One of my personal goals this summer was to start reading again. And though I've managed to achieve my summer goal and have read four books in the past week alone, I forgot just how many books there are that I want to read but just can't seem to get into. I get excited about one, read a few pages, and then quietly move it to my do not finish pile.
Sometimes, I wouldn't be able to follow the plot. Other times, I'd stop reading for a few days and forget what was going on entirely. A lot of these books are super-hyped, and most of my friends are raving about them too. So I feel extra guilty for not liking them enough to push through. Ultimately, I figured it was time to try something new.
I decided to see if my favorite AI productivity tool, NotebookLM, could help me with this. I figured, if I wasn't going to sit through the whole book the traditional way, maybe I could get a little help. After all, it's designed to help you interact with your uploaded documents better. I already use it to break down lectures and readings for college, so I figured... why not try uploading a book I've been wanting to read but keep abandoning? And that's exactly what I did.
The setup was fairly simple
It took only a few clicks to get going
If you're wondering "why exactly did you use NotebookLM for this, anyway?" there's only one real reason: NotebookLM's source-grounded answers.
Though there are a ton of video summaries of books you can watch online, they don't always go in-depth and tend to spoil major plot points without giving any real context. Similarly, a lot of AI tools can browse the web for you and tell you everything you need to know about a book without you needing to read a single chapter, but there's no telling where that information is coming from, or whether it's actually accurate.
NotebookLM changes this completely. It builds a personalized AI that's only familiar with the information you feed it via sources or the conversation history. So, creating a NotebookLM notebook for this experiment was extremely simple.
Since I primarily read on my e-Ink tablet nowadays, most of my books are in the epub format. Unfortunately, NotebookLM doesn't accept .epub files as sources, so I converted the file to a PDF and uploaded it. I recently watched an entire show using NotebookLM, and since the tool doesn't accept videos itself as sources, I needed to hunt down episode summaries and recaps to upload to my notebook. So compared to that, uploading a book was almost too easy.
The book I decided to use for this experiment was Legendborn by Tracy Deonn. I've attempted to read it at least three times but just can't seem to get into it. At the same time, the plot itself sounds really compelling. So, I figured it was the perfect book to read using NotebookLM!
So, how did I actually use NotebookLM to "read" the book?
My step-by-step approach
Once the PDF version of the book was uploaded to my notebook and processed by NotebookLM, it was time to actually begin reading. Before diving into NotebookLM's most iconic features like Audio Overviews and Mind Maps, I decided it was best to start with the basics: just asking questions.
So, I started by sending the following in-depth prompt:
I've tried reading this book multiple times but I've never made it past the first few chapters. I'd really like to finally understand it, not just skim through, but actually absorb what it's about.
Think of it like this: I want to walk away feeling like I've read the book, even if I didn't read every single page myself. Can you help me get familiar with everything I need to know? I'm looking for a full overview. Tell me about key characters, plot points, themes, chapter summaries, and any important quotes or turning points. Tell me everything!
It took NotebookLM longer than usual to whip up a response, but that's not entirely surprising. It was helping me read a 400-page-long book, after all. The response to that one prompt alone was probably enough for me. Since NotebookLM is AI-powered and has access to the entire book, it can intelligently make connections, some of which I'd never have picked up on myself.
For instance, the protagonist of the book struggles with PTSD, and NotebookLM pointed out that her grief is actually inspired by the author's own experience, as mentioned in the author's note. I almost never read author notes, so I'd have completely missed that.
The book I was reading was a young adult fantasy novel. And if you read fantasy often, you know how tricky it can be to keep up with all the terminologies, world-building concepts, and more. I almost always get lost in them and end up constantly flipping back to understand what's going on. In NotebookLM's response to the prompt above, it listed everything I needed to know under a Key Concepts and Terminology section.
By the time I was done reading the response, I obviously had questions of my own. So, I asked NotebookLM all of them, and had in-depth answers to each question within seconds. I also asked it to create a chapter-by-chapter breakdown by sending the following prompt:
Give me a detailed chapter-by-chapter summary of this book. For each chapter, include key plot points, character developments, and any important themes or symbols introduced. Also, point out any callbacks to earlier chapters or foreshadowing, if present.
NotebookLM's Audio Overview and Mind Map features also helped
Of course, I had to explore NotebookLM's standout features
Once I was done asking all the questions I had, I felt pretty content with how much I'd gotten out of the book. I felt familiar enough with it to be able to talk about it with my friends, but I still wanted to see how else NotebookLM could help me. So, I generated an Audio Overview of the book, as well as a Mind Map. Since I was already familiar with the book at this point, the Audio Overview didn't really tell me anything I didn't already know, but it was a great recap of the book as a whole.
The Mind Map was... exceptional. It had five nodes: Key Themes, Characters, Magic Systems, Organization and Conflict, and Key Plot Points. It was a great way to visualize the book's structure. I just wish I'd looked at it before asking NotebookLM all the questions in the Chat panel, since it would've been a good way to have a mental roadmap beforehand.
I might just do this again
Now, just to very, very clear: I won't ever use NotebookLM to actually read a book I care about or want to read for fun. But for the books I just can't get into or am only slightly curious about, this might just be my new favorite hack.
