How to Learn Anything Faster with AI and Look Naturally Gifted

In this video, we'll break down my 13 top tips for learning anything faster with AI. Let's begin. Tip number one: think of AI as your personal tutor. One of the most underrated uses of AI is not just using it for search queries or doing your job. This is using AI as the world's best personal tutor. And since one of the factors that brings satisfaction to life is learning and development, we don't want to simply translate our thinking into AI. We actually want to use AI to learn more things. Because if we can learn while creating something, then life becomes much more interesting. And we are also developing our brains, instead of letting them atrophy due to over-reliance on AI. Honestly, thanks to the power of modern AI, you can learn as much in

one weekend as it would probably take you an entire semester in college. And importantly, you can do it with pleasure. And do it while creating a cool project. Tip number two: start with a specific project. Whether it's learning programming or anything else in business, I'm not a fan of learning for the sake of learning. Because if you're not really interested in the subject, studying for the sake of studying ends up feeling like a "must." Ultimately it feels like a duty. In addition, information is less easily absorbed if we cannot connect it to real-world experiences. That's why I'm so convinced that to learn anything, you should simply start with a project.

The project we'll use as an example in this video: Over the weekend, I built a local home security system with cameras and a mini PC that I installed Linux on-it took a while, but I figured it out. And I connected the cameras to the home security system. So now we have cameras in the house, and instead of broadcasting the signal over the internet, where theoretically someone could snoop, they transmit the data to a local system that is n't even connected to the network. And it can run local AI models to analyze video recordings without the need for an internet connection. Before starting this project over the weekend, I knew absolutely nothing about all of these things. I knew nothing about Linux. I've heard of it, of course, but I've never installed it. I

knew nothing about how security cameras worked I didn't even know it was possible to run local AI models on a cheap computer. But while working on this home improvement project, because my wife asked me to. I was able to learn so much about Linux, cameras, networks, SSH, security protocols, the history of all these things, and it was just incredibly interesting. It was a really enjoyable and energetic way to spend a Saturday afternoon, and it opened my eyes to a whole world of computers that I had never even imagined before. So if you're trying to learn something quickly with AI, my advice is to start with a project, something you're trying to create, build, or are personally interested in. If you are, for example, a student and you are only doing this

because you need to do well in exams, then that is fine too. I have many videos about effective learning methods, like spaced repetition, active recall, and the like, that I made in my time. But for the rest of the video, I'm going to assume that there's something that's important for you to learn, not just to get a high score on an exam, but because you're genuinely curious and want to figure it out. Tip number three: use voice, not typing. Okay, so we have a project. Now we just need to talk to the AI to set it up to teach us interesting things while working on the project. In the context of this particular home security system, I happened to watch

a video by a guy named Network Chuck, who is one of my favorite YouTubers. In this video he talked about Linux, about a thing called Frigate, about the possibility of using some kind of camera protocol to connect video to Linux and Frigate, whatever that means. I watched this video a few months ago, so it was in my head, "Okay, I know this is definitely possible." So we can just open up any AI we like and just dictate all of this to it. I currently use Whisper Flow for dictation because I don't want to type, but all AIs actually have. So, either use something like Whisper Flow, or just use the voice feature built into the AI. I'll show you exactly what my request looked like so you can see that there's nothing. In short, I want to create some kind of

home security system. I heard that you can install Linux on some device. I heard there's something called Frigate. I heard that you can use some kind of protocol to connect cameras to this device because I watched Network Chuck's video about it. And this is a cool home improvement project that I want to do because my wife told me about it. But along the way, I'd like to learn some interesting things about Linux and how it all works, because, you know, I know a little bit about computers, but I've never worked with Linux. So, yeah, can you help me figure this out? What's up with that? So, super basic query. I didn't use any query engineering. I literally talk to him like he's a tutor or a friend who's incredibly smart. In any case,

