PERSON
Douglas Hofstadter
The cognitive scientist who spent five decades arguing that analogy is the core of all cognition—and who watched, with a mixture of vindication and vertigo, as AI systems began producing analogy-shaped outputs through a process fundamentally unlike the one he had identified as constitutive of genuine thought.
In the winter of 2025, something happened that Douglas Hofstadter had spent forty-six years arguing was impossible or so distant as to be irrelevant: a machine made an analogy that mattered. Not a surface comparison any lookup table could retrieve, but a structural mapping—connecting technology adoption curves to punctuated equilibrium from evolutionary biology—that illuminated both domains and changed the direction of an argument in a book about the nature of intelligence itself. To grasp why this constituted something between a vindication and a catastrophe for Hofstadter’s life’s work, one must begin where he always began: with the simplest possible act of perception, and the extraordinary machinery hidden inside it. His argument, refined from
Gödel, Escher, Bach through
Surfaces and Essences, was that
analogy-making is not one cognitive act among many but the atomic unit from which all cognition assembles itself—that classification, memory retrieval, language comprehension, and scientific