You will never be current again — a four-question triage for deciding what deserves your finite learning attention, and what you can safely stay behind on.

Somewhere on my laptop is a folder of saved tutorials I will never open. I add to it most weeks. The saving feels like progress — a small deposit into a future, more capable me. It is nothing of the sort. It is a talisman against the anxiety of falling behind, and I have been carrying some version of it for thirty years.

Here is the strange part. The World Economic Forum says 39% of core skills will change by 2030 — and that this figure has fallen, from 57% in 2020 and 44% in 2023. The measured churn is slowing while the panic about it grows louder. That gap is not an accident. An entire industry — courses, newsletters, “10 tools you must know this week” — runs on you never feeling caught up. The feeling is the product.

After four platform shifts, I have stopped trying to keep up. What I do instead is closer to what we do in an incident queue: triage. You do not treat everything. You decide, deliberately, what gets attention and what does not — knowing the second list is the longer one.

The Learning Triage

Four questions, in order. A candidate skill has to pass the earlier ones to earn the later ones.

1. Check the half-life

Technical skills decay fast — IBM puts their half-life around two and a half years, against five for broader skills. In the client-server years I learned Novell NetWare administration and TCP/IP in the same winter. One evaporated with the product; the other has paid rent for three decades. Tools expire; the principles underneath them mostly don’t.

Ask: will this still be true in five years, or is it merely current?

2. Check the adjacency

Learning compounds when it lands next to something you already know. A delivery lead picking up observability is stacking it on years of incident experience — every concept has somewhere to attach. The same person learning a fashionable framework from zero is starting a savings account with pennies.

Ask: does this multiply what I already know, or reset me to zero?

3. Check the practice surface

Carnegie Mellon’s research on online courses found learning by doing roughly six times more effective per activity than watching or reading — and the relationship is causal. I once watched a team get sent on a Kubernetes course months before any cluster existed for them to touch. By the time one did, the course had evaporated. If there is nowhere to practise, you are not learning; you are consuming content about learning.

Ask: where, specifically, will I use this in the next month?

4. Check the price of ignorance

This is the question the keep-up industry never asks, because the honest answer kills the sale: most ignorance is free. I have never properly learned front-end development. Total career cost so far: nothing — someone within shouting distance has always known it. Ignorance only costs when nobody near you holds the knowledge and a real decision depends on it. That is rarer than the newsletters suggest.

Ask: what actually breaks if I stay ignorant — and who nearby already knows it?

The uncomfortable bit

Most things fail the triage. That is the point of a triage. Being behind on almost everything is not a personal failing to be fixed by one more course; it is the permanent condition of working in technology, and it always has been. The only real choice you get is where to be behind. Make it on purpose.

Sources used: WEF Future of Jobs Report 2025 (39%, falling from 57%/44%) · IBM skills half-life research · CMU — “Learning is Not a Spectator Sport” (the doer effect) · Koedinger et al. — causal doer-effect study · Oliver Burkeman — Four Thousand Weeks (2021)