NotebookLM is the BEST AI Tool for NEW USERS in 2026 🤯

Most people open up Notebook LM, they upload a few documents, they summarize it, and then they say something like, "That's pretty cool. " And they don't realize that they're missing out on one of the most powerful AI tools available to people today because they don't know how to use it properly. They're using it like chat GPT or Gemini 3 Pro and it's a completely different tool in its own lane that you can't even compare to those because in my opinion and for very specific use cases it's absolutely better than anything else on the market. You see the problem with large language models like chat GPT claw Gemini 3 pro is that they're creators. They try to predict the next best word based on what you prompted. It drifts. It hallucinates. It makes up stuff. It does not get everything right. But a lot of people think that it does and that's a major problem. People expect AI to be 100% accurate and that's an unrealistic expectation. But the closest you're going to get to that is an AI system like Notebook LM because Notebook LM is not just a large language model. It's powered by But what it is a rag system ra. And so here's an analogy. I want you to imagine that you hire the smartest person in the world and you ask them a question and they say, "You know what? I'll get right back to you on that boss. " They go to the library, they read a million books and then they come back and they say, "This is your answer. " Could be right, could be wrong. But then Notebook LM is like the employee you hire. You place them in a room. You give them five books that contain all of your business information, all of your medical practice information or anything else. and you say, "This is my question, but you can only use the information that is in this room. " If it's not in these files, then tell me you don't know. Which one are you going to trust with the information they give you? Obviously, the one you locked in the room because you know that when it tells you it doesn't know that it's being honest, it doesn't have the information. And that's not a limitation. That's a safety feature that protects you from ending up in a space where you're dealing with an AI's hallucinations and you're accepting them as facts, which a lot of people are doing right now. And so now, let me help you shift the way you think about notebook LM so that you stop using it like chat GPT and you start getting real results from it. So the first thing is with large language models, prompt engineering is going to be important. how you write the prompt, the role you give it, the order of instructions, the order of operations, the output format. You have to be very particular with how you write the prompt. That's very important. And then we have problems like contest engineering where we have to continually reexplain certain facts and details to an AI because of memory problems. But even more critical than prompt engineering and context engineering is what I call source engineering. So instead of focusing on becoming a prompt engineer or contest engineering, you need to start thinking about becoming a sourcing engineer because source engineering is about curating, cleaning, and structuring the files and the data you need to feed the AI to get the actual outcome that you want. And without going into too much detail, I have a custom GPT that I've built called synthetic data because I understand that there's only so much compute that I'm going to get from these systems. And the only way to get the leverage that I actually want is to give it the data. Because the biggest difference between how amateurs and pros use AI is the data they're providing, the structure they're giving it, the curating and the cleaning of the files. Chat, GPT, Gemini, and Claw have been trained on the entirety of the internet. That means that comparatively speaking, it's trained on a small amount of high quality intelligent content and the rest of it is trained on memes. Now, imagine if you could reverse or inverse those proportions and you can have an AI that's largely trained on high quality and intelligent content only with little to no memes or irrelevant content. That's what you can do with Notebook LM to a certain degree. Now, I want to walk you through a couple of different levels for how to use Notebook LM. So, the first one is just the student. And so, with the student, this is all about abstraction. You upload a math textbook. You ask it to explain a formula. help you with the problem, etc. You're just extracting data. You upload a history book. You ask it questions about what happened in 1752. The best thing you could do in those situations is upload a history book, create a podcast overview, and listen to it. It's going to be far more engaging than simply trying to sit down and read through all of that documentation. And you can create some very detailed podcasts. You can create a deep dive that's a lively conversation between two hosts unpacking and connecting topics in your sources. You can do a brief, a critique. You can even do a debate, a thoughtful debate between two hosts, illuminating different perspectives on

your sources. And this is good because this keeps you from falling into a vacuum or confirmation bias. just hearing what you want to hear or finding somebody to agree with you with which is what a lot of people do when they use AI these days. The next use case is you can use notebook LM as a study buddy. You can click on customize quiz and you can ask it to create a quiz and it's going to ask you questions about the uploaded content. And the third use case is simply a TLDDR. You see I have several different or I have 253 YouTube videos uploaded with several different documents. I can ask notebook LM summarize all available data into 10 talking points. And if I'm studying this information, this is going to allow me to see what's the best places for me to place my emphasis so that I can understand the concepts that are being laid out. This is a mind map of my YouTube channel. If I click this and expand it, I can see that full bird's eyee view of everything that my YouTube channel and my content is about. And this gives me the clearest path for learning from my own content and doing so many other things. But that brings us to level two. In level two, you go from just being a student who's using Notebook LM to learn and you're somebody who's trying to find connections