NotebookLM was developed at Google by bestselling science author Steven Johnson and members of Google’s AI team. I’ve long admired Johnson’s work and compelling ways of thinking on a variety of topics, so it’s no surprise that NotebookLM has become one of my favorite AI tools.
It’s also unsurprising that it’s also a favorite of many educators with whom I speak. That’s because NotebookLM does all the things we normally associate with good education-focused AI use. It puts teachers in the driver seat, allowing them to decide which sources the tool draws upon for all the materials it creates.
Even more significantly, NotebookLM is, at its heart, a research assistant, and is fundamentally designed to help users organize and learn from various sources they gather rather than do their thinking for them. In other words, it truly works as a notebook for the AI age.
Here are some ways in which teachers can use NotebookLM.
These days, NotebookLM is a good starting point when you begin researching any topic. It gives you the option of choosing “fast research” or “deep research.” Deep research takes more time but goes more in-depth. Regardless of which option you select, you are then given a series of sources that you can choose to keep or discard.
Below, I talk about some of the cool things NotebookLM helps you do once you’ve gathered this research, but before getting to that, I want to say how helpful this tool is in its own right. Sure, other AI tools can search the internet for sources, but here you not only get the sources, it all is automatically saved and organized within an easy-to-use interface. I find this helpful when doing research for books and articles as well as for research for class prep, but of course, this organization is only the beginning.
Both deep and quick research options work very well, depending on your needs. I’ve found that starting with a quick research option, exploring those sources, and then having NotebookLM do deeper research to unearth more resources, can be very effective.
2. Interact With Research and Educational Materials
Once you get research results or add your own documents and materials, which is also a great option within NotebookLM, you get a whole bunch of AI-enhanced ways to organize and learn from this research.
First, you can interact with a chatbot that uses these sources, and only these sources, as the basis for its responses to you. This can be a fun way to delve into a topic and, literally or figuratively, interrogate a text.
For example, I was reading a paper about a complex physics experiment and didn’t fully understand certain implications. I asked a few questions, and the NotebookLM chatbot read through this paper and other similar ones I had selected and clarified various points.
Of course, even with the guardrails that NotebookLM inherently places on these interactions, it might not be perfect, but it’s still a helpful learning tool as you conduct and think through research. It could also theoretically be a good and fun way for students to explore class materials you assign, though I don’t personally use NotebookLM in that way.
3. Create Class Materials From Specific Documents
You can also then use these documents to do many of the tasks AI is already helping teachers with, such as generating quizzes, flashcards, mind clouds, and even slides. This can be either a tool for organizing your own thoughts and research, or the first step to creating some class materials.
None of the materials I’ve used NotebookLM to create are perfect, but it is generally better than other programs I’ve tried. Since it is based on specific materials I had approved, I also feel much better about using this as a starting point than I would working with a more general AI tool.
4. Create Podcasts and Videos
Creating on-demand podcasts and videos that summarize documents and readings you select is one of the features for which NotebookLM is known. The podcast feature, in particular, drew the lion’s share of press after the tool launched, and from a purely technical standpoint, it’s amazing.
You put in a few documents and, soon after, you can listen to a casual podcast with multiple hosts, breaking down the topic for you in the manner of a modern laidback podcast. The video is somewhat similar but offers visuals and graphics along with narration.
As mind-blowing as both features are, neither is my favorite, as I still don’t trust this kind of summary from AI. That said, this feature is one that a lot of teachers use. So even though it’s not my personal favorite feature, I’m intrigued by it and agree there’s a lot of educational potential for this kind of tool.
I also recommend everyone try creating their own podcast using NotebookLM because, from a pure wow factor perspective, it’s still hard to beat.