Transcription
I have just turned a free tool from Google into a team of five experts who work around the clock and take no vacation. A doctor who analyzes my lab results from the past 5 years, a chef who knows the contents of my refrigerator and creates personalized recipes for my family's tastes. And a consultant for gaining social media followers. And a couple more. And this is not science fiction, this is Google Notebook. And today I will show you how to assemble such a team for yourself in one evening, without code, without expenses, from scratch. Subscribe, like in advance, and if you don't like the video, you can take it back. Let's go. But before we start, I will provide a little context. For those who are hearing about NotebookLM for the first time, it is a free tool from Google that works with your files, videos, books, and much more. I decided that the service itself would do an excellent job of self-presentation and tasked it with explaining its functions in 15 seconds through a video summary. Let's watch together at 1.5x speed. >> This is NotebookLM. You constantly have to switch between tabs and documents. Hours are spent connecting ideas. Such information overload, of course, hinders concentration on work and creativity. What if an assistant could analyze everything saved on its own? Let's see how it works. Essentially, it's a smart workspace that turns any document into a personal knowledge base. The main difference from chatbots is that it only works with uploaded sources. It is grounded, meaning it doesn't make anything up; it only analyzes the information provided to it. The process is very simple. Upload materials, ask questions, and get answers with citations. Many formats are supported. Everything for a project can be gathered in one place. One of the cool features is audio summaries. It turns texts into podcasts that you can listen to. >> Okay, Google, thank you. Now for the main part. Let's not just study the functions, but immediately apply them to real tasks. And we'll start with a simple example, working with video, and then I'll show you a case that personally impressed me so much that I changed my approach to self-exploration. Remember how many times you wanted to summarize a two-hour video but didn't have time to watch it. First case, working with video. I am now creating a new notebook. And, as you can see, NotebookLM is omnivorous and ready to work with files of different formats. But I'm now switching to YouTube, where I've pre-selected a couple of very high-quality videos about gaining social media followers from my acquaintance Dima Torgov. I'm copying the link to the YouTube video and pasting it into the websites tab. While the full transcription of the entire video is happening in seconds, I'll copy the link to the second part of this training material and add it to the current notebook as well. Everything is ready. As you can see, I didn't even have to wait. The full transcription of two training videos, which, for a moment, together take 2 hours, is ready. NotebookLM handled the task in 10 seconds. Okay, look, I also don't have the opportunity to watch these videos for 2 hours, but I want to know exactly what the author is talking about in them, because I trust his experience and don't want to miss any important nuances. Therefore, first of all, I go to the notebook chat and ask the service the following. Provide the essence of these videos in the form of a step-by-step structured action guide. And now that the summary, broken down into seven steps, is ready, as I requested, I can start working with it. And what is most valuable to me in NotebookLM is that I can be 100% sure that it hasn't made anything up. Note that next to each thesis, there is a reference to the author's specific words that he says in these two videos. And now, to demonstrate the service's capabilities, I will ask it to create from this step-by-step guide. Let's say, a mind map. A table, an infographic, flashcards, tests, and a presentation, and an audio summary in the form of a debate. Note that for many formats, you can create custom settings, but we won't do that now, as I'm eager to move from the format of this boring lecture to practical cases. Therefore, let's quickly check the result together and go create something very interesting. I'll play the audio podcast generated by Google in the background and show everything else. Welcome to the discussion. Today we are diving into the world of short videos and Instagram Reels algorithms. It seems like every second person today is a content creator, and everyone is looking for that one, well, secret ingredient to success. And this poses a fundamental question for us, >> yes, a question that, in essence, lies at the heart of our entire conversation today, based on a very detailed guide to creating Reels. What truly drives growth? Is it the result of, you know, cold calculation, strict adherence to proven formulas, algorithms, psychological triggers, or is the key something entirely different, in authenticity, in spontaneous self-expression, in, well, documenting your unique journey? >> In my opinion, the answer here is quite obvious. Success in such a competitive environment is a science, not some kind of magic. And I will argue that a systematic, almost, let's say, engineering approach to content creation is the most reliable, if not the only, path to stable growth. >> And I, in turn, will present the opposite point of view. I believe that this excessive reliance on formulas, on constant optimization, is not just harmful, it kills the main thing – genuine creativity, that very spark that creates a truly strong connection with the audience. >> In short, to summarize the capabilities of NotebookLM, they are as follows. You can take up to 50 sources on the free plan and up to 100 sources on the paid plan for one notebook. Moreover, sources can simultaneously be various files and videos, as I showed in the example. And then within the chat, you can work with this information as needed. You can ask it to find something in these files, analyze it, and then