between ideas. You're more of a detective. You're researching and investigating. And so you're going to be uploading several different documents. And maybe it's not your YouTube channel. Maybe you've created a notebook that has information that you've gathered from the internet because you can come inside of notebook LM and you can do a web search, a Google Drive search, or you can do deep research, a fast search, great for quick results or in-depth report and results. And you can ask things like what are the most effective side hustles for 2026? And now Gemini is doing deep research and going to return an indepth report right here inside notebook LM. And then we can choose which of those documents that we bring into this notebook and which ones stay outside of this notebook which is a key feature of notebook LM. You control the data. You control what goes inside the notebook. But now we can compare documents from this year, trends from this year to trends that are coming in the next year. We can look for discrepancies. contradictions. You might be trading stocks and maybe Apple or Tesla or SpaceX or whatever is a company that you're interested in. You can upload their quarterly earnings calls and you can look for different things that have been said that might give you an indication of where the company is going. If there's a problem that's being hidden, a lot of different things you can do. You can investigate using Notebook LM. There's just so much you can do with it because it gives you this ability to take all of this information and easily grasp it at one time using these tools, so to speak. Another way to use Notebook LM as an investigator is to do thematic searches. So, this is a notebook titled the psychology of video hooks and audience retention. So, I have several different deep research reports and web pages that have been uploaded as well as videos from YouTube that I found about this topic that I watched and liked. And I can ask it a simple question such as what is the one objection they all address. Now that's a simple question but to find one objection that all 62 agree on is a task that would take chat GPT Gemini claw just it would take a long time but notebook LM because it is a retrieval augmented generation system and it's only going to just pull from these documents. It's not going to take that long at all based on the provided sources. The single objection that content strategies, hook formulas, and retention guides all address is the viewer's immediate subconscious question. Why should I care? Now, if we come back to this other notebook, we can see that this deep research is completed and we have 31 sources. And so, now this is the deep research report at the top. And at the bottom, there are 31 sources cited and there are three not cited. If I click on the three not cited, we can see that there's an article from Shopify, Zip Recruiter, and Blue Host. I don't know why they weren't cited in the report, but these are the 31 that were cited. And I can go read any of these because Notebook LM provides the link right here. At what point does a workflow stop being worth the effort? So, I can take a glance at that and I might decide this isn't relevant to what we're discussing here, so I don't want that. I might even look at the title and just say that I don't want it. But after I've gone through these, I've chosen what I want or don't want, then I can just click this import button and immediately my sources have been updated with relevant information from the

internet that I chose to add as well as the deep research report. Here we have a content strategy patterns and monetization data table that was created using the available sources. And so you can see the sources here, the action items, the key insights, and the content type. And if I want to, I can export this data to Sheets. And now that I have this information inside of Sheets, all I have to do is click this ask Gemini button right here. And I can use Gemini Alpha to process spread tables inside of Sheets. So for this next example, we're going to do a web search. Select web. And then we're going to say Wilded Heart by John Eldridge. And just to give you some brief context, Wild Heart is a book that I read when I was about 23 25. It's in my top five for uh personal development books. Absolutely love it. It's about discovering biblical masculinity. But I'm just going to ask what metaphors from Wild at Heart could I apply to my product launch? And one of the things that it discusses or the primary in the book is that every man has three core desires written on his soul. Number one is the desire for a battle to fight. Number two is the desire for an adventure to live. And number three is the desire for a beauty to win. And everything that you want to know about a man is tied up in those three core desires. And so I'm interested to see how Notebook LM cross-pollinates these ideas because cross-pollination is probably next to Notebook LM in general. Another underrated, overlooked idea for using AI. And so this creative agency is in the ideas department. and check out the video that I pinned at the end of this video uh where I go in detail about how to set up your own personal notebook as a creative agency that has different departments, rules, instructions, and prompts. You're going to love it. But based on your founder blueprint, uh the bridge gap between the spiritual masculine journey of the book and the creator to entrepreneur shift in your business strategy, the caged lion. Rounded Heart argues that a lion in a zoo is safe and wellfed, but it has lost its soul because it's no longer wow. It has been tamed and caged. Your business application, the cage, the social media algorithm. It feeds the creator likes, views, small bonuses, but it keeps them trapped in a small space where they have no control. This is just making me want to read the book again. The initiation creator to founder metaphor. A boy cannot become a man on his own. He must be initiated by other men. He must go into the wilderness, face a trial, and return with a new name identity. Business application. The boy. The content creator who chases viral trends and magic prompts. Childish things. The wilderness. The scary transition of selling a product. It feels safer to just post free videos. But the adventure and the wealth is in the wilderness of entrepreneurship. The launch hook. You've mastered the playground, social