Transcription
Have you ever wished there was a way to turn an endless social media dune scroll into a learning session? Found an interesting article? If only there were a way to create a podcast and listen to it on your commute home. Upload a bunch of research papers and generate a video presentation while you are making dinner.
There is a way. Here's what I'll cover in this video. Five major updates that shift Notebook LM from a notetaking and productivity tool to a complete learning and research platform. I'll show you real demos of every single new feature using my actual real projects, no theoretical examples. Are you ready? Let's dive in.
You've probably seen Notebook LM's audio overviews feature, which have gone massively viral over the last year or so. But sometimes that's not enough. Sometimes you need to not only hear but see the data, the diagrams, the quotes, the images.
Here we are in the Notebook LM interface. I'm going to go ahead and create a new notebook where I can now add a new batch of sources. So, I'll go ahead and click add, and I'll open up this folder that I have here on the side, which is all the files and folders related to one of my biggest current projects, which is a book that I'm writing, which will come out next year on the practice of an annual life review. So, as I've been doing research for my book, I've been finding different academic papers and research studies. But here's the thing, I'm only about two months away from my manuscript deadline. I don't really have time to comb meticulously through new research. I feel this pull like there could be something interesting, an anecdote, a story, a piece of data that could support the arguments I'm making in my book, but I just do not have time to read them in the traditional way.
So, what I'm going to do is select all and drag these five PDFs, all of which are academic papers, into the sources window. So the first thing I like to do is just to click on each individual source because it will generate this little summary of just this one source. So now I have just kind of a rough sense of the diversity of topics that these five sources cover.
Moving on from the sources panel here on the left, I'll go ahead and close that so we have more space. I could go straight to the chat panel here in the middle and start asking it questions. But as I think about that, I don't even know which questions to ask. I need a little more context on what I'm working with here. What is the what does the landscape of information look like that's represented by these five sources?
So, let's actually go all the way over here to the studio panel. And you'll notice that there's a new option right here called video overview. So, what I'm going to do is click the little edit icon and give it a little bit of guidance. I'll do this by holding down a key on my keyboard and then just speaking because I use a text transcription tool that's also AI powered called Whisper Flow. So that's what you'll see here. I'm writing a book on annual life reviews including why they're so important, how to do them successfully, what they have to do with achievement and productivity. Can you please give me an overview and summary of these five sources in video form so that I can decide whether or not they have any relevance to the book I'm writing and whether I might be able to include them as sources. There you go. Then I'm going to hit generate.
That took about 10 or 15 minutes, but now I can click on this video right here and let's see what I came up with.
>> Welcome to the explainer. Today we're going to talk about building something really powerful. A framework for an annual review that actually sticks. One that can genuinely change your life. All right, let's kick things off with a quote that yeah, it might feel a little too real for a lot of us. It's from a 2015 study on professional identity. And man, does it nail that feeling that's just so common in work culture today. So, does that sound familiar? If you're nodding your head right now, feeling that constant tug-of-war with work, then you are absolutely in the right place. That tension, that's our starting line for figuring out how to build a more intentional life. Okay, so let's dig into this feeling because it turns out it's not just you. It's a real thing. And that right there is the big idea. An annual life review is an act of authorship. You are literally taking the time to read the chapter that just ended to really understand it and then to consciously plot out the next one. So what we've built here is a complete architecture for a more intentional life. It's a structure for reflection that leads directly to action, which leaves just one question for you. What story are you going to write? And what's the very first if-then plan you're going to create to start building it? >>
You can find a link to the full presentation in the description below if you want to check it out. What we just witnessed is honestly one of the most mind-blowing things I've seen. I can say as someone who spent almost the last 2 years immersed in the topic, uh, that this little video overview was about, that that was excellent. Think about what just happened. I just threw five pretty random papers. It not only summarized those sources, but it linked them together in a way that would have taken any human many, many hours at least. This is such a powerful medium for learning because visuals plus audio plus narrative is really how humans have shared knowledge for millennia. It's such a more natural way for us to absorb information than reading a cold, stale document.
