I Learn Faster Than 99 Of People Notebooklm Claude Code Obsidian

90% of the people listen to experts and never change a single behavior. They never close this loop. They always stay in a learn mode. I built a system that closes it. We set a goal and we feed expert content into Notebook LM and we connect Cloud Code to it. Then we set a goal and we get a protocol with experiments based on our answers. And in this video, I’ll show you how to build this and how you can set it up in just 20 minutes. We’re going to feed 300 of Huberman podcast from a terminal, run a cited interview, and turn it into experiments in your morning routine and build a reusable skill for any expert. So, let’s get started. Let’s start with a demo. And my goal is to improve my health, improve my focus. I want to have more energy. I want to go to the gym. I want to be fit. Now, that’s my goal. Another part is we need to do the research about how to achieve this goal. This part is handled by Notebook LM. And Notebook LM is tool where you can load this bunch of YouTube videos or other sources and start asking questions. Now, the challenge is is how to actually load those sources into Notebook LM. Let’s say I really love Huberman and I want to learn from him as an expert. And I want to somehow uh get a grasp on all of his videos about health. And How can you do that? You can’t just tell You can’t just go into Notebook LM and tell Okay, so let’s just add the this YouTube channel. You can’t really do that. The tool that’s going to help us is Cloud Code. And I connected it to my Notebook LM and I created a skill, which you can also download in the description. And it listed my all 53 notebooks. Now, that’s very cool. That’s very cool. That’s one part of the solution is this connection. What can we do is, let’s say I want to create a new notebook about Andrew Huberman and I want to get all of his videos regarding the health into a new notebook. And I’m giving it also a YouTube channel. And the goal here would be just the first step is to upload create a new notebook and upload those 300 podcasts about health. So, we can have a verifiable and great knowledge base based on real facts and citations. Okay, so we got uh 400 videos and right now here is the list and we want to understand which of them are health related. I really love this part about Cloud Code that I can pick the actual sources collaboratively with it. I can tell which sources I want and which I don’t. So, here is the plan. We’re going to filter out uh different videos. Now, let’s say we want to like upload like a recent one recent 200 videos. And let’s just execute this. And now we’re going to see how Cloud creates a new notebook and upload those videos for me. I imagine yourself just going to each YouTube video getting a link and just uploading it manually. I just such a big pain. Right now, uh okay, so we can actually go back. We created both notebooks. That’s but we are interested in the recent one and we want to see the sources appearing. Okay, so we’re going to have two of them. One of them is archive and one of them is recent. So, let’s see the sources are appearing. So, we already have 60 sources. So, all of them are appearing and those are the real videos from his YouTube channel appearing just now in real time. How cool is this? It’s already 96 sources. It’s so amazing and you can see all these videos about improving learning improving health with a strong brain-body connection. And you have all of the access to this knowledge and you don’t have to watch the podcast. You can save so much time by just doing that if you want to learn about a new topic. We have almost 200 sources and we are actually done. We are actually done. And the way to go about this is let’s say Okay, we want to ask a questions. Let’s say I want to improve my health. Uh where do I start? That’s one way to interact with Notebook LM. You can do it in here in the interface. You can also create all the different artifacts such as reports, videos uh audio review. You can also create a customized podcast. You can do flashcards if you’re studying something. Right now, Cloud finished the task and here is the report. So, here we have answer, right? We have a quality of sleep and morning sunlight. And here we got a citation to the video. We can click here and we can actually get back and trace the exact citation. And here it tells about uh light exposure, sunshine, movement. Here is the part about stress modulation. Now, this is already cool. We can talk to 200 podcasts uh from Huberman. And you can do it for any expert, actually. From any YouTube channel. In just one command. But here is the exact problem that this is where everything stops at the learning stage. You can ask questions, but we are missing the steps where we actually plan uh to do those experiments. Uh let’s say we want to wake up earlier, we want to go to the gym. And then we also want to act. We need to schedule somehow this into a calendar. And we need to see we need to review what is the effect on our health. Are you achieving our goals? And these parts are completely missing from this picture. If you just imagine a standard research where you open a bunch of tabs uh or even if you’re using Notebook LM in this way, you can ask questions, but then so what, right? How do you act on that? I can’t really set a reminder from here. How do I change my actual behavior? And this is exact problem. We have all this knowledge from Notebook LM, podcasts, articles, chats, research. But there is a gap between actually implementing it in our life. Uh and the way to implement it in our life is to schedule it into our calendar, set reminders, uh do morning routine, run some experiments. And this is what’s really missing. And it’s a very challenging problem. How do you actually have a system which helps