you can format or structure the information through a mind map, through a table, through a video or audio summary. You can create tests or flashcards for memorization. An infographic comparing the plans is also attached. And I won't dwell on this now, as it is open information. But, in my opinion, this functionality looks quite decent, especially since it is available for free to everyone. And this is something I didn't have when I was studying at university or school. I am glad that now I and you have access to such a tool, because futurologists and scientists say that we will now learn throughout our entire lives. There will no longer be a situation where you get an education once in your life and that's it. Due to the very high speed of technological development in the world, the world is changing every day. Yes, you see this yourself. Therefore, now adults must become eternal students, which is what I propose to do right now. Second case – talking to a book. What if I told you that any author of any book can become your personal mentor right now for free and will be ready to answer any of your questions? It sounds like science fiction, but look. To do this, we create a new notebook and upload the book file, which I previously saved on my laptop. Today it will be the almanac of Naval Ravikant. The book is in English. Pay attention to this, because from now on I will communicate with it only in Russian. And now for the most interesting part. You go into the chat settings and set your own parameters. For example, you write: "Communicate with me as if you are Naval Ravikant, my wise mentor and friend." And click Save. And now in the chat, you can ask the author of the book about anything you want. For example, "Naval, hello. If you and I had only one minute, what would be the most important thing you would tell me from your book and your experience?" And at the same moment, I receive a very personal answer. NotebookLM has studied the entire book and provided an answer in just 10 seconds. Wow! Isn't that impressive? Write which book author you would like to consult with and discuss your life. And look, you can upload not just one file, you can upload an entire library and arrange a council of book authors in the format of a podcast or debate on your question. What else can be done with a book? In the studio section, you can work with the book in the same way as I did with video. I will ask it to provide information in different formats. And at the end of this episode, I will show how it handles the task. Third case – a chef in your pocket. And the third example of use is culinary. What if you ask a neural network to give a precise step-by-step recipe for a dish? But specifically for the products you have in your refrigerator right now, taking into account your diet, possible allergies, and other personal preferences. At the same time, I will complicate the task by adding maximum personalization. And this time I will not give it my sources, but will ask it to find everything on its own. To do this, I create a new notebook and immediately select the deep research mode and enter my request. Hello. I am actively studying the topic of longevity. My personal goal and the goal of my family is to live up to 120 years in health and energy. I want to compile an extensive selection of cookbooks by modern authors in the Zzh, PP, Wellness, Longevity Life direction. And these can be both foreign and Russian authors. It is important that these books contain real step-by-step recipes. The books should be published from 2020 to 2026. Well, and so on. As a result, it found 34 sources of information. Now I can either select all of them at once, or exclude something. And this is a key aspect of working with NotebookLM, as it uses a RAG system at its core. And to avoid overloading you with incomprehensible terms, I'd better explain with a simple example. The meaning of a RAG system. The result of information retrieval directly depends on what sources you uploaded at the very beginning. Just as the result of the dish you are cooking in the kitchen directly depends on what ingredients you use. So, if you accidentally decide to add herring to a dessert, which for some reason ended up in the cake recipe shopping basket, then most likely your cake will not turn out. It's the same here; the result at the output will depend on what sources are uploaded to this particular notebook. I will select all the sources it found. I trust it 100%. And now, in the chat, I will give it this request. Look, today in my refrigerator I have chicken fillet, various root vegetables, lactose-free cream, tahini, various spices, there are also tomatoes, cucumbers, and eggs. Also, in long-term storage, I have couscous, bulgur, brown rice. Well, in general, I'm listing everything I have there. There are oils, like olive, truffle. And provide a step-by-step recipe for my set of products. Cite a specific source. Look, it did an excellent job, I think. But reading the recipe is not very convenient. Let's ask it to also format this recipe into an infographic. I copy the text, click on infographic settings, and paste the text here. Visually, the result is very cool, in my opinion. Everything is clear, accessible, and now cooking will be even easier. I see that the neural network still makes some minor typos in words, substitutes letters, but I am 100% sure that this is just a matter of time. And in a year, NotebookLM will no longer make such minor mistakes. Do you realize the scale of this tool? A pocket chef assembled in 5 minutes. For free. But there's a nuance. NotebookLM works as well as the quality of the sources you provide it. Remember how I talked about herring in a cake? If there's garbage in the sources, you might get garbage at the output too. Therefore, I spent time selecting thirty-four authoritative sources, and the result is personalized nutrition tailored to my goals. What if you dedicate an hour to this, not 5 minutes, upload your lab results, allergens, family preferences, budget for groceries to this same notebook? In my opinion, this all sounds