But if you think about it, most learning doesn't happen like this, sitting quietly at a desk for hours at a time with zero distractions. It happens when you stumble across that interesting article on your lunch break or you're listening to a podcast while walking your dog. As long as Notebook LM was limited to desktop only, it really meant missing most of these what I call microlearning opportunities. But not anymore. They now have a mobile app.
All right. So, let's say that I'm taking a little break during my workday. Here's a kind of infographic that caught my eye about seven proven models for change management. That's something I'm really interested in these days. So, I'll go ahead and tap it and then I'll hit the share button down here at the bottom right and then share via. And you can see I now have the Notebook LM app as one of my share destinations. So, I'll go ahead and tap that. So, you'll see the first thing it asks me is which notebook I want to add this to. So I could easily add it to one of my existing notebooks. But let's go ahead and create a new one with the button down at the bottom. You can see down here at the very bottom of the screen, it's the same three panels that you see in the desktop version, which is sources, chat, and studio.
So I'll go ahead and move over to the chat panel, and just ask using the audio transcription tool built into my smartphone, what is the single most effective method for change management recommended by this source? and how could I use it to help organizations adopt AI and then hit send. So let's see what it came up with.
Interesting. So the response is, the source does not recommend a single most effective method for change management. It depends on the suitability for the specific type and size of the change initiative. That actually makes a lot of sense. So I'm glad that it kind of pushed back on me. So let's go down a little bit and see what it recommends for my specific situation of helping organizations adopt AI. So here we see for small behavioral changes related to AI adoption, encouraging employees to use a new AI powered tool for a specific task, nudge theory could be effective. So I could ask more about that. For individual performance changes, the ADKAR model might be appropriate. So you can see it's it's given me not just a single recommended approach, but kind of a buffet of different options that I could try. This is really taking what was originally a static piece of content that was completely not customized to me that someone posted on LinkedIn and bringing it into my context, customizing it and personalizing it for what I am trying to accomplish, which just saves me a tremendous amount of cognitive effort, which is effort that I can then put into actually implementing this advice. So, this is fascinating. This is already giving me some really concrete examples, some practical takeaways that I could take into my consulting with companies.
Let's try one more thing, which is to head over to studio and go ahead and generate an audio overview. You can see right down here it says this may take a few minutes, but there's no need to stick around. So, I could put my smartphone down. I could even exit the Notebook LM app. I could lock my phone and the system would keep generating in the background, which is nice.
That took about 4 minutes and we can now listen to it.
>> Okay, let's unpack this. Imagine pouring I mean countless hours, huge energy, significant resources into what you really believe is a brilliant idea for change. >> Yeah. Something transformative. >> Exactly. >> So imagine how you could use this. You generated this over five or 10 minutes on your lunch break during work. You could then on your commute back home that same day spend the 10 or 11 minutes it takes to listen to that podcast. You've now successfully linked together and integrated what were previously these disconnected little bits of time throughout your day into a cohesive learning journey even though you don't have hours to spend in your day. I hope you can see through this that this boundary between work time and learning time is just completely dissolving. This app is really bringing the power of AI-driven learning and AI-driven research to all those little microlearning opportunities that you have throughout the day and throughout the week where a desktop or laptop computer just doesn't fit.
But this also raises a bigger question. Why should it be so difficult to find high-quality reliable sources, especially when experts have already organized that research for you? The Notebook LM team last year added the ability to make a notebook public. What you're seeing here is a featured notebook. This is a publicly available notebook that was put together by best-selling author Eric Toppel, who is a professor at Scripps and has written a lot and researched about longevity. How do you live long and how do you live well?
So, think about how valuable this is. How long would it take you to go individually and find and vet and add each one of these sources? Instead, an existing expert, in this case, Toppel has done it all for you. So, you can see here it automatically creates a summary of the collection of sources here in the middle just to kind of orient you around what you're looking at. But I have a specific question, which is I have quite a bit of cancer history in my family in both my parents and my grandparents. What is Eric Toppel's best advice for how I can live my life now in my 40s so that I can avoid cancer for the rest of my life?
