Managing Sources

Lecture 3: Manging Sources

Lecture 03: Managing Sources

Lectures 1 and 2 introduced the context window and interactive prompting. Lecture 3 focuses on context, information, and prompt instructions that persist beyond a single chat session. This includes everything else sharing your context window — settings, remembered facts, uploaded files, and retrieved web pages - and introduces retrieval-augmented generation (RAG) as a solution for grounding context that would overflow the context window. Examples include how to use these tools to reduce sycophancy and “AI style,” understanding what your chat history is (and isn’t) used for, saving your best prompts as reusable Skills or Gems, and strategic withholding–how sometimes the smartest prompt leaves something out. We introduce Retrieval-Augmented Generation (RAG) for grounding AI in your own documents, and the documented ways trust in a RAG store can fail. We close on a cybersecurity note with prompt injection: why an AI can’t always tell your instructions apart from malicious text on the internet.