Google’s Hidden Powerhouse - Notebook LM
Chat with your own documents with Google Notebook LM
Google’s Notebook LM hasn’t received as much attention as some of the other generative AI tools, but in my experience, it’s a real powerhouse in the right situation. You may have uploaded documents to other AI tools like Claude or ChatGPT, which allowed you to ask questions related to those documents. Notebook LM is like that on steroids.
Here’s how I’m using Notebook LM. My doctoral seminar involves weekly readings of articles on various topics. For example, one week we read about different views of what theory is in the social sciences. Generally, students read five or six articles and we discuss those during the seminar. Students also have to write a synthesis of the articles each week.
This is the first seminar in our doctoral program, so this is all very new for the students. There are a lot of terms (or terms used in new ways), concepts, and ways of thinking that are completely new to the students. As a result, they often get overwhelmed and confused. The newness of it all sometimes makes it hard for the students to understand the essential takeaways from the set of readings. Of course, I explain my thoughts on what’s essential, but it’s still quite a bit to process.
Since I’ve been doing this for a LONG time, sometimes it’s hard for me to step out of my own perspective and look at the articles from the students’ points of view. This can cause its own set of problems.
This is where Notebook LM has been invaluable as a tool for helping me better distill the readings in helpful ways. Let me take you through my process.
The first step is to create a new notebook. I decided to create a notebook for each week to better focus the responses on that week’s topic. Also, I’m not sure if Notebook LM could handle the entire set of seminar readings. Once the notebook is created, I upload the readings into the notebook. This is all very straightforward.
Once this is done, I’m ready to chat. Keep in mind that the chat is now going to use the articles I uploaded as its key information source. In other words, responses are going to be based on the articles.
Each week, I use three main prompts: 1) key takeaways, 2) potential areas of confusion, and 3) a synthesis. It took a few tries to get the right prompts, but now I have them saved in an Apple Note and just copy and paste them into each week’s notebook.
The image below shows the notebook for Week 2 of the seminar.
Let’s zoom in on the Areas of potential confusion note. The image below is scrolled down to the end of the note because this is where the big payoff is. Most of the rest of the note were confusions I already anticipated. It’s nice to know that AI agrees, but what I really want is for Notebook LM to uncover some additional possibilities. Notebook LM brought up two potential points of confusion I hadn’t considered: the iterative nature of theory building and the balance between theory and method. Discussing these with the students may help avoid some necessary confusion and angst.
Here’s the thing. The entire process takes about ten or fifteen minutes. That’s it. Creating the notebook, uploading the documents, and running the prompts only takes five or six minutes. Then it’s another ten or so to scan through the results. The return on that investment in terms of the quality of the seminar is massive.
The ability to save the responses as notes is something that sets Notebook LM apart from similar tools, such as Claude’s Projects. Another unique feature of Notebook LM is its interesting built-in prompts, which automatically save as notes. Clicking on Notebook guide in the lower right accesses these. These prompts create:
FAQ
Study guide
Table of contents
Timeline, and
Briefing doc
I don’t always use these, but sometimes they’re useful.
Maybe the most fascinating and unique feature of Notebook LM is the audio overview. It’s available through the Notebook guide. The audio overview is scary. It’s a natural sounding conversation between two “people” (AI simulations) discussing a synthesis of the papers. It is good to the point of being scary.
You can check out the conversation here: https://open.substack.com/pub/aigoestocollege/p/google-notebook-lm-talks-about-theory
I’m still trying to decide how much of this to share with my students, although I’ll disclose my use of Notebook LM next week. So far, Notebook LM’s creations have been high quality, but I’m still a little hesitant to share what I’ve created with my students. My concern is based on the possibility that Notebook LM may short-circuit some of the struggling that’s critical to the learning process. That being said, I’ve decided to share the existence of Notebook LM with them during the next seminar meeting. (They may already know about it though.)
Notebook LM may be useful for almost any situation in which you have to deal with a set of documents such as course materials, projects, grant writing, career development and the like. As always, be careful about uploading any private or sensitive information.
Google hasn’t made Notebook LM available to everyone yet, but the last person I recommended it to was able to get access right away. Check it out for yourself at https://notebooklm.google.com/.
If you have any questions or comments, you can leave them below or email me - craig@AIGoesToCollege.com. I’d love to hear from you. Be sure to check out the AI Goes to College podcast, which I co-host with Dr. Robert E. Crossler. It’s available at https://www.aigoestocollege.com/follow. Thanks for reading!