QSS 30.26 Digital Language Theory
It has long been a goal to merge computation with natural language, and by many measures the latest AI systems seem to have achieved this. But how do models like ChatGPT enable us to talk with our computers? While this is partly a technical question, AI’s facility with natural language also presupposes a broader theory about what language is—and what, therefore, terms like “writing,” “communication,” and “understanding” (among others) mean. This course will survey various attempts to make language computable over the past two hundred years. We will consider the above terms as we sample from the history of natural language processing, various literary theorists and linguists, as well as artistic uses of language modeling. Readings range from semiotics and post-structuralism to computer science and language art, language models from Markov chains to GPT.
Department-Specific Course Categories
Quantitative Social Sciences