You’re Already Being Sorted
AI is dividing the workforce into two permanent tracks with different salary ceilings. Most people don’t know which one they’re on — and their job title won’t tell them.
I’ve noticed something hard to unsee once you spot it. The jobs that keep getting backfilled around me — couriers, nurses, electricians, garbage collectors, general labor, blue-collar trades — aren’t the safe knowledge-economy roles everyone assumes are protected. They’re the ones that need a body in a place, doing something AI can’t touch yet. Meanwhile the office roles I hear about quietly losing headcount aren’t getting laid off. They’re just not getting replaced. That’s the sorting happening in real time. When I look at my own week — what gets prioritized, what gets cut, what a client actually needs from me versus what could be templated — I land mostly on the judgment side. Not entirely. There’s process in what I do too. But the work that’s hardest to automate is the work I spend the most time on. That’s not an accident. That’s the track I’m on.
PwC released their 2026 Global AI Jobs Barometer last week — 1 billion job postings, 27 countries, two years of data. The headline finding: AI isn’t replacing workers uniformly. It’s creating two permanent tracks. They call them “professionalised” and “democratised.”
In professionalised roles — think recruiters, analysts, lawyers, project leads — AI absorbs the routine work, and what remains becomes more valuable. The judgment calls, the client relationships, the decisions that require context no prompt can fully replicate. Those roles are seeing twice the job growth and 42% faster salary increases.
In democratised roles — think IT service managers, junior operations staff, report generators — AI makes the work itself easier to perform without deep expertise. Which means the expertise premium shrinks. You don’t need a specialist anymore. You need someone who can run the tool.
Here’s the thing nobody says clearly: your job title doesn’t determine which track you’re on. Your daily tasks do.
| |
“Your job title doesn’t determine which track you’re on. Your daily tasks do.”
|
Think about what you actually spent your time on last week. If most of it was structured inputs producing predictable outputs — first-pass research, report generation, intake processes, summarizing meetings — that work is getting automated. Not because your company made a policy decision about it. Because someone on your team already figured out how, or will by next quarter.
If most of your week was judgment calls, ambiguous problems, reading a room, making decisions that don’t have a right answer — that work is getting more valuable as everything else gets cheaper.
So here’s the real question: of the ten things you did last week, how many could a competent AI complete from your notes and a good prompt?
If the answer is more than six, you’re not automatically in trouble. But you need to be actively building toward the work that doesn’t have a clean answer — and using AI to absorb the rest, so you have the time and cover to do it. The window is open right now. In two years, the sorting will be done.
|
The takeaway
The question isn’t whether AI will affect your career — it’s whether you’re building toward the work that becomes more valuable as AI absorbs everything else.
|
|