.
Get comfortable being slightly out of your depth. Pick something nobody around you understands yet. Spend the time. Become the person who knows something about it. That’s how you get positioned for jobs that don’t have names.
Make adaptability your actual skill
Greg McKeown’s Essentialism made the case for doing fewer things, more deeply. I’d add a corollary specific to this moment. Do fewer things deeply, but make sure one of those things is being adaptable.
Adaptability sounds vague. It isn’t. It’s a measurable, trainable capability that shows up in concrete behaviours.
Adaptable people read constantly across multiple domains, not just their own field. They have hobbies that put them in beginner mode regularly. They keep relationships across different industries and disciplines. They run small experiments at work and accept that some will fail. They notice when their default approaches stop working and update faster than their peers.
I started learning Spanish a few years back. Still not particularly good at it. The point was never the Spanish. The point was the practice of being a beginner, of fumbling through something hard, of building the muscle of not knowing and continuing anyway. That muscle transfers everywhere. When AI showed up and required me to relearn how I write, I had recent experience with the discomfort of being early on a curve.
If you only know how to do the thing you’ve always done, you’ll be obsolete the moment that thing stops being needed. If you know how to learn fast, you’ll keep finding the next thing.
Position yourself at the intersection, not the centre
A pattern I keep noticing in the people doing the most interesting work right now. They sit at the intersection of two or three fields rather than the centre of one.
The lawyer who deeply understands AI. The designer who codes. The doctor who can analyze data. The marketer fluent in psychology. The engineer who can write convincingly about what they’ve built. The intersections are where new disciplines get invented, because the work that doesn’t fit cleanly in one box is often the work that didn’t exist before.
If you’re already deep in one field, the move isn’t to abandon it. Add a serious second skill that intersects with it in interesting ways. The combination becomes its own competitive position. Most people don’t bother because it’s harder than picking one lane and staying in it. That’s exactly why the intersections are valuable.
The bottom line
You can’t see the jobs of 2030 from where you’re standing. Nobody can.
Stop preparing for the job. Prepare the person.
Build the transferable stack. Get comfortable being a beginner. Pick the intersection nobody else is standing at. Learn faster than the world changes.
Do this and the next decade has a place for you. Skip it and the next decade replaces you. The roles are coming either way. The only question is whether you’re built for them when they arrive.
Build now. Or get left behind.
