NotebookLM is one of the few AI tools I'll never get tired of talking about. It impressed me early on when it was just a Google Labs experiment, and it hasn't wavered since. In fact, it's only gotten better, and after sitting down with the engineer behind the tool's best feature to talk about his setup, I walked away even more convinced that Google has something genuinely special going on with NotebookLM.

The reason why I mentioned all of this is because I need you to understand that what I'm about to say doesn't come from someone who's been looking for an excuse to ditch NotebookLM. I'm still a loyal user, but the way I've used the tool has changed a fair bit over the months. One of my biggest NotebookLM use cases was using it as a second brain, and the tool was excellent at it. However, after spending the last few months building out Claude Projects as my primary knowledge system, I've come to the conclusion that Claude Projects is more than a viable alternative to NotebookLM as a second brain.

Claude Projects and NotebookLM are fundamentally different tools

Same inputs, very different outputs

Now, something I'd like to be incredibly clear about is the fact that NotebookLM isn't a Claude Projects alternative, and Claude Projects isn't a NotebookLM alternative. They're both completely different tools. Claude is a conversational AI chatbot that users use to do everything from writing code to drafting emails to brainstorming ideas. The models powering Claude are trained on vast amounts of internet data, and while they're incredibly capable, they can and do hallucinate. NotebookLM takes the completely opposite approach. It's built entirely around source grounding and doesn't pull from general knowledge. It reads the information you upload and then refuses to go beyond it.

While that might sound like a limitation, that's precisely what makes it such a powerful tool. Anthropic added a Projects feature to Claude that lets you create dedicated workspaces where you can upload files and set custom instructions. When you're having a regular conversation in a Claude thread, it doesn't really know much about the specific project you're working on. It might remember bits and pieces from past chats if you have memory turned on, but it doesn't have deep, structured context about your research, your frameworks, or the fifty pages of notes you've been building up over the last few months. Projects were built with exactly this in mind. They give Claude a dedicated workspace where your documents are always loaded, your instructions are always active, and every new conversation picks up with the full picture rather than fragments of it.

Now, of course, this raises the question of how it is any different from NotebookLM. You upload documents and you chat with them. Isn't that the same thing? Well, on the surface, sure it is. But the difference becomes obvious the moment you actually start using both for the same purpose. NotebookLM keeps the conversation strictly tethered to your sources. That's the tool's entire purpose, and it's a double-edged sword at times. Claude Projects doesn't make that promise. Instead, it takes your uploaded knowledge and combines it with everything else Claude already knows to help you think further than your sources alone would take you. It connects your notes to broader ideas, challenges your assumptions, drafts new material based on your frameworks, and treats your uploads as a starting point rather than a boundary.

The fact that it doesn't stop at your sources is its biggest advantage

A second brain that thinks, not just remembers

The reason why I've been such a huge advocate of NotebookLM is its source-grounding. But also, when you're using a tool as a second brain, you don't necessarily want it to only tell you what you already know. You want it to take what you know and push your ideas further, like your actual brain would. You want it to connect your current ideas to things you haven't read yet, challenge them with perspectives you hadn't considered, and fill in gaps you didn't even know were there. You can't really achieve that with NotebookLM since it stays within the boundaries of what you've uploaded. Ask it something that goes beyond your sources, and it'll tell you it doesn't have enough information to answer. So with NotebookLM, your second brain can only ever be as smart as the material you've fed it.

On the other hand, Claude Projects lets you engage with your own sources just as deeply, but it doesn't hit a wall when the conversation naturally extends beyond what's in your documents. It keeps going and draws on its broader training (and external information) to supplement your thinking. The reason why Claude specifically works great for this use case is because, while the Projects feature pairs your knowledge with what's generally available, it doesn't take your previous conversations outside that project into account. So your second brain stays self-contained, and feels like an actual dedicated knowledge system rather than just another generic chat window with some files attached.

And unlike NotebookLM, which just retrieves information and cites it, Claude Projects goes a step further and reasons across your materials. It drafts new ideas (though as a writer, I won't support that), challenges your thinking, and helps connect dots between documents you normally wouldn't have found yourself. Given that a second brain isn't supposed to be just storage and retrieval (which is all that NotebookLM can offer you), Claude Projects is simply better suited for the job. Not because it was built for it, but because the way it handles context, reasoning, and knowledge turns out to be exactly what a second brain needs.

Claude Projects conversational continuity is something NotebookLM can't match

Pick up where you left off

In my eyes, something a second brain absolutely needs to have is the ability to pick up where you left off without re-explaining everything from scratch. I've found that Claude Projects is excellent at this, and it's the biggest area where it beats NotebookLM. You can have dozens of conversations within a project you've created, and every single one of them shares the same context. This means you don't really lose momentum between sessions and don't need to spend a significant chunk of time explaining the same context again and again.

NotebookLM doesn't really work this way. Each chat is more or less its own thing. You can query your sources, but the tool doesn't accumulate a sense of what you've been working toward across multiple conversations. While the Chat History feature NotebookLM has makes it a bit easier to look back at what you've discovered before, it's not the same thing in the long run. While NotebookLM can use that as context, I've found that it's nowhere near as seamless as how Claude Projects handles it. You're constantly doing a small amount of re-orientation, and over weeks of sustained thinking on a topic, that friction adds up more than you'd expect.

The rest of Claude's ecosystem is what seals the deal

A second brain that actually has hands

Everything I wrote above should be enough to convince you that the Projects feature itself has impressed me enough to make the switch. What seals the deal is the fact that Claude Projects doesn't exist in isolation. It sits inside a much broader ecosystem that makes using it as a second brain significantly more powerful. While Google's been working tirelessly on making NotebookLM work better with other Google tools, I'd say Claude's ecosystem is already there in some ways.

You can connect your second brain to your actual workflow, which makes it a lot simpler to go from thinking about something to actually doing something about it. You can pull in emails from Gmail, add something to your Google Calendar, reference a Notion page, or even pull your Slack conversations without needing to leave your project. Your second brain goes from being a place where your ideas sit to connecting to the tools where all your work happens. And then there's the output side of things. Claude can generate interactive visuals, create documents, build spreadsheets, put together slide decks, and more within the same conversation where you were just brainstorming.

While that extends to NotebookLM too, with its Slide Decks feature and the other Studio outputs, Claude's aesthetic is a bit more polished. The interactive visuals alone are something I haven't seen any other AI tool match, and when you're using it as a second brain, being able to instantly turn a messy train of thought into a clean diagram really helps.

For building a second brain, Claude Projects just works better

While I had my doubts about the Claude Projects feature, I've found myself using it more and more. It's become increasingly clear that it makes for a better second brain than NotebookLM.