Though both the NotebookLM and Gemini trains had long left the station before I jumped on either of them, I got there eventually. And then Google put them on the same track through a new feature called Notebooks. At first, I couldn't help but notice similarities to Claude's Projects because of their persistent context. But after using them for a few weeks now, it became clear where they actually slot into my workflow - connecting the learning to the practicing.

Google solved a problem I didn't know I had

The middleman in my workflow was me

The problem was that I was exporting my own notes to feed them back to myself. I've been using both tools happily for about a year now, and never really questioned hopping between them. That's what I already do with most of my other workflows anyway - from a note-taker to a chatbot, to a design tool, back to the notes tool.

NotebookLM was where I'd upload course material, generate study guides, build out a knowledge base, make interactive mind maps, and basically just create a learning environment for myself. Gemini was where I'd actually do something with what I'd learned (and honestly in the back of my mind, I still see it as a Google on steroids that can fill in any gap in your knowledge). So, two separate tabs.

Moving content between them was manageable enough that I never flagged it as a problem. I'd save NotebookLM responses as notes, export to a Google Doc, and fetch that Doc from Drive inside Gemini. Same thing with Gemini chats: export to Docs, fetch from within NotebookLM. I use Google Drive almost every day to get my notes across devices and software already, so it wasn't a big deal. Plus, prior to this, Gemini already had a NotebookLM integration…sort of. You could attach a NotebookLM notebook to a Gemini chat and have your sources available - but it was a one-way link, nothing synced back, and anything you worked through in Gemini stayed in Gemini.

Then Notebooks happened

The bidirectional sync is everything

I still hop between them, but Notebooks removed the necessity for manually shifting content between them. The Notebook is the shared workspace now, and that's the part that actually changes the workflow.

Say I create a notebook in NotebookLM on the theory behind onboarding screen design. That notebook now automatically shows up in the Notebooks panel in Gemini. Moreover, it contains all of the sources and custom instructions from NotebookLM. The only thing not in there are the NotebookLM chats and content generated through Studio. Any Notebook created in Gemini also populates automatically in NotebookLM, and every conversation I had within the Notebook in Gemini, appears as a source in NotebookLM. The sync is now a two-way street, and a step up from the previous NotebookLM integration.

For my use case in particular, NotebookLM is my top spot for learning about design. I add a bunch of my coursework in there to help me memorize concepts and principles, and so far, NotebookLM has been solid for helping me retain information, especially through extras like the Quizzes and Flashcards features. But there's a secondary element to what I'm studying, and that's practice. This is where Canvas comes in…

Canvas is one of my favorite features in Gemini…actually, in any Google product. It's a separate workspace that adapts to whatever you ask Gemini to create. Need an editable text doc? It turns into a document editor. Need a slide deck? It morphs into a presentation space. But my favorite is rendered code - it turns into a vibe-coding space of sorts that gives you interactive prototypes of whatever you want to build.

So how does Canvas tie into Notebooks for me? Well, now I get to create a Canvas inside my notebooks. Take visual hierarchy in design for example, something I'm currently learning about. I started in NotebookLM, added my links and course docs, and had some chats to cover the basics. Then I opened the notebook in Gemini and it already had all the relevant materials and instructions for helping me practice visual hierarchy. This is thanks to Canvas, which lets me build UI screens based on my own materials, plus Gemini's outside context which just further refines it.

The loop closes back

The next session starts smarter than the last one

As I've mentioned, when you finish a Gemini session inside a notebook, that chat syncs back to NotebookLM as a source automatically. So my design reasoning, the questions I asked Gemini, the structured outputs it gave (usually through Guided Learning, another one of my favorites in Gemini) all of it feeds back into the NotebookLM notebook as their own, toggleable sources. Next time I open that notebook in NotebookLM, it's basically way smarter than it was before because it now has Gemini's reasoning, training data, and web search content.

It's worth mentioning that visuals generated in a chat in a Gemini Notebook will not show up in the source in NotebookLM, despite NotebookLM now accepting image inputs. When a Gemini chat is added to your NotebookLM source panel, it essentially creates a text transcript of the conversation. Visuals generated in Gemini are hosted on a separate generation server, so they're not part of the text file that shows up in the notebook sources. Unfortunately, this does mean that I'd need to save my images and stills of my prototypes to Google Drive, then import them into NotebookLM separately.

Two tools, one brain

I'm still hopping between the two apps - that part hasn't changed. But there's a difference between tool-hopping by necessity and tool-hopping because each one does something the other genuinely can't. Notebooks made it the latter.