Essays and observations

We have built educational systems that generate data about students rather than insight into how they think. The measurement is not neutral — it actively shapes what teachers pay attention to, and it is pointing in the wrong direction.

Most of us have been trained to produce the right answer. Very few of us have been taught to notice when we don't actually understand something — or to sit with that discomfort long enough to resolve it.

There is enormous optimism right now about AI transforming learning. Much of it is misplaced — not because the technology is weak, but because the mental model of what education is for has not changed.

Most productivity advice optimises for output. I am more interested in what builds the kind of person who produces better output — and why the distinction matters enormously in practice.