let's see what happens. OK. Wow, that's too much information. I don't like it when AI gives me a bunch of things. Oh, listen, buddy, this is too much. Can you give me information one item at a time, and I'll let you know when I'm ready to move on to the next one, please? Oh, that's better. Much easier to learn. So, the camera sends video over your home network to your computer. Frigate. What the hell is this Frigate? Works on this computer, watches and records videos. That's all. So, you need two things. You need a camera and a computer running Frigate. We'll ignore Linux, Docker, RTSP, POE, AI, networking, and everything else for now. When you're

ready, say "next" and I'll give you exactly one more piece of information. Class. This is it, this is what I need. Wait. What the hell is this Frigate? Okay, Frigate is essentially software that turns your computer into a smart You install it on your computer, connect your cameras, and Frigate can show you live feeds, record videos, detect people, cars, dogs, packages, and save important events. Wow, that's cool. So, Frigate is software for security cameras. And then I'll explain exactly which computer Frigate actually runs on. So, I could just continue on or, you know, follow my curiosity and explore some offshoot. So, I'm going to follow my curiosity and explore a new direction. Tip number four: follow your curiosity. Okay, reading this made me genuinely

curious. Here, here, here is the question I have. Wait a minute, are you seriously telling me that there is free software created by someone else that literally makes a home security system for me? This, this seems kind of strange. How, how does this even work? Yes, that's exactly what I'm telling you. Frigate is open source software. A group of developers created it, published the source code, and anyone can use it for free. It's a bit surprising that the expensive-looking thing a security company sells you often consists of nothing more than a video stream, a recording program, detection software, and a handy app.

Frigate provides you with a software layer. He doesn't magically create an entire system out of nothing. You still have to provide a computer, cameras, storage, and network. But once you do, Frigate takes over a significant portion of the security system's work. Perfectly. Now I'm following my curiosity again. Wait a minute, can you explain what Linux actually is and why it's interesting? Okay, it's an operating system, it's open source. Hardware, operating system, software. Why is Linux interesting? Because it immerses you deeper into how computers work. Class. On the Mac, Apple, of course, hides most of the internal mechanisms. Linux allows you to freely explore everything

from the inside. Wait, you can have a tiny computer sitting quietly in a closet without a monitor, keyboard, or mouse that runs your house 24/7, and you control it remotely? That sounds cool. Here you see that the goal is a home security project, but along the way I ask these questions, simply because I'm curious, and ChatGPT, Claude, Gemini, Grok, or whatever your favorite AI model is these days, is incredibly good at communicating in your language and explaining things in a fun way. By the way, if you're anything like me, you probably spend a significant portion of your day. Emails, messages, AI prompts, notes to yourself, social media content, whatever it is. And this is where the sponsor of this video-WhisperFlow-comes to the rescue.

WhisperFlow is a great voice-to- text tool that integrates into any app on your computer or phone. You just say whatever you want, and it automatically inserts the text, corrects typos, and formats everything properly. So, for example, if I'm writing an email, it might look like this. Hey Nicole, we have three questions about social media this week. First, can we postpone the release of the reels until Thursday? Wait until Friday, because the video comes out on Thursday. Secondly, can you collect the September metrics for Instagram and TikTok, whichever is faster? And third, send me the publication schedule for October as soon as it's ready.

Nicely formatted, turns everything into a numbered list. I can send it via email, Slack message, or even WhatsApp if I wanted to. I didn't touch the keyboard, it's much faster than typing, even though I type very fast, speaking is still faster. And it automatically got rid of errors like " Thursday, not Friday". He just takes it all into account and does it incredibly quickly. This is magic. Some people say, " But my phone or computer already has built-in dictation." To which I will answer: have you seen how terrible the built-in dictation works on your device? Hey Angus, can we have a team meeting at 3? Wait, it's actually 4 o'clock, and I'm coming from West Kowloon, so I might be a little late. So, keep that in mind. And yes, it's just absolutely

awful. He left all those "uhs" and "mms." Instead of "wait" he wrote "Waite", which is the last name of one of my friends from university. He doesn't know that West Kowloon is a place here in Hong Kong that WhisperFlow knows about, but Apple's built-in dictation doesn't. So, if you do even a little typing, you're better off just using WhisperFlow. WhisperFlow is free to use, and if you want to use it all the time, you can upgrade to WhisperFlow Pro, which gives you unlimited words. The video description will contain a link and code that will get you 1 month absolutely free. So you can experience it to the fullest.