you to bridge this gap? And this is where Cloud Code comes in. Where we have this Notebook LM, which is a source of knowledge for us where we can trace by the citations. The knowledge is verifiable. Real experts uh provide real insights. Now, and I think of a Cloud Code, which lives in a terminal, as something that can help me to actually act. Actually act on this knowledge to help me to manage my calendar, to schedule my focus blocks, to help me to track my goals. Am I achieving those goals or I’m not achieving those goals? And let me show you an example of how we can use Cloud Code to help you to close the loop. Let’s say Okay, so we can access this knowledge from Huberman, from this expert, and we can design a protocol, which we actually would do later, about towards achieving our goal. And now we need to act. Okay, so we can schedule events into calendar. We can embed this into our morning routine to see actually are we achieving our goal or not. Cloud can ask questions about that. And we can actually run real experiments to track whether or not we are achieving this. And then we can have a health dashboard, let’s say, to see everything at a glance. Here is the example of my experiments which I track. Let’s say how this experiment about gym consistency. I recently started going to the gym. Now, I have this experiment. It’s just Obsidian note. But the type experiment and it relates to my health dashboard. Now, here I fill out some information. And here is all of my data, so I can track it. Track the results. I have my volume. I have my gym sessions per week. Here is a target. Here is how actually it went. And now the most important thing for me, uh the goal of going to the gym for me is to have more energy, have a better mood. And here is the actual data. So, we can see here the volume. How much I do in the gym. And here is the data about my mood and energy. And here I run some analysis. Here is the energy after the rest day. On average, it’s 6.1. I rate it from 1 to 10. And then after a gym day, the energy is consistently higher. And the mood is already also higher. And the sleep quality is higher. Now, I can track those results. And the only way I don’t forget about this is because I have this morning routine skill, which lists all of my experiments which are active. And then I instruct Cloud to actually ask me about Okay, how is this experiment going? Like any observations here? And then based on the observations, we can schedule to next actions. Okay, let’s go to the gym or just log how are you feeling about that. And we fill this daily note where Cloud asked me about Okay, so how is my mood right now? What is my energy? And then we update the goals uh log. Now, let’s try to actually do the action. Let’s go through the stage where we Okay, let’s design a few experiments for our health based on all of this 200 episodes of a podcast. I’m back to my Cloud Code and I’m just telling what is my goal. So, my goal is to improve my health and I want you to ask questions uh from the recent episodes of Andrew Huberman to understand Okay, so what are the dimensions for the health and I want to basically run an interview uh based on all of his expertise to assess my let’s say current health to this notebook. And I also want to track the questions in Obsidian. And we have a Notebook LM base for that. And I want to tell you to create a dashboard. All right. So, right now we are creating this folder where we’re going to have questions saved. So, we can keep track of them and we can always get back to the source. We’re going to save those actual results and responses from Notebook LM into our Obsidian and based on that, we’re going to design interview for the main, let’s say, dimensions to assess our health. It runs six obedience in parallel to ask the questions. Here is example of the questions and it saves into this temporary file. And I want to save them into my Obsidian so I can see the citations and backtrack to the source to the exact transcripts. Right now, we just imported those 200 sources and I believe those are transcripts. So, you can refer back to the actual citations. All right, so Claude has finished and got response back into our Obsidian wall. So, we got six questions and we got this dashboard where here we have individual questions which we asked. Here are the questions about the supplements, about the exercise, sleep optimization. And actually, you can see that they are also here in the Notebook LM interface. So, it’s kind of synced in a both way. We can use both interfaces. And those questions Claude asked. I didn’t like actually do that. And then we got those responses back into our Obsidian wall. Let’s say about this like I don’t know, maybe about the supplements. That’s a question. And then you can see it back in our Obsidian wall. So, that’s the supplements protocol and that’s exactly the same response. Now, let’s look maybe at this one. Okay, that’s the first citation for sleep and we have it also here in our Obsidian. And I also asked Claude to analyze new citations make sense and seven out of eight citations there are strong matches. So, overall citations are accurate and well-grounded. So, with that you can see the actual answers. You can backtrack to the source to the actual transcript. And that’s like huge huge transcript file and you have this dashboard like where you can see all of the questions which we asked and you can see the sources. Here is example of the video and here is the transcript. And we can see where it was cited by. And this is where we are right now. We actually got well-grounded responses from the Notebook LM, from the expert. And now we can try to design our own personal protocol based on this research. And then after that, we’re going to create experiments which we’re going to review every morning. And to