Articles

A learned system can tell you what your software did. It cannot tell you who is answerable for what that behaviour now obliges you to do.

The loop can write correct code unattended. What it can’t do is grant itself permission — and in a regulated estate that gap is a gate, not a better checker.

When AI entered customer service, nobody said ‘just give the agents ChatGPT and carry on.’ In software engineering, we’re making exactly that mistake.

The SDLC isn’t dying, it’s migrating from documents into the execution engine. Most organisations stall at machine-speed delivery under human-speed governance. The missing piece is a function nobody is staffed to own: interpretation binding.

AI is not just changing how work is done. It is changing the rate at which change is proposed. The constraint has moved from strategy to absorption.

Token Jail is not vendor lock-in. It is what happens when implicit authority meets machine-speed execution. The dependency lives in the operating model, not the API.