NotebookLM Workflow
NotebookLM is a source-grounded research workspace. Build a notebook around a peer post, add outside sources, and ask questions that get answers with citations.
The Big Idea
Unlike regular chatbots, NotebookLM only answers based on sources you add. When it makes a claim, it cites the specific document and passage. This makes it perfect for:
- Understanding a peer's argument before responding
- Finding credible sources on a topic
- Verifying claims with evidence
- Building a research trail you can cite later
Key concept: Grounding. NotebookLM's answers are "grounded" in your sources — it won't make things up because it only pulls from what you gave it.
Step-by-Step Workflow
Go to notebooklm.google.com and click "New Notebook."
Naming convention: Use a descriptive name so you can find it later.
Peer Response – [Peer Name] – [Topic] – Week X
Example: Peer Response – Gabriel Bell – AI Scarcity – Week 3
Click "Add Source" and choose one of these methods:
- URL: Paste the direct link to their blog post
- Copy/Paste: Select "Copied text" and paste the post content
This becomes your primary source — the argument you're responding to.
Try it: See the NotebookLM Example for a complete walkthrough with a copyable student post URL ready to paste.
If your peer is responding to another post, add that too. Build the conversation trail:
- The original post your peer responded to
- Any sources your peer cited (if you can find them)
This gives NotebookLM the full context of the conversation.
Now ask questions in the chat. NotebookLM will answer with citations to your sources.
Good starter questions:
What are [peer name]'s three main claims? Quote the relevant passages. What sources does this post cite? List them with their claims. What assumptions does this argument depend on? What's the strongest version of this argument? (Steelman)
Notice how the answers include numbered citations pointing to your sources.
This is NotebookLM's research superpower. Click "Discover Sources" and enter a query related to your peer's argument.
Example queries:
- "IMF report 2024 AI job displacement"
- "consumption paradox automation wages"
- "evidence for AI productivity gains"
NotebookLM will find web sources. Review them and add the credible ones (2-3 is usually enough).
Now that you have outside sources, ask questions that test your peer's claims:
Does the IMF report actually say 40% of jobs are exposed? Quote if available. What counterarguments exist to [peer's thesis]? What do my sources agree on? Where do they disagree?
This is where you build your own argument — by testing theirs against evidence.
When NotebookLM gives you a particularly useful answer, you can save it as a Note. Notes can then be converted to Sources.
This builds your research trail — a record of your thinking that you can cite or refer back to.
- Click the "Save as Note" button on useful responses
- Edit notes to add your own thinking
- Convert key notes to sources if you want them searchable
You can draft directly in NotebookLM or export your research to Google Docs.
To draft in NotebookLM:
Based on my sources, help me draft an 800-word response that: 1. Summarizes [peer]'s argument fairly 2. Presents my counterargument: [your thesis] 3. Uses evidence from my sources 4. Marks any claims that need additional citations
To export: Copy your notes and key quotes, then draft in your preferred writing tool.
Tips for Better Results
Quality Sources Matter
NotebookLM is only as good as the sources you add. Prioritize:
- The peer post you're responding to (always first)
- Sources your peer cited (to verify their claims)
- Credible outside sources (academic, reputable news, official reports)
Be Specific in Questions
Vague questions get vague answers. Instead of "What does Gabriel think about AI?", ask "What specific evidence does Gabriel cite for his claim that AI will dismantle wages?"
Use Audio Overview for Orientation
NotebookLM can generate a podcast-style summary of your sources. This is great for getting oriented — but don't cite it directly. It's a summary, not a source.
Important: Always verify quotes and statistics in the original source before using them in your post. NotebookLM is grounded, but you still own the accuracy.
See It In Action
For a complete walkthrough using a real student post, see the NotebookLM Example page.