He underscores that no intervention can offer a 100% guarantee against cancer, but a comprehensive lifestyle approach is his best advice. So here you can see his advice around diet, avoiding ultra-processed foods, limiting sugar, prioritizing good carbohydrates and fiber. I'm already realizing that this is comprehensive but also a little bit overwhelming. So I think I'm going to now draw on Notebook LM's broad knowledge of even other topics beyond this notebook and ask something very specific and practical. I live in Mexico with my family and we have someone who cooks lunch every day for us in the afternoon. Could you please give me a recipe list and a meal plan for a full week of her cooking that takes into account all the advice in this chat?
Okay, it's a set of guidelines. So, it's summarizing the guidelines up here, but then it has done what I specifically asked, which is what do we actually eat on Monday, on Tuesday, on Wednesday. So, you can see baked fish with roasted root vegetables, black bean and quinoa bowl, chicken, and colorful stir fry. I mean, this looks honestly pretty delicious. I think I'll even go one step further and say, "Please get all of the meals described in this meal plan. Identify every single ingredient that I will need to prepare them, and then put them into a grocery list that is organized according to the sections of a typical grocery store." And there we go. We have produce section, all the vegetables we'll need, fruits, meat, and seafood counter, dairy, pantry.
What these public and featured notebooks really mean is that once someone has put together a thoughtful, thorough collection of sources on any topic anywhere in the world, instead of having to sort of passively listen to a podcast or watch a YouTube video, you can go in like this and interact with their entire body of work in one central place. These sharing capabilities mean that we can all now kind of reallocate our time and effort from knowledge gathering to knowledge synthesizing.
Speaking of which, I've actually taken advantage of the new public sharing options to create an official Building a Second Brain notebook. It includes all my books I've written, all my articles, all my podcasts, all my YouTube videos, including ones like this one, and even more that hasn't been public, which it's taken me over a decade to research and produce. But there's still a bottleneck. Even with expert curated notebooks like this one, what happens when you need sources for your specific project? The hardest part of research, or at least the most consuming part, has always been the hunt, finding credible, high-quality resources in the first place. Notebook LM now has a way for you to discover new sources right within the interface. Let me show you what that looks like.
So, I'll go ahead and create a new notebook. And then you'll see there's a new button right up here in the add sources little dialogue window which says discover sources. So, notice what's happening here. Instead of having to open up a dozen or two random tabs, instead of having to do searches on Google or on YouTube or look on social media or wherever you find your sources, you can just do it right here. So for my book on annual reviews, I recently heard about a concept called infradian rhythms. So I'd like to know more about that. So I'll go ahead and ask it. Find high-quality research-backed sources about infradian rhythms, including multi-day, weekly, monthly, and yearly time scales. Try to find anything that might be relevant and useful for a book that I'm writing on annual life reviews.
All right, so you can see that just took about 20 seconds and it's curated around 10 different sources. I think it usually defaults to 10. So, you can either just go ahead and leave all the boxes checked. You could also uncheck certain ones if you don't think they're relevant. And to know whether they're relevant, you can just click this button right here, and it will open the original source in a new window. These all look pretty valid to me. You can see it's a mix of web pages, blog posts, as well as YouTube videos, which is nice. I'll go ahead and hit import.
All right, that took about 30 seconds. And now that these are loaded into the main Notebook LM interface, I can click on each one as I usually do and just get a little summary of what these are about. So just based on that, this looks like a really excellent selection of sources. This summary here in the middle is also a summary of all 10 sources. So if I didn't want to go in individually, I could just get a, you know, an overview of the entire group of them all at once. But I think the first thing I'm going to do since I'm I'm still trying to just wrap my head around this collection of sources that it's discovered is to click each one of the studio types. Audio overview, let's do a video overview. Let's do a mind map. And let's do a briefing doc. So, at this point, I would probably step away, go have lunch, go for a walk, maybe do some exercise, and just let my AI research assistant go and create each of these different types of outputs.