media. Now it's time for your initiation into business. This course isn't just information. It's the transition from content creator to media executive. That might seem simple and it might seem small to you, but that is astronomically large. Let me explain why. Notebook LM is projects for Gemini Pro. The next feature about Notebook LM that I want to discuss with you is actually inside of Gemini. If you come to the message box and you click on this plus button to add files and you scroll to the very bottom, you'll see notebook LM. Now, if I click this, it's going to show me the available notebooks that I have. I'm going to reference my prompt library. And I'm just going to give it a basic prompt. and I'll say something like, "I need help coming up with an idea for a YouTube video or fleshing the idea out. I want to talk about the Notebook LM updates that have happened in the last 30 days in a way that's meaningful and valuable to my audience so they understand the true power of this tool. Can you help me? " Now, I didn't tell it to pull the prompt, but because the notebook is there, it's definitely going to pull the prompt. And so now we see that it's giving us a primary angle and two backup angles. And if we come back over here, the notebook LM and if we scroll down to the primary objective, you can see that it takes a generic topic idea and a specific viewer and return a primary angle and two backup angles. So I can confirm that Google not only read the document, but it's actually executing that prompt in real time. And if we look over here in the AI studio, there's an audio overview, video overview, a mind map. Then there are

reports. And when we open reports, you can create your own. You can create a briefing doc, study guide, or a blog post, a strategic plan, a brand manifesto, a beginner's guide, a concept explainer. And so I might say, you know what, I want to create a brand manifesto. And so I click on this icon right here. I choose the language, and then I write a prompt. Now, let's say that I want to create a prompt. I'm just going to create my own prompt right here. I want to write an introductory newsletter that also functions as a brand manifesto. Write a prompt that will accurately guide the AI and also include my voice and style based on my YouTube videos so that it sounds like me. And so now, Notebook LM is going to write the prompt for me. There's no reason for me to ask Chat GPT or claw to write the prompt when I know that no one can write a prompt that is as on brand for me as Notebook LM with access to 259 of the most relevant documents to Corey Mlan. As a content creator, there's no way I can get all of my YouTube videos uploaded to chat GPT or even claw for that matter. But Notebook LM, easy peasy lemon squeezy. And here's a little trick that a lot of people don't know. But you can upload documents to notebook LM that don't have anything to do with the topic of the notebook and they have everything to do with how you work with the notebook. For instance, if you want your responses formatted a certain way, you can upload a document that tells notebook in detail how to format its responses. upload a document that tells it the voice to write in when you ask it to write content. And so these documents that you upload function like system prompts. Notebook LM reads them and then whenever it responds, whenever it does anything, it does it according to the documentation. A quick example of this is my YouTube made simple notebook and I have my channel uploaded here, but I also have some markdown files uploaded. And so for instance, I have this right here, the idea angle system. And if I open it up, it's going to give us a summary of it right here. But then at the very bottom, you can see that this is a full detailed mega prompt. And it's walking through all of these different steps when it's coming up with ideas for a video. So I'm going to close this and I'm going to upload a new document. help me come up with a new idea for my next video. And so we're going to see if it uses the idea angle document. And so I can tell you right now that it's following the system prompt flow original idea generation. Let's develop a new concept to initialize the idea angle system. Please provide the following three inputs. The topic, the target viewer, the transformation. Once you provide these details, I will generate three specific angles. primary plus two backups with title, thumbnail, and hook concepts. Content creators who want to launch digital products, courses, communities, and build a sustainable income, creating content that drives traffic to their offers. So, now that we've provided those three, Notebook LM is going to use the system that I created to go ahead and build out the entire YouTube video for me. Then of course, Notebook LM doesn't do the best writing yet. So, I'm gonna take it over to Claude, polish it, and then I can record it. It's based on my YouTube channel, so I know it's grounded in my content. It's grounded in everything that I say and do. I could even, and I think I will. I'm going to kind of jump in here and cut it off. I know it's giving me an idea, but there's a video that I believe ChatGPT and Gemini recommended that I should do about how to create a digital product with AI in like 10 minutes or something like that. Oh, how to create a digital product in 24 hours that actually sell. Listen, I'm telling you guys, Notebook LM is the truth. It's powered by Gemini 2. 5 Flash, a smaller model of 2. 5 Pro. I believe I read somewhere a model picker is coming. So once three pro is inside notebook LM, everybody else can just hang it up. This is going to be where I'm hanging out at. So option one, the identity shift. Option two with three levels of content. Man, that came directly from one of my videos. I'm going to go with option two. So once again, we have it letting us know where it is in the flow. Angle two, the lazy launch. Excellent choice. This is the architecture. So now it's going to give me a list of titles. How to create a digital product in 48 hours using AI. I'm going to go with one, but I'm going to change this to 24 because you can do it in a day. You can probably even do it in two hours. This is the hook scripting and structural notes. I'm

