Answer: Best ways to do just-in-time learning. Part 3: getting a teaching summary with NotebookLM

 Last time... 

A woman studying a complex anatomy topic. P/C Pexels.com by RF._.studio _

      

we talked about asking your favorite LLM to give you a summary (say 500 words on a specified topic).     

But maybe there's a way to be even MORE specific.  Remember our Challenge:  

1. What are the best AI-inspired (or AI-driven) new Micro- and Meso-learning that you've found?  Have you actually tried this method?  How well has it worked out for you? 

Asking for an essay, especially one that you can have read aloud, is pretty handy.  

The downside is that you're getting the full LLM, and that may (or may not) be what you're looking for.  

Using Google's NotebookLM lets you start with a specific set of resources--presumably ones that you've hand picked to be reliable, accurate, and on-topic--and then create a summary in text, audio, or video formats!  

Here's what I did.  

1. Start a new Notebook.  Visit NotebookLM and create new notebook.  Then add sources by clicking "+ Add" in the upper left corner.  As you can see, I've added 5 papers I found by using regular Google Scholar 

The papers are summarized in the main Chat window (in the center).  In short, they're: 

1. Selgrath et al. (2024) reconstruct kelp forest changes over centuries in central California

2. Leichter et al. (2023) examine the impact of nutrient availability

3. Garcı́a-Reyes et al. (2022) demonstrate the predictive power of winter oceanographic conditions on summer bull kelp canopy cover in northern California

4. Butler et al. (2021) document the soundscapes within and around a marine protected area off La Jolla 

 

(A minor hassle here in the design of NotebookLM.  The references in the Chat window are all correct, but it's hard to know which of the links in the Sources window map onto the refs in the Chat.  You have to click on the source to see which is which, and it's not an obvious mapping. It also didn't tell me that source 2 and source 4 are duplicates, which is why there are only 4 sources listed in the summary. Argh!) 

In any case, if you click on the "Video Overview" in the upper right corner, NotebookLM will create an 8 minute video summary of the contents of the sources.  Link to the YouTube video.  

You'll see the effect in  a minute or so... 

Let me repeat--the summary (while great to use for learning)--is NOT a replacement for reading the papers.  There's all kinds of detail that's not covered in the video overview.  That's why it's an overview, and not the papers themselves. 

As we mentioned earlier, you can also generate an audio overview.  Here's the M4a file of the summary of these papers. (14 minutes)  This is handy for putting onto your phone and listening as you walk.  (One minor annoyance--Google speech doesn't know how to pronounce the city name "La Jolla."  It's NOT "la-jala," it's "la-hoya.")  

ALSO note that you can customize both the video and audio overviews.  Click on the 3-dots in the upper right and you'll get a set of customization options: 

Obviously, you can get the overview in another language (handy, if you're studying that language) OR you can get it focused on a particular learning task, such as preparing for a quiz on the topic or a specific sub-topic within the collection of papers.  

I have to admit, even though I know a fair bit about this topic, I learned something by listening to the audio summary.  NotebookLM really does a good job of synthesizing something coherent and useful from the sources. 

 Last time we also mentioned using the Feynman technique for learning.  Using the prompt: 

[tell me about the Feynman technique for effective learning] 

.. this will cause Gemini to jump into the "Feynman teaching mode" on the topic of California kelp.  

If you give NotebookLM this prompt, it will give you a summary of the material, which you can then use to guide your own Feynman method. 

How use the Feynman Technique with NotebookLM

After you've uploaded your sources (documents, articles, etc). You can select a topic you'd like to understand in more detail.  

Prompt NotebookLM to explain: Ask the AI to explain the concept as if to a child or a beginner to force simplification and the creation of analogies. For example, you can use a prompt like, "Explain the relationship between purple urchins and the loss of bull kelp in California using the Feynman technique, as if teaching it to someone with no background in the topic."  (Don't worry about marking off the concept from the query--NotebookLM will figure out what you mean.)  

Evaluate the explanation: Review the AI's explanation for clarity and accuracy. Does it all make sense to you?  If not, you need to... 

Identify gaps in your understanding: If the explanation is still unclear or if it highlights weaknesses in your own understanding, make a note of these areas.  Once you've identified these issues... 

Return to your sources: Go back to your original documents within NotebookLM to find the information needed to fill those knowledge gaps. 

Refine your explanation: Go back to NotebookLM and re-explain the concept, incorporating the new information you've and simplifying the language further. 

Repeat the cycle: Continue this process of explaining, identifying gaps, researching, and refining your explanation until you fully grasp the concept. 

Make sense?  That's the value of having your own tutor on the topic you choose.  

Of course, your next step would be to find additional resources to extend your learning.  You could ask NotebookLM to: 

     [ recommend other papers to read on this topic] 

Or, you can use the Discover button (upper right of the Sources panel) to add additional sources.  

 

SearchResearch Lessons

This is Part 3 of our "how to learn rapidly" series.  We learned that: 

1. Using NotebookLM can give you control over the content you're trying to learn. Add the materials you need to learn to the Notebook, then ask for a summary, or ask questions of it as you wish.  

2. You can create audio or video summaries of the Notebook sources. This is remarkably handy as a way to learn-on-the-go.

The NotebookLM method promises to give you much higher quality summaries since it's basically answering questions (or generating your summaries) from the materials that you've selected. (Which you presumably selected for their accuracy!)  

There are many more AI-augmented learning methods out there.  Keep your eyes open and let us know if you find any that are particularly effective for you.  Keep those comments coming! 

Keep searching!