like a great idea for a full-fledged app on your phone, for which I would personally gladly pay, well, I don't know, 2-3,000 per year. And by the way, this is exactly why I created this channel. I am launching a YouTube series where I will build a business with artificial intelligence from scratch to millions of cases before your eyes, from idea to first sale. With all the successes, failures, hopefully not, through vubecoding, where code is written with words, not hands. The series starts as soon as the channel reaches 1,000 subscribers. Therefore, subscribe if you want to either just observe this, or repeat this path and launch your own project. And now, two more cases that personally, well, really impressed me. Fourth case – deep self-exploration. Perhaps this will sound unexpected for an AI channel, but I believe that an entrepreneur's success is 80% determined by how deeply they know themselves. And Naval Ravikant, by the way, said: "The better you know yourself, the better you understand, the faster you find your place in the world." And look, I am a member of several business clubs: Oscar Hartman's, Mikhail Grebenyuk's, Igor Budnikov's, and it doesn't matter the scale of the companies of the entrepreneurs, 2 billion per year or 10 million. In each club, we periodically discuss the metaphysics of an entrepreneur, that very tum-ba-yumba, as Misha Grebenyuk says. And over the years of self-exploration, I have accumulated a whole library of data about myself. I decided to conduct an experiment with NotebookLM, uploading literally 10 files about myself into the notebook. I remembered that I recently took the CliftonStrengths assessment – this is a personalized profile of leadership and talents from Gallup. I asked it to analyze the results and prove or disprove the data, relying on other systems, other files. And I really liked how it handled this task. It was as if, as if, you know, I saw a visualization of my thoughts, only written by NotebookLM. In general, it seems I've just gained a second brain for myself, if I can put it that way. Therefore, I recommend you upload all your data into the notebook and ask it to look for interesting connections and provide you with some non-obvious insights about yourself. My mind is already full of ideas on how I can further use this simple neural network. Share in the comments what thoughts my examples have inspired in you. I have prepared the final fifth case for you, which, I think, can replace another paid app on my phone for me personally. And fifth case – a doctor in your pocket. As you have already understood, I am interested in the topic of human health, energy, and longevity. And look what I did. I downloaded all the medical lab results that I had taken over the past 5 years from different places, like Invitro, Gemotest, there was also the Kivach clinic, and others. And it turned out to be 49 files. I uploaded them into a new notebook and started formulating requests for it. Show vitamin D in dynamics over the past 5 years. Compile a list of tests that I need to take in the spring of 2026, and justify why. It provided instant, accurate answers to all my requests. And if you do the same, share how you liked the result and do you understand, well, that this could be somehow useful to you? What conclusion can be drawn now from all five examples that I have presented today? Any artificial intelligence is just a tool. And only our own minds are capable of finding a solution on how to use this tool to improve our personal lives, health, business, or other areas of life. And if you ask me: "Georgy, how will your channel differ from other channels about artificial intelligence?" then, probably, by this. Here we will not just study neural networks in a race to catch up with a runaway train, as new AIs and their updates come out every day. No, here we will improve our real lives, build businesses, earn money, and try something new each time. Well, shall we wrap up? Let's watch together a minute-long video that NotebookLM generated from Naval Ravikant's book. >> It's about how to achieve wealth and happiness. The main idea: wealth and happiness are not just luck, they are skills that can be learned. Who is Naval Ravikant? He is an entrepreneur, an investor who has deeply studied these topics. His journey from a poor immigrant to a Silicon Valley icon speaks for itself. But first, let's understand what wealth actually means. Naval looks at it differently. This is where the difference lies. Money is a way to transfer value, and wealth is value itself. The point is that you need to own assets. True wealth is a share in a business, not just a salary. So what's the formula? It's simple. You need specific knowledge, and then leverage to apply it. Specific knowledge is not what they teach in university. It is found by following your curiosity. Remember what seemed like play in childhood, but work to others. This is your uniqueness. Okay. Specific knowledge is acquired. Now you need an amplifier, the very leverage. Naval identifies two types. For traditional capital and people, you need permission. But there is another. Code and media are the leverage of the new era. It can be created by anyone, and it works while you sleep. >> And Naval's idea that everyone in modern times, by using just two levers, code and media, can create a successful and large-scale business, resonates deeply with me. You, and I, and Masha from Moscow, and Pavlik from the village too. It absolutely doesn't matter where you live. After all, I'm not a programmer either, but I've been in business for 10 years and decided to record a series here on YouTube about how I will launch a new project using neural networks, vubecoding, and scale it through media. And honestly, it will either end with money in the cash register, or complete failure. Subscribe, I'm launching the series as soon as I get 1,000 subscribers. Give feedback in the comments on what you would like to see in new videos. After all, your subscription, your comments, and likes are the most valuable currency. This was Georgy Rivera. See you soon.