Quanta Magazine highlights philosopher Harvey Lederman’s claim that large language models offer a sandbox for studying which mechanisms and pressures might have produced human introspective capacity—turning LLM research into a tool for cognitive science.