All right, so the briefing doc was ready the soonest, which is usually the case. This sounds really fascinating. So, I actually now want to see if AI can help me understand how all of this research, which is really quite extensive, might be directly applicable to the book I'm writing. So, I think what I'll do is convert it to source by clicking this button right here. And you can see it was added right there. And then I think what I'll do is actually deselect all the other sources and only select that one. And then I can ask it, can you tell me ideas for how this research and these findings might be relevant and useful in a book I'm writing on annual life reviews? And there we go. So, it's given me several ideas. Uh, integrating multi-year life cycles. That is something I had not considered that I could definitely incorporate. So this is excellent.
And in fact, I even want to go beyond this now. So what I'm going to do to kind of do even one more of those passes, one more of those loops, is let's go ahead and remove all these research sources because we've sort of summarized and distilled them pretty well. Then what I'm going to do is do another round of discovering sources. This time though, I'll search Google Drive and I think make the scope of what I'm looking for a little broader. Search my Google Drive and find any potential source that's related to our relationship to time or the impact of different rhythms or patterns in time on human psychology or motivation or cognition.
So, this is interesting. It's finding um the outline of the current book that I'm writing. So, I think I'll actually unselect that. But I'm seeing some of my Readwise highlights from books I've read, some highlights and excerpts from the book The Body Keeps the Score, which I hadn't considered relevant to this. Some blog posts I've written, which could have some of my personal story anecdotes and writing samples that I've done in the past. So, that seems interesting. All right, that took a few minutes because a couple of these sources are hundreds of thousands of words long. But now what I'm going to do since I have another batch of sources is say draft an outline for a chapter in my book that is all about the science and research of humans relationship to time and how rhythms at different scales including hourly, daily, weekly, monthly, yearly have an impact on how we think, feel, and perform.
And there we go. The rhythms of life. Understanding our relationship with time, self, and performance. So, I'm going to go ahead and close these two other panels so that I can see. Looks like one section is an introduction, time, our most elusive dimension, the micro rhythms, the meso rhythm. So, it's broken it up into kind of small, medium, and large size rhythms, which makes a lot of sense. Wow. Excellent.
What this really means is that the research bottleneck is shifting. The question which has been the question for so many years and decades and centuries, which is where is the good information? Can I get my hands on the good information? That has shifted now to what unique insights, what unique perspective, what unique angle can I extract or distill from this information based on what's unique to me, my life history, my existing knowledge and expertise, my way of seeing the world or of seeing this subject.
But here's what happens next. You've got all these high-quality credible sources, but once again, there's a new different problem. You can't see how they connect. You're reading them still in a way linearly one at a time, quickly forgetting what you just read in the previous one and therefore potentially missing patterns and connections between them. Sometimes your brain needs to zoom out from the nitty-gritty details to the big picture so that you can then know which details you want to zoom back into. This is where interactive mind maps are so incredibly useful.
Let's use the example of the notebook we were just looking at about the rhythms in human life. You can see a few things. First of all, it's somewhat interactive. You can click and drag the mind map around and then it's broken down for you what are the main subtopics, the main parts of this big topic called biological rhythms. So, let's say I want to know more about one. For example, seasonal cycles in human psychology. You can see that it kind of opens up and shows me the next layer of subtopics. Every time you see one of these little right-facing arrows, that means you can click it and see another layer of detail. So, let's see. Cognitive performance. That's the one I'm most interested in. And it looks like we end up with really specific concrete examples like Alzheimer's, seasonal affective disorder. I'm actually most interested since my book is not really about um medical conditions. I'm interested in healthy adults and how their attention might shift. So, if I click on that sub-subtopic, it actually opens up a new chat that is specifically related to that thing that I clicked. It looks like uh it's giving me a bit more detail, which is even in healthy adults, sustained attention and working memory varies, uh, their cognitive performance varies according to the seasons. It looks like it's based on meteorological changes like sunlight and temperature, but also ecological factors.
