Tech vs non tech

We keep sorting people into the wrong two boxes: technical and non-technical, as though that tells you who will get useful output from AI.

In 2026, it increasingly does not. What matters more is between AI practitioners and non-practitioners.

A practitioner is simply someone who uses these tools often, on real problems. They learn where the tools are useful, where they break, how much context to give, when to push back, and how to judge whether the answer is any good.

Some of the technical skill they lack is now supplied by the tools as they go: syntax, scaffolding, examples, implementation paths.

That does not make them engineers overnight. But it does make them more dangerous in the useful sense: generalists who can pull expertise together and turn an idea into something that works.

A non-practitioner never quite gets there, even a deeply technical one. Without the hours in the seat, they cannot see the wider picture: what the tools are good at, what they are bad at, where they fail, and how to point them at something worth doing.

Anthropic found that across hundreds of thousands of agentic coding sessions, success is less tied to whether someone has a traditional coding background than to whether they can frame the problem properly and judge the answer.

So stop asking whether someone is technical. Ask whether they have actually practised.

Because that is the one thing you cannot read your way to, delegate your way around, or buy later.

Gregg Bayes-Brown puts this into perspective in our podcast: https://youtu.be/ZlFhciMqj80