Ended my 15-lecture Information Theory course for upper-level physics undergraduates by applying it to large language models. Basically: entropy, compression, typicality, and why “Claude” is called “Claude”. (Made in collaboration with the eponymous LLM.)