I've never been particularly good at choosing one thing.

I learned to break complicated things into smaller ones and became comfortable not understanding something immediately.

I became interested in organisations, incentives, markets and why technically correct answers often fail once people enter the equation.

I worked close enough to technology to discover that many 'digital' problems are actually operating-model, process and coordination problems wearing screens.

The scale widened: institutions, payments, regulation, public finance and transformations where no single team controls the outcome.

Writing became the place where I could follow questions that did not belong to a client brief or job description.

Increasingly, I want to turn some of those questions into things: experiments, products, frameworks and small bets on ideas I think deserve to exist.

A system that works only when everyone behaves exactly as expected does not work.

People may follow instructions for a while. Systems eventually reveal what they actually reward.

If something complicated sounds simple, someone probably did a great deal of thinking before you arrived.

Elegant slides are cheap. Implementation is where assumptions acquire consequences.

The interesting part often begins one question after the sensible stopping point.