LLMs can feel almost magical. They answer questions, write code, summarize research, and even help build games. But underneath it all is a system you can understand.

This course takes you inside large language models without burying you in unnecessary complexity. You’ll start with tokens and the idea that an LLM is, at its core, an incredibly sophisticated prediction machine.

From there, you’ll explore transformers, GPT architecture, and the training process that turns raw models into useful AI assistants.

You’ll also investigate the stranger side of AI. Why can models solve difficult problems but stumble over seemingly simple ones? Concepts like the Reversal Curse help reveal where current systems still fall short.

Along the way, visual exercises and practical challenges help turn abstract ideas into something concrete. By the end, you’ll have a clearer mental model of how modern generative AI actually works—and how to think about it more critically.