Last year, I was catching up with an old friend over a round of golf. We were swapping life updates, and at some point he laughed and said something like, “I genuinely don’t know what to call you anymore. Are you still in finance? Teaching? Running the school? Writing? Pick a lane, mate.”
I laughed too, but his comment stuck with me.
For most of my adult life, I’ve worn what I used to think of as a slightly embarrassing label: jack of all trades, master of none. Finance, then teaching, then managing a language school, then a couple of small businesses, and now writing. None of it on a straight track. Plenty of zigzagging.
But I think the world is now changing in a way that makes that zigzag look less like indecision and more like an advantage — at least for some kinds of work.
What follows is one writer’s reflection on that, not career advice. I’m aware it’s a convenient theory for someone with my CV, so take it with the appropriate grain of salt.
The era of the deep specialist looks different than it used to
I grew up believing the path was simple. Pick something, get good at it, stay in your lane, climb. That was the deal my parents’ generation lived by, and it more or less worked for them. The rules underneath that deal have shifted, though.
The World Economic Forum’s Future of Jobs Report 2025 puts numbers on the shift. Employers predicts that around 39% of workers’ core skills to be transformed or outdated by 2030.
That isn’t a statistic about robots taking over. It’s about the half-life of expertise getting shorter in many fields. Plenty of professions still reward thirty years of deep practice — medicine, engineering, classical music. But for a growing share of white-collar work, the narrow knowledge that used to anchor a career now refreshes more often than it used to.
If your entire identity is built on one tool or one job title in one of those changing fields, that’s a more exposed position than it was twenty years ago.
AI competes hardest with narrow, well-defined work
When I was younger, the assumption I came to make was simple: the more specialized your job, the safer you were. Become an expert. Become irreplaceable. AI has complicated that picture, at least at the routine end.
Generative AI seems reasonably good at the kind of work that lives inside a single, well-defined domain. Drafting standard contracts. Writing routine code. Producing first-draft marketing copy. Pulling structured analysis out of a spreadsheet. The narrower and more pattern-based the work, the more AI tools can chip away at it.
Anthropic’s Economic Index, which tracks how Claude is actually being used across the economy rather than how people speculate it might be used, backs this up: real-world usage is heavily concentrated in software development and technical writing. The narrow, technical, pattern-rich work is where the tools have landed first.
The same WEF report notes that two-thirds of employers now plan to hire people specifically because they have AI skills, and 40% intend to reduce their workforce where AI can automate tasks.
It’s worth being honest about the other side of this, though. AI also likely raises the productivity ceiling for many specialists — the radiologist who reads scans faster, the lawyer who clears document review in an afternoon, the developer who ships features twice as quickly. In plenty of fields, deep expertise plus AI tools is probably still the strongest combination. The vulnerable position is narrow expertise without the willingness to adapt the tools or the surrounding skill set.
As far as I can see, AI has consistently struggled with is the messy in-between. Connecting two unrelated fields. Spotting that a pattern from biology applies to logistics. Knowing when the data is misleading you because you’ve also worked with the people behind it. That kind of cross-domain pattern matching is something generalists tend to be wired for.
The case for the curious career-jumper
David Epstein’s book Range: Why Generalists Triumph in a Specialized World makes the case that, in fast-changing and complex environments, breadth often beats depth. He writes, “The more constrained and repetitive a challenge, the more likely it will be automated.”
This squares with what the WEF report identifies as the core skills for 2025. Analytical thinking sits at the top, followed by resilience, flexibility and agility, then leadership, creative thinking, and motivation and self-awareness. None of those are domain-specific. All of them favor someone who has had to start from scratch a few times.
Why the self-teachers may have an unfair advantage
There’s a subtle thing going on with formal education at the moment. It’s still valuable, and I’m not knocking degrees. But the speed at which knowledge updates means that any specific curriculum is, by the time you graduate, a snapshot of what mattered three or four years ago.
The self-taught are different. They learn because they want to know, not because someone told them they had to. They poke at YouTube tutorials, follow rabbit holes on Reddit, build small projects, ask AI to explain things back to them, and then move on to the next thing that catches their attention. Almost by accident, they’ve been training themselves in the very skill the WEF lists among the fastest-growing for the next five years: curiosity and lifelong learning.
I see this in myself. I never studied writing formally. Most of what I know I picked up by reading writers I admired, copying them badly, getting feedback, and adjusting. The same pattern shows up in every other career I’ve had. Finance was trial by fire. Running the school was the same. Repeatedly being the new person in the room, having to figure things out without a clear instructor, builds a particular kind of confidence.
It’s the confidence that says, “I can probably learn this,” rather than, “I already know this.”
In an economy where what you know matters less than how fast you can re-learn, that’s not a small thing.
The bottom line
I’m not arguing you should ditch your specialty tomorrow. The most valuable people I’ve come across actually combine both. They go deep in something, then keep dabbling outside of it.. Specialists in plenty of fields will probably keep thriving, especially the ones who treat AI as a power tool rather than a competitor.
But here’s the harder question, and the one I keep circling back to. It’s easy to read a piece like this and feel quietly vindicated about a winding CV. Retroactive permission is cheap. The actual cost of the generalist path isn’t sentimental — it’s structural. It means choosing, repeatedly, to be the least competent person in the room. It means never quite finishing the climb you started. It means watching peers compound seniority while you reset.
So the question isn’t whether the zigzag looks smarter in hindsight. The question is whether you’re willing to keep zigzagging from here, into a future nobody has mapped yet, when there’s no guarantee the next pivot pays off. Comfort with that uncertainty is the actual skill. Everything else is just storytelling.
If you wanted permission, this isn’t it. If you wanted a prompt, maybe it is.