Thanks to WhisperFlow for sponsoring this video, and we're back on topic. Tip number five: create visuals. Of course, this is a video about how to learn anything faster with AI, not about the ins and outs of Linux or home security systems. So I want to share a few principles of effective evidence-based learning that we can apply to learning anything with AI so that it is truly learned and remembered. There are several different ways to make complex things clear to you, depending on the context. One of my favorites is that instead of asking the chat to just tell stories, I ask it to create visualizations. Okay, I still don't quite understand how this works.

Can you turn this into an HTML visualization or something like that will help me see how the whole system interacts? Okay, great. A simple visual explainer. Let's see. Oh, of course. How a home security system works. The camera sends video to a small computer. You have cameras, a home network, a mini PC. The entire system is on one screen. Cameras, home network. You know, I don't really like it. I like it better the way Claude does it. Can you make it more interactive so I can click on charts etc.? Do you understand? Okay, let's see.

Show the flow chart. Yes, it's a little crooked, but you see, I can click on the camera. Okay, this sends the video to the home network, to a mini PC, on Linux. Frigate-it's an app, and then, oh, I watch it through the app. So, by simply asking for a visualization, I immediately began to Our brain often perceives visual information better than text, so it makes learning much easier. You can almost always get Claude, ChatGPT, Gemini, Grok, or any other tool to turn the material into an interesting visualization that helps you understand everything. Tip number six: tie it to a fictional story.

One of my favorite ways to understand something is to simply ask the AI to relate what I'm trying to learn to the world of Harry Potter or the universe of Brandon Sanderson, because those are two fantasy worlds that I know very well. So let's find something a little more complicated than that. How does Linux actually work? What about Ubuntu, Debian, and others? What is this latest news on Twitter about Arch? Explain all this to me using analogies from the world of Harry Potter, etc. Again, there is nothing special about this request. I'm not doing some, you know, 2023- level prompt engineering. Linux is magic itself, and Debian, Ubuntu, Arch are different schools of wizardry that teach and structure their magic differently. Arch

is essentially a very bold, stylish Hogwarts faculty built on Arch. Now schools of magic are beginning to appear. Debian is like an ancient, serious Hogwarts. Lots of traditions, a huge community, a conservative approach to change, and extreme attention to making sure everything works reliably. Debian has huge software repositories and very strict release processes. So, Debian is Hogwarts, founded 1000 years ago. Ubuntu is Hogwarts, but someone has put up signs everywhere, updated the student handbook, modernized the dorms, and hired a

customer support department. And what's cool about it is how it becomes relevant to your damn project. Your little server is essentially a brownie. So let's say to him, "Dobby, show me your vault." Dobby replies, "Dobby, I have a camera working." Dobby, install this So now I've instantly connected this historical thing about Linux or whatever else you want to do to a world that I'm very familiar with, and I can understand it through analogy. Using analogies, stories, metaphors, etc. again helps us not only understand things, but also remember them. Now I'll definitely never forget the difference between Debian and Ubuntu, having a visual metaphor: " Oh, sure, it's like Hogwarts, but with its own

peculiarities in my imagination." Here's how to learn anything faster with AI. Tip number seven: look for drama. Another point about learning anything is that if there is an emotional charge to what you are learning, it is remembered much better than just a dry fact. Dry facts are boring, but stories are interesting. And what makes a story interesting? Well, it's stakes and drama. We love drama. And so, to follow up on this particular example, can you tell me about any interesting controversies or any interesting recent drama around all of this that might be interesting? Let's take a look. This points to perhaps Linux's biggest meta-drama. Linux proponents cannot agree on how Linux should work, which is both its greatest weakness and its greatest strength.