just going to go with this here as the title. It's going to then rewrite the hook using the hook. md. Once again, there is a full documentation for how to write the hook. Different hooks, different ways to write it. And this was all created using notebook LM pulling in different sources, different documentation about hooks, about different designs. And what I can say is this is that my hooks aren't what they should be yet, but they have definitely improved since I started using this system. And so now Notebook LM, it's writing the script as well. Me scroll up and just see if I can check out the hook. So, Dflow scripting, how to create a digital product in two hours using AI, highly aggressive promise, two hours versus the original 48, which makes the click-through rate potentially higher, but delivery must be extremely precise to make this 2-hour promise real. We will focus the script on creating a high value ebook or guide. So, this is super smart. It's thinking ahead. Like, if you're trying to do a course, you really can't create a product you can sell for hundreds of dollars in two hours. You just can't. It's going to take several days and weeks uh to get it done properly. But video format, archetype, the teacher, cheat code, tutorial, fastpace, the blueprint. This is the script. This is the hook. Most creators are broke because they're stuck on the content hamster wheel. You think the way to make money is to get more views. It's not. The way to make money is own an asset. Talking head. But here's a trap. I'm going to ask it to focus on the hook. Can you please write me three hook variants using the hook module engine? And all we're checking for here is that it uses the actual documentation that we've uploaded for writing a hook and then doing so because the way that document is set up is it takes a lot of the guesswork out of actually writing a hook so that it can't miss. That's the whole point of it. I write these documents so that they're very deterministic and the AI doesn't have to do any thinking. Just follow the directions. Just follow the yellow brick road. So number one, the direct promise PVSS formula. This is probably the one I'm going to go with. Most creators spend six months building a course that nobody buys. Today I'm going to show you how to build a profitable digital product in exactly two hours using AI. We're going from blank page to active checkout link before lunch. That's good. Stop trying to get paid by the algorithm. The creators making real money aren't chasing views. They are building assets. I'm going to give you the exact AI workflow to turn your knowledge into a product you can sell today without filming a single video. That's good, too. Uh, the cheat code demonstration archetype. I challenged myself to build an entire digital business from scratch in one sitting using this specific AI prompt chain. I wrote, packaged, and launched the guide in 120 minutes. Here's a step-by-step blueprint so you can copy it. And so basically when you let it know the hook, everything else is going to flow in line as far as the transcript and the rest of the video that it's crafting. And so Notebook LM is now going to write out the rest of the video. And if I wanted to, I could come over here and create a slide deck based on this output that it gives me. I can turn it into a source. Turn I mean, excuse me, I can add this output to the studio and then I can turn that studio source into a document source. Then I can turn around and create a slide deck based on that source alone. All right. And so now what we want to do is save this to notes. I'm going to come over here to the lefth hand side where it says select all sources and unselect them. And then I'm going to add this to our profitable digital product script to a source. I'm going to make it a source. So now notebook LM is going to add it over here. And so now all we have is this one source right here, the 2our profitable digital product scripting guide selected. Then if I click on this right here for uh slide deck, paste in my design prompt and I can either go with presenter slides or a detailed deck. Click generate. Notebook LM is going to create the B-roll slides for actually presenting this information. And if I wanted to turn this into a digital product, a course where I screen record and show people how to do it, but also teach them at the same time, I now have the assets to do that. And the design is just going to be uh next level. This would take me hours to generate inside Canva, but Notebook LM is going to get it done in about 2 minutes. So, if we open this up, you can see that Notebook LM has created an absolutely stunning slide deck. And I give ChatGBT the credit for rewriting this design because I kept doing it and Claw just kept writing it the same way where it was actually writing the slides. And I just want a design element. And so, it added the glass morphism like I wanted. And

everything else, everything else is just beautiful. I absolutely love this slide deck. It even pulled from one of my videos about how to make a ebook with Google Docs and it gave us a clear image of the file edit view format tools because this is where you go. You go to tools then voice typing. This is beautiful. Even though it got some of the smaller text wrong, that's a minor thing. This is good. This is all gold. And so, like I said, if I wanted to, I could easily take this and use this for presentation purposes. And I'm inside of Notebook LM right now. I just want you to notice that I am inside of Notebook LM and I'm running through the slides right now. So I could actually create the digital product inside notebook LM if I wanted to. And so hopefully by now you understand that. And there are so many other things that you can create with Notebook LM. And I highly advise that you watch the video at the end of this where I show you how I turn notebook LM into a creative agency. Especially if you're a single creator. There are different levels to how you use Notebook LM. There's the student approach. There's the researcher approach. There's the creator approach. And there's just so much that you can do to control the environment, the data, the sources, and to create predictable, highquality outputs that aren't confirmation bias, but that actually increase your learning, increase your intellect, and also increase your income if you wanted to. If you got value out of this video, make sure you hit the like button, subscribe to the channel, and as always, take care. Have a good day and I'll see you in this next video where I show you eight notebook LM use cases you probably didn't know existed.