
Intro & Evolution of Natural Language Understanding, Pt. 1
Christopher Potts, Professor and Chair of Linguistics at Stanford and professor by courtesy of Computer Science, opens Stanford's XCS224U course on natural language understanding with a history of the field. The lecture traces how NLU moved from early rule based and symbolic systems through statistical methods to the current era of large neural language models, framing the course's central question of what it actually means for a machine to understand language rather than just produce fluent text. Potts situates the field within linguistics and computer science together, pointing to the recurring tension between engineering progress on benchmarks and deeper questions about meaning, reference, and context. Drawn from Stanford's graduate course CS224U, the talk runs 74 minutes and sets up the terminology and historical arc the rest of the course will build on, giving newcomers the vocabulary to follow later lectures on specific models and tasks.