One camp says, " Make it polished, standardized, and simple enough that grandma can use it." "The other camp says": standardization inevitably means someone gains control. Give me the details and I'll decide. "Another one says : I don't care about either of you." My server hasn't been rebooted since 2023. "And then someone who uses NixOS or whatever comes out of the basement and says: actually, the entire Hogwarts configuration should be described declaratively in a reproducible file." "And everyone stops inviting them to parties. Steeply. Again, this is interesting. Tip number eight: seek out opinions, preferably controversial ones. I'm currently reading a book called "The World: A Brief Introduction" by a certain reason I'm

reading it is because the other day I literally told AI that I wanted to better understand how the world works, because I realized: there's a lot going on in the news, and I have no idea what's going on. And I feel like I don't have a holistic understanding of the world, politics, and geopolitics. Sure, I know a little bit here and there, but because I went to medical school right after school, I never had any proper classes, except for when I was 13, about history, politics, or economics. And so, to really understand the world, you need to have a fairly solid knowledge base about what's going on. So, I told the AI all this, and he recommended this book to me-"The World: A Brief Introduction" by this guy, Richard Haass. And I started

reading it. I say this, firstly, because you can learn anything by simply asking AI to recommend resources. But, secondly, there was a moment that I particularly remembered, related to NATO, the North Atlantic Treaty Organization. You've probably heard of NATO, a kind of alliance between the United States and a number of Western countries, aligned against Russia. Regarding NATO, I knew roughly what I learned in Cold War history class when I was 14: NATO was created during the Cold War as a kind of created the Warsaw Pact as a kind of response to NATO. That was my understanding of what NATO was about: it was like an alliance where an attack on one country meant an attack on all the others, and they came to each other's defense. But what I didn't

realize, and what he writes about in the book, is that actually at the end of the Cold War, somewhere between 1989 and 1991, depending on who you ask, the Russian threat suddenly seemed to disappear-at least for that time. So NATO as an organization didn't have anything to focus on in particular, and so they started expanding, bringing more countries into NATO as member states and so on. This moment stuck in my memory because Richard Haas, the author of the book, shares a controversial opinion. His controversial view is that NATO should have been dissolved in 1991, when the Russian threat disappeared, and he believes that in fact the expansion of NATO membership since 1991 was one of the factors that led Putin to say "to

hell with Europe and the West", leading to far more drama than if NATO had been dissolved. They could find better ways to engage with Russia, and he controversially believes the world would be a better place if NATO had not expanded since 1991. This is interesting. The author expressed the opinion that this idea is controversial, and therefore it is inherently interesting, because what seems dramatic and interesting is imprinted in our memory. So whatever you're trying to study, if you can find drama, interesting or controversial points, or points where historians or politicians disagree with each other, and ask AI to give you the " Man of Steel" arguments for each side of the debate, you can get a pretty clear picture of a lot of different

issues. Because, especially in the news, almost everything interesting has two sides to the debate. We can argue all we want about which side is better, but, especially when it comes to education, especially if it's something controversial, you want to be able to really present the opposing side's best arguments, right? If you think the Labour Party is much better than the Conservative Party or something else, you should understand why so many people vote Conservative? What are their best arguments for this? And then-what is the best response to these best arguments? It used to be difficult because you had to search on Google, maybe you saw something on Reddit, or in a blog post, and you had to trust the author. But today, since AI has essentially read the entire

internet, you can just ask it to argue with itself. At least for me, it gave a much deeper understanding of anything, just by asking AI to first find controversial points, and then strengthen the arguments of both sides. Tip nine: trust AI, but with some caution. Now all of this may raise the question: yes, but can artificial intelligence be trusted? What if he's A- hallucinating and just making up facts? Or some say that certain AIs have been trained or tuned by system prompts to a certain political position, so neutrality should not be expected from them. So, one problem is, wouldn't he just lie to you? And the other is, might he have a bias that you don't notice?

