NotebookLM

Summary

NotebookLM is a Google AI product that uses your own uploaded documents as sources to answer questions, generate summaries, and create audio overviews — grounded entirely in your content, not general web knowledge.

Sources can be: Markdown files, PDFs, Google Docs, web URLs, or pasted text.
Answers are always cited back to a specific source — it will not hallucinate outside your uploaded content.
Audio Overview generates a ~10 min podcast-style conversation between two AI hosts summarizing your sources.
Briefing Doc and Study Guide outputs give structured summaries and Q&A from your material.
Each Notebook is isolated — sources from one notebook do not bleed into another.
Useful as a synthesis layer over this Obsidian vault: upload a topic folder’s notes to query across them.

Your vault has many notes on a topic and you want to ask cross-cutting questions (e.g. “what are the differences between Claude Code and OpenAI Codex?”).
You want to turn your notes into a study guide or audio briefing before a meeting or review.
You want to verify what you actually know (grounded in your own notes, not AI guessing).

Go to notebooklm.google.com and create a new Notebook.
Add sources — upload Markdown files, PDFs, or paste content directly.
Ask questions in the chat panel; answers include source citations.
Use the output panel to generate: Audio Overview, Briefing Doc, FAQ, or Study Guide.

To bundle a vault topic folder into one file for easy upload:

cat "AI/Claude Code/"*.md > claude-code-notes.md

Upload all AI/Claude Code/ notes → ask “what slash commands manage context?” → cited answer from your own Slash Commands Reference.
Upload all AI notes → generate Audio Overview → 10-min podcast summarizing your knowledge of Claude Code, OpenAI, and Gemini.
Upload meeting notes or research → generate Study Guide → structured Q&A for review.