Reskilling
The Half-Life of a Skill Is Shrinking Faster Than You Think
The thing you spent three years learning to do well may be genuinely obsolete before you've had a chance to feel expert at it.
The Idea
There's a concept that rarely gets discussed outside workforce economics circles: skill half-life — the time it takes for half the knowledge required to do a job competently to become outdated. In the 1980s, engineers reckoned their technical training would stay relevant for roughly a decade. Today, depending on the field, some researchers put that figure closer to two to five years. In AI-adjacent roles, it may be even shorter. What's genuinely surprising here isn't the pace of change itself — most people sense that things are moving fast. It's the asymmetry in how this affects different kinds of knowledge. Procedural skills — how to use a specific tool, run a particular process, operate a piece of software — decay fast. But the cognitive habits underneath them: systems thinking, knowing how to learn something new, judgment about when a tool is the wrong tool — these compound rather than erode. They become more valuable precisely because everything around them is churning. This reframes the reskilling conversation entirely. The dominant narrative — learn to code, get certified in AI, take the course — treats skills as fixed assets you acquire. But that model is itself outdated. What the evidence increasingly supports is that the people who adapt best are less skilled in one thing and more practiced at the meta-skill of updating their own mental models. Not what you know, but how quickly and honestly you can unknow it.
In the World
In 2019, Amazon announced a programme called Upskilling 2025, pledging to retrain a significant portion of its workforce for more technically demanding internal roles — machine learning specialists, data analysts, solutions architects. The headline number was ambitious. But when researchers and journalists began examining how it actually worked, something interesting emerged: the employees who moved most successfully into new roles weren't necessarily those who completed the most modules or clocked the most training hours. They were disproportionately people who had already crossed a professional boundary once before — who had, at some earlier point in their career, made a significant pivot and survived the disorientation of being a beginner again. The psychologists have a term for this: transfer learning (borrowed from machine learning, fittingly). The more times you've rebuilt your competence from scratch, the less threatening the process becomes. The fear of not-knowing loses its grip. You stop mistaking unfamiliarity for incapacity. This pattern holds beyond Amazon. Studies of German manufacturing workers retraining after automation, of mid-career professionals entering technology roles, and of nurses pivoting into health-tech all point in the same direction: prior experience of deliberate reinvention is the strongest predictor of successful reskilling — stronger than age, education level, or the specificity of the new skills being taught.
Why It Matters
If the half-life of skills is genuinely compressing, then the single most valuable investment you can make in your working life isn't the next certificate or the hottest new tool. It's engineering regular, low-stakes experiences of being a beginner — picking up something unfamiliar, feeling lost, and finding your way through anyway. This matters because most adults, especially competent ones, unconsciously arrange their lives to minimise that feeling. Expertise is comfortable. Being the person who doesn't know yet is not. But every time you avoid that discomfort, you're also letting the muscle for navigating it atrophy slightly. It also changes how you might think about your organisation or team, if you're in a position to shape one. Reskilling programmes that deliver content are common. Cultures that reward people for transparently not knowing something yet are rare, and far more valuable. The question isn't whether the people around you can be trained in new tools. It's whether they've practised being wrong and rebuilding before — and whether the environment makes that feel safe.
A Question to Ponder
When was the last time you were genuinely a beginner at something that mattered to you — and what did that experience reveal about how you actually learn?
Get a new one of these every morning.
Start learning with Thinkable