There is a tempting version of executive maturity that says: the more senior you become, the less you touch the technology. You stop building; you start governing. You stop debugging; you start delegating.
I have spent two decades in technology leadership, and I am not sure that version is right.
What changes, and what does not
Some things obviously change. At a senior level, your time is better spent on decisions — architecture direction, team structure, investment trade-offs — than on any single implementation detail. The leverage shifts from “I can build this” to “I can judge this, staff this, and make it sustainable.”
But one thing does not change: the quality of your judgment depends on the freshness of your understanding. And understanding, in technology, goes stale. Not over a decade — over a few years, sometimes fewer.
Why personal experimentation matters
I run my own infrastructure at home: virtualized Linux systems, private LLM inference, model routing, agents, and automation. It is personal work, not enterprise production — and that is precisely why it is useful.
Because it is mine, I can afford to be wrong. I can try something that would never survive an enterprise risk review, break it, learn from it, and move on. The result is a working mental model of what is actually possible, what it costs, and where the real limitations are.
That mental model shows up in professional work. It changes the questions I ask vendors. It changes how I evaluate an AI-assisted development workflow. It changes how I set expectations with my teams when they propose something new.
A practical discipline
If you are a technology executive and you have not personally run code, a model, or a system in the last few months, consider why. Not as a productivity suggestion — as a judgment-quality one.
The discipline I follow is simple:
- Build something small and real each year — a tool, a deployment, an integration — that you would not have understood fully two years ago.
- Write down what surprised you. The gap between expectation and reality is where the learning lives.
- Keep it separate from production. Personal lab work should never be presented as enterprise capability — the value is in the judgment it gives you, not the badge.
The bottom line
Delegation is a strength. Detachment is a risk. The executives who stay current are not the ones who never stop working, but the ones who keep a small, honest window into the technology they lead.