The Calculation Default: What René Descartes Teaches Us About Reasoning Models

René Descartes sought to rebuild knowledge from the ground up, trusting only what could be established with mathematical certainty. His method of systematic doubt gave us the cogito — but it also bequeathed a particular vision of reasoning: clear, distinct, calculable.

AI reasoning models inherit this Cartesian legacy. They excel at problems that can be formalised and computed. But what happens when they encounter the vast domain of human experience that resists formalisation — moral dilemmas, aesthetic judgments, existential questions?

This episode explores the “calculation default” — the tendency to treat computability as the measure of all reasoning — and asks what Descartes himself might have made of machines that reason without thinking.