I think both of these points are quite valid. In my opinion, you shouldn't worry too much about either of these aspects -it all depends on how important the specific issue is. So, for example, if I'm visiting a random museum in Japan and I'm curious about what the conflict was 1,500 years ago between different dynasties or empires-it's not a big deal. I don't care if he makes up a few little things. I don't care if he has a somewhat anti-Western, pro-Western, or pro-Japanese stance -it doesn't really matter. Overall, he still teaches me quite sensible things about the world. For the most part, AI is getting better every day at not hallucinating or randomly making up facts. So many in the " hallucination" camp are focusing on the

free version of ChatGPT 2023, rather than modern paid advanced models that have a much better grasp of the facts. And I think the argument about political bias and trust also makes sense. This is also quite logical. That's why I like to ask him to reinforce the arguments of both sides, because he usually does a very good job of that. They say Claude leans a bit to the left, but if you ask him to give the strongest right-wing arguments, say against abortion or against immigration, he will also give a pretty solid answer. So you might say, "Okay, now I understand why people

think that." "I may not agree with it, but I see the logic behind why people think that." If it's something like this, sometimes I just ask Grok. Grok is known to be a bit more relaxed. And so if I'm really interested in what Twitter is saying about a certain thing, because Twitter has strong views on both sides of the political divide, I'll just ask Grok. But again, it depends on the stakes. If you're like me, and mostly just an average person trying to figure things out, the stakes aren't that high. If, for example, I were an assistant to a state leader and I was preparing a briefing for him about Crimea, Ukraine, Russia, or anything else, then, of course, the stakes would be much higher, and I would definitely check my sources. But for most of us,

most of the time, I don't think it's worth worrying about, especially when the stakes are low. Tip number 10: Let AI correct you. The best way to learn anything is a technique I've been talking about for years called active recall. All the research ever conducted on how to learn puts this method at the top. This basically means you are checking yourself. You learn not by just reading something, but by reading it first and then trying to recall it later or simultaneously. Active recall is often combined with something called spaced repetition, which seems to counteract the forgetting curve, where you forget a lot of

information very quickly. So, you test yourself on the same day. Then you check yourself in a day, then a week, and then a month. And with this combination of spaced repetition and active recall, you can learn almost anything you want, very quickly . I've been talking about this on this channel since 2018. Now, when people approach me on the street, they often say that, firstly, they watched my videos, and secondly, they watched them when they were at school, college, or university. And that it was learning techniques such as active recall and spaced repetition that helped them pass their exams. If you're watching this video right now, I'd love to know in the comments if this is true for you too. But these two techniques, active recall and spaced

repetition, are by far the best ways to learn anything quickly. And the great thing about AI is that you can ask it to help you with that. So, active recall is essentially when you give yourself a test, and there are a few different ways to do that. The first one, which I really like, is to simply state what you're thinking and let the AI correct you. So, for example, I'll start a new chat, and let's say I want to learn something random, like the history of banking, just why not? Maybe I'm visiting a museum or something. That's why it's now worth interacting with artificial intelligence when visiting museums, making them much more interesting than I thought before.

Okay, so I want to learn about the history of banking. What I'm going to do is this: I'm going to lay out my rough understanding of the issue, and then I want you to correct me and explain where I'm wrong. Okay, so the history of banking. When do you think the first bank was created? I don't know, around 1000 BC, 500 BC? What kind of empire was this? Oh my god, are we talking about the ancient Greeks? Damn it, when were coins invented? I vaguely remember a few thousand years BC, maybe. But, as soon as coins appeared, Ancient Greece was probably the main one in this regard. They probably had banks in the Roman Empire. Like in Game of Thrones, they have the Iron Bank of