make sure that we are actually doing them, we’re going to put the events in the calendar. All right, so let’s do a health assessment interview from those six Q&As. So, let’s pick I would say the most impactful questions across several areas. Okay, that’s actually very good. That’s actually very good. That’s actually very good. So, right now I’m going to answer those and we’re going to come back with the experiments. We’re going to try to prioritize the experiments. Okay, so I actually submitted the response and right now I told to go and look up my fitness experiments and get the data from there. That’s the power of Obsidian and logging your data. You’re going to tell oh, go grab this data and it’s going to like have it without me even explaining it. And they’re currently getting our health profile and you can see okay, it’s the current state and also the target state. How can we get better? And there is a gap assessment where it’s high, medium or low. Okay, and it’s actually like very very good assessment of myself, I would say. I would say it exactly regarding the going for a quick walk to get the sunlight in the morning, getting more cardio into my program. For stress management, you can improve meditation, make it more frequent. The biomarkers, I hadn’t really thought about getting them, but here is a good insight. You can actually it’s unknown and if you don’t know, that’s the worst. You can actually get and do the blood work. And here is the assessment. Yeah. And right now, the proposed top three highest leverage experiments. We have a sleep regularity. That’s the biggest gap. You can see it from here. Now, there is another experiment is morning sunlight. And then there is third experiment is adding zone two cardio. Can you show me what kind of experiments you’re going to do? Present a plan to me. So, I will ask LLM to present a plan to me to be aligned so that it doesn’t do weird things but I am always in control. Yeah, you can just present plan here. Okay, so here we have our hypothesis. We are using scientific method. We are using scientific method and the success is okay, 80% of the days within 30 minutes of target wake up time. And there is another experiment morning sunlight and here is actual protocol. Wake up, open window, go outside for a 10 minute, go grab coffee. And what I’m going to do here is actually put into our daily daily tracking those fields. Okay, so let’s create those experiments as a file. Yes. And here is my daily note. I have here my goals, experiments and we’re going to write down and write here. Let’s actually test that. So, that should appear here as in progress. So, make them in progress. Right now, it creates experiments which are proposed. I just want to right away make them in progress so we can see them right here and review them every morning. Okay, I’m fixing the front matter and right now, finally those experiments are here in the progress. So, you can see them. Here is our hypothesis, here is our protocol and here are the observations which we’re going to be filling out each morning. Now, let’s say if I want to start my morning, I have a skill which is called daily and it has my morning routine. That’s my workflow for the morning and it checks my goals. It checks my experiments. Okay, and here is the actual active experiments for the morning check-in. And those one which we are created just now. And here is the core thing that now we are getting asked about okay, about the morning sunlight, about the caffeine delay, about the gym, about wake up time. How is it all going? And in the morning, you can answer those questions and then Claude going to write them to this file for you and you can keep track of your experiments this way. And now from there, we can actually okay, schedule our gym, schedule our, let’s say, morning walk to get and execute on those experiments. And this is the whole protocol in action. We had this assessments based on the interview. We designed our experiments. We embedded them into our morning routine and we scheduled the task throughout our day to actually do those experiments. And now you can keep track of them in your health dashboard. Let’s say I have this dashboard and here we can have our goals. Here we have our experiments which we can track here. Now, and what you just saw, we just turned research into experiments in just one session. And if you want to build this yourself with a group, I’m running Claude Code and Obsidian Lab. You can come with your own goal and you leave with a working system that runs experiments every morning. Link in the description. Now, let’s move on. And my vision for this is treat Notebook LM as external knowledge across experts, across domains. If you were into product management, there is Lenny’s podcast. I actually did the same for it and I just uploaded 200 sources, 200 podcasts and you can chat with them in the same way. Now, let’s say okay, if you want to learn about something else, you can together with Claude in the chat try to like maybe search for YouTube channels and filter the videos which are relevant to your goal which you want to achieve. That’s going to serve as this context for Claude Code as a source of verifiable truth. And Claude Code can read any notebook and can actually help you with closing the loop where we move on from the learning to actually planning, acting and reviewing our results. And most people never leave learn. And I believe that’s a really as the highest leverage right now which you can do is absorb any expert knowledge and act on it the same day. And that’s the best way to learn, actually by doing something. Now, and I want you to try it yourself. I share Notebook LM skill in the description so you can just grab it and install it in the next 10 minutes and you’ll be ready to go. And I’ll see you on the next one.