NotebookLM has become one of my most-used tools this year. I use it for studying, research, and as a note-taking app. I’ve also been using it in unconventional ways, such as pairing it with other apps, creating courses, “watching” YouTube videos, and turning it into a design instructor.
Like most users, a lot of my time is spent in the Studio section of NotebookLM. But I’ve mostly ignored its Video Overview tool because, honestly, I’m not a fan of AI-…
NotebookLM has become one of my most-used tools this year. I use it for studying, research, and as a note-taking app. I’ve also been using it in unconventional ways, such as pairing it with other apps, creating courses, “watching” YouTube videos, and turning it into a design instructor.
Like most users, a lot of my time is spent in the Studio section of NotebookLM. But I’ve mostly ignored its Video Overview tool because, honestly, I’m not a fan of AI-generated videos. Still, curiosity got the best of me. Video editing is one of my favorite pastimes, and I’ve tested just about every editor out there, but I wanted to see what I could create with a tool that isn’t meant for video editing at all.
Why even attempt to use NotebookLM as a video editor?
Testing the limits of an AI research tool
I didn’t have a specific need to use NotebookLM as a video editor; it was mainly curiosity. I’ve been exploring different ways to repurpose the tools I use daily, and it made me wonder if the same AI I rely on for research could help me shape visual stories. NotebookLM was obviously not designed to be a video editor, but it does have a video generator tool.
I’ve had success in using NotebookLM to help me learn new design tools and to help me guide my design process. So I already knew it was capable of assisting in creative work. And when I realized that the Video Overview function had been sitting untouched in my notebooks, I wanted to take it beyond chat prompts and try something more creative. I ended up using NotebookLM the “regular way” for this project too, though, prompting it for editing ideas and advice to clean up the generated video in a real video editor.
What is Video Overview in NotebookLM?
Condensing research into digestible visual overviews
NotebookLM’s Video Overview feature is meant to make large quantities of dense research more digestible. It generates narrated slides from your notebook’s content, summarizing the key talking points and presenting them visually. You don’t have to script or edit anything; it automatically pulls from your sources, generates a voiceover, and pairs it with relevant visuals. It doesn’t generate video footage or animated graphics, but rather illustration slides strung together in a sequence.
Setting up NotebookLM as a “video editor”
It’s all about the sources and instructions you feed it
The first time I tried the Video Overview feature, it gave me a pretty standard infographic-type video on some UX design resources, complete with the voiceover. This was what sparked the idea to try and use the feature more creatively. I already have a more creative notebook - one filled with fictional book files - so I decided to use that for the base of this experiment.
You can’t actually edit these generated videos in NotebookLM, so by “edit”, I mean making the most of its customization options to shape the final output. Selecting the little pencil icon on Video Overview will bring up the customization window. You get a choice between Brief and Explanation - I went with Brief since it creates a shorter output.
Then, and this is what enabled me to get a decent animation, there are different video styles to choose from - I quite like Anime, Paper Craft, Heritage, and Retro Print. Lastly, there’s a text box where you can instruct the AI what to focus on in the video. It’s mainly for directing the content and discussion in the voiceover, but that’s exactly what informs the types of illustrations NotebookLM creates, so definitely give it some instructions for what you want to see. This was one of the prompts I used:
Focus on the themes of fear, folklore, and supernatural belief. Use dark, muted, blue-toned colors.
The results were quite awesome - the illustrations accurately demonstrated the main themes of the story. But it wasn’t perfect. I don’t think NotebookLM registered my color palette request since the generated video only loosely followed it. And I also didn’t like how so much of the graphics were obscured by text, so I asked it to exclude text in my next video, but that didn’t work. My guess is that NotebookLM prioritizes text in these videos for educational purposes.
If you’re going to give this a shot, I’d say focus your customization prompts on the themes you want NotebookLM to discuss in the video rather than the visuals you want to see - the graphics will reflect the themes of discussion anyway. So instead of typing “blue colors”, rather describe a topic that would generate blue visuals, such as the ocean.
Cleaning up my video
With the help of NotebookLM
You can download your videos, just keep in mind it will have a NotebookLM watermark. I imported the video into a browser editor so that I could easily flip back to NotebookLM for some quick editing tips. I used my design notebook for this as it already has some information on video color editing, motion techniques, and so on. You can go all out here and even instruct NotebookLM to be a video editing guide for a specific software, but that wasn’t necessary for a simple video like this.
The illustrations were really the only thing I was interested in, so I removed the audio and cut the parts that had too much text. Then I prompted NotebookLM for some color editing best practices and transition/animation ideas based on the theme of the story and the graphic style, and it pulled some tips from my sources. It was just a matter of making the relevant adjustments in the editor, which took about ten minutes (I just used CapCut Online).
NotebookLM is more creative than we give it credit for
Will NotebookLM replace my video editing apps anytime soon? Absolutely not. But its video tool is a lot more capable than I first realized. It’s basically an AI video generator that you can prompt through the customizations, which lets you “edit” the direction of the video. Furthermore, in proper NotebookLM fashion, you can use it to help inform your editing techniques for when you import the video into a real editor later on. The results aren’t breathtaking by any means, but it’s definitely something worth trying if you want to explore alternative uses of NotebookLM or want new material to edit.