Braavos, and it's kind of like medieval. So, banks definitely existed, right? And how did the banks work? Did the crown keep the banks under control? When did the Bank of England come into being? Why banks? OK. What do you say? Obviously you can see that I don't know much about the history of banking, but I'll just throw this at the AI and see what it responds with and how it corrects me. Your intuition is actually quite good. The main thing to correct is that banking is much older than coins. Oh, cool. Banking began as bookkeeping, and coin lending came later. So, the mental model "as soon as people got coins, they got banks" is perfectly logical, but the cause-and-effect relationship is almost

inverse. The first banks were probably temples and palaces. Let 's go back to Mesopotamia, around 3000-2000 BC. Steeply. Coins appeared much later than you think. Coins did not exist for several thousand years BC. The first widespread.Oh, 7th century BC. OK. Especially in Lydia, in what is now western Turkey. So, writing and debts appeared in 3000 BC. e. Lending arose before 2000 BC. e., and coins only around 600 BC. e. It's a huge gap that reveals something profound. Money as debt and accounting are older

than money as coins. Interesting. That's also a bit too long. If I were actually doing this, I 'd say, "Let me make it easier because I don't like the huge arrays of text that ChatGPT produces," but never mind. And now I have a banking history. As you can see, in this context I'm practicing a bit of active recall, straining my brain to remember what I know about banking. And I'm completely calm if I make a mistake, because it's just AI, I'm not filming it for the public. And then the AI can correct me, which is cool because it tells me where my assumptions were right and where they weren't, and that's a much better way to learn than just asking it to tell me the history of banking. Tip number 11: Let AI give you a quiz. Again, continuing

our theme of active recall, this can be done while studying anything. Okay, great. So, I read about the history of banking. I want you to make up some multiple-choice questions for me so I can test myself. Okay, let's go. Coins, paper credit and record keeping, central banks and paper money. Okay, well, obviously it's B. I'll just copy and paste that. Where did some of the earliest banking transactions occur? I think they said it was in Mesopotamia. What was one of the main roles of temples and palaces in early Mesopotamian finance? I think it was this. Approximately when did Easy, 700 BC. e. With what place is this closely associated? They said-Lydia, which is in western Turkey.

Ancient Greek bankers were called trapezites. What did they usually do? Probably this. Which statement best describes the Roman banking system? Oh, damn. Me, I stopped reading at this point. These multiple-choice questions are a bit tricky because you can often guess the answer from the wording. And so on. Oh, 11 out of 11 correct answers so far. You stop at 11, so you're back at the 12 to 20 stage. So you can do it like this. You can ask him to do this one question at a time. You can do it all in one go. You can even set it so that it doesn't offer answer options. For example, you might say, "Okay, I want you to ask me one question at a time that tests my

understanding, but without multiple choice options." Just ask me a question, and I'll tell you what I think the answer is, then you correct me and ask the next one. Why could a complex banking system exist before the invention of coins? Because coins came quite late, and as you told me, it seems that wheat, barley, and silver were used as currency long before coins appeared. Exactly. Great, next question. What problem did medieval bills of exchange solve.Yes, you get the idea. We use the process of active recall. In the past, before AI, you would have to either create your own flashcards to memorize, which takes forever, or

download someone else's flashcards that aren't tailored to what you're personally interested in, unless you specifically.What is this? Mosquito. Unless you are specifically preparing for some standardized test that a lot of people take, like USMLE, NEET, or something else. Tip number 12: Ask AI for help if you get stuck. So, going back to my home security project over the weekend, I was working on these things. I asked AI to make a shopping list that I needed. I ordered them on HKTV Mall, which is like Amazon in Hong Kong. I waited for the things to arrive. I unpacked them. And in the process, I took pictures of everything and

asked AI to help me with what to do next. For example, I unpacked the PC and hard drives and connected everything. And then I realized that I did n't have a monitor to connect it to. So, I asked the AI, and guess what it told me? He said I have a monitor. I was actually able to use my TV because the AI remembered that I had this particular TV. So I found an HDMI cable, connected it to the TV, plugged in the keyboard, and was now able to get started without any problems. And while this computer was connected to the TV with my old keyboard somewhere in the playroom, I was like, "Okay, great." He is working. Windows is configured. What should I do next?

