A mental model that works is also a trap. Not because it gives wrong answers, but because it keeps giving right-enough answers — and right-enough is the enemy of noticing when the territory has quietly shifted beneath you. The philosopher Alfred North Whitehead, writing in *Process and Reality*, argued that reality is not a collection of fixed things but of events — ongoing processes of 'becoming' rather than stable 'beings.' Most mental models, however, are built like photographs: they freeze a moment of understanding and treat it as a permanent object. The danger isn't that your model of a competitor, a user, or a market is wrong. The danger is that it was accurate once, and that prior accuracy is now functioning as evidence that it's still true. The practical move isn't to distrust your models — it's to treat each one as carrying an invisible expiry date, stamped not by time passing but by the number of confirmations you've collected without a serious attempt at disconfirmation. Run your strongest mental model through one genuinely adversarial question this week, not to replace it, but to find where it's silently calcifying.

What would someone actively trying to break your most-used mental model point to first — and when did you last let that person exist in the room?

Drawing from Process Philosophy — Alfred North Whitehead