From antiquity to 2000. That’s roughly how long it took to turn “why” into something a machine could compute. Aristotle started it. To know a thing, he argued, is to grasp its why — its causes, not just its surface. Knowledge was explanation, never description. Hume broke it. Causation, he said, is never actually observed — we see one thing follow another and call it cause out of habit. What felt like bedrock became inference. Lewis rebuilt it with counterfactuals: C caused E if, without C, E would not have happened. And in 2000, Judea Pearl gave the whole question its mathematics — causal diagrams, interventions, a formal grammar for “what if.” Twenty-three centuries of philosophy, finally executable. I have that first edition on my desk. What strikes me is what’s still unfinished. Most of today’s energy goes to the machine’s why — explaining why a model produced an output. Important work. But there’s an older why that stays stubborn: why a person chose this over that. Why a customer stayed, or left, or never came at all. Pearl built the engine. Pointing it at human decisions, rather than algorithmic ones, is the part still open. Aristotle would have found that question familiar. @yudapearl