So my question here is, what time of year is cognitive performance the highest in humans on average? Interesting. So it looks like in healthy adults, at least the sources that I have so far do not specify what time of year, cognitive functions are best, but in Alzheimer's patients, their symptoms are more severe in the winter and early spring. So interesting. I've reached a little bit of a dead end, at least in terms of what I am trying to learn. But this could absolutely be the starting point of a whole another loop, right? I could click right here and say save to note. So you can see it saves that response as a note. I could then convert it to a source. And then I could even do another round of discovering sources and try to find sources that do indicate what time of year cognitive performance is highest.
Now you've seen the five major updates, but before we talk about when to use them strategically, let me just rapid fire through some kind of smaller changes that also do affect how you'll actually implement this. New audio overview types. So you can now choose a deep dive mode where it really gets deep into the details. You can choose a brief mode if you want something short and quick. You can use a critique where it's actually criticizing and trying to disprove the content of the sources or you can use a debate format where there's two people debating back and forth. You can choose the length of the audio overview that you want, short, medium, or long. There is now a student version of Notebook LM that is designed to be used in educational settings like schools, which has stricter privacy for minors. It doesn't train on the data that you put into it ever. And they even have a $9.99 per month student discount that you can use to get this more affordably. And last but not least, it can now work with PDFs. You can either upload the PDF in its entirety or you can just provide a link and it will go find the contents of that PDF itself.
Now, zooming out for a second, here's the strategic question that determines whether this actually changes how you work and how you study. When should you use Notebook LM instead of an LLM such as say ChatGPT or Claude? There's really three very specific scenarios in which Notebook LM beats ChatGPT or Claude. The first is when working with massive information sets. Notebook LM is based on Gemini, which has a 25 million token context. When you have a lot of documents, a lot of data that you're trying to dump in there, Notebook LM is a great choice. The second scenario is when accuracy and precise citations really matter. If you're doing any kind of high-stakes work, whether it's in science or in health or it's for a test or it's for a client, and you really need to know that the responses it's giving you are correct, that it's not hallucinating, and you need to be able to click a citation and verify for yourself that it's correct. Notebook LM is one of the best, if not the best AI tool that is really source grounded. They don't try to hide sources. They don't try to make you forget about where this is coming from. They actually make sources first-class citizens in your work with AI, which is really rare I found among other AI tools. And third, Notebook LM is I would say the best choice these days if you learn best through audio or visual mediums. There are other AI tools that have visual and audio formats, but you kind of have to piece them together. You have to do some custom stuff. You have to move content from one place to another. Notebook LM is almost the only one that has all of that integrated into a single very user-friendly interface.
But here's the deeper principle that I think Notebook LM is built on. Your unique collection of information, your information, your notes, your thoughts, your ideas, your research has become more valuable in the age of AI, not less. Notebook LM and similar tools actually represent the evolution that I predicted way back in my book Building a Second Brain. We're increasingly moving from collecting information to cultivating insights. The CODE method that I first formulated years ago is still just as valid. I would say even more valid than before. You have to capture information. You have to organize it. You have to distill it with the help of Notebook LM and other AI tools. And eventually at some point you really need to express it. All of that can now happen at AI speed. Your current knowledge isn't being replaced by AI. It's being amplified by it.
Notebook LM is a very powerful tool. But to truly transform how you work with information, you need some foundational systems in place first. Things like how to manage your to-do list, how to handle your email, how to organize your digital world, and even just take notes effectively. I've created a comprehensive playlist on digital productivity and information management that walks you through all the key strategies from streamlining your daily workflows to implementing frameworks like my PARA method to building project management skills that give you real control and focus. The future belongs to those who can combine systematic information management with AI augmentation. What I recommend is that you start with these systems and then you supercharge them with AI. Subscribe below if you'd like more guidance from me on exactly how to do that.