Install Linux or what? How do I make it so I don't have to keep it connected to this TV forever? I would like to manage it without a monitor via SSH, perhaps using Tailscale. What do you think? And so he gave me a recommendation. So now I'm setting everything up. "How do I restart Windows using only the keyboard?" Windows key + R, then type this: shutdown /r/o/t 0. Boom. And then I needed to install Linux, but I didn't have a USB flash drive, because who uses those these days? And I'm like, "Hmm, I don't have a flash drive. Can I use the hard drive that I have, a spare hard drive?" And he gave me

clear instructions on how to use the spare hard drive. During the process, one of the Mac programs used to burn ISO files for OS installation crashed for some reason, I simply told the artificial intelligence about it, and it gave me terminal commands that could replace the program's functionality. I wouldn't have been able to figure this out via Google before. I would just say, "Oh, damn it. I'll have to wait another day for the flash drive to arrive from Amazon or somewhere else." So, simply following the AI's instructions and asking for clarification on things I didn't understand, I plugged the drive into my computer, was eventually able to install everything, and then booted into Windows in recovery mode, something I hadn't done in decades. And then

this black background and all that stuff appeared on the screen. So, as I go through this, I take photos, send them to the AI, and ask it to suggest what to do next. This is the type of technical difficulty that always arises when you do things like this. So literally every step of the way, I have no idea what's happening on the screen. I just take a photo, send it to AI, and ask for help. And just by following this process for a few hours on a Saturday afternoon while my wife was napping and my baby was sleeping, I was finally able to install Linux on this little box, connect and format this hard drive that now records video from security cameras, install Ubuntu Server, then a thing called Docker, then a thing called Frigate, and finally connect

cameras to it to create a home security system. And now it's just great. We have a system that works. I'm really excited to experiment with it. We have set up the basic version of the camera. I learned a lot about Linux, cameras, security protocols, SSH, and all that kind of stuff during this process. And now my next project that I want to do is find a way to, you know, actually, you know what? I'll do it right now. So, I have a home lab on a Linux server, like in my bedroom. I would really like to automate the opening of our bedroom curtains at 7:00 every morning. How do I motorize my curtains and have them automatically open at 7:00

every morning? Ah, Home Assistant, Linux server, okay. SwitchBot Curtain 3. These are upgraded engines. You attach them to the cornice you already have. Oh, cool. This is cool. Okay, not bad. So, this will be my next home lab project. And again, working on this particular project, it would be cool to open up, damn it. And then, as I work on this particular project -it would be cool to automatically open our curtains at 7:00 AM, so that would be a nice alarm clock-I learn more about the hardware. You may have to do a little work with a soldering iron. Maybe I'll find out how it works and what it is. And it's really cool to have these little projects that you can do at work or in your personal life, and learn interesting things while working on

them. And importantly, if you get stuck on the road, just take a photo, send it to artificial intelligence and ask it to help you. So now there are no excuses why you can't do cool things and learn interesting things at the same time. And finally, tip number 13: know when to stop. Of course, when it comes to learning, there is an endless rabbit hole that one can dive down without end. My barometer for when to stop is simply following my own curiosity. At a certain point, if I'm in a museum and I learn about a coin, why it's important, why it's dramatic, interesting, and controversial, I say to myself: okay, cool. I feel like I've learned enough and I'm moving on to the next one. So, the goal of learning is not to make it feel like work. It should

feel like a game. This should be enjoyable. And if you're not having fun, find a way to make it interesting or just stop, because when you enjoy the process, you learn much faster and better than when you force yourself to do it. Education should not be an obligation. Once you get outside the school or university system, where it's a kind of obligation. You have to want to learn. And it's a bit like reading, is n't it? Many people force themselves to think that they should read more books. But one of my favorite quotes is from Naval Ravikant: read what you love until you love reading.

Reading doesn't have to be a routine. Just like learning shouldn't be a burden. You can totally read and study what you are truly interested in. To start reading, you can take up, I don't know, Harry Potter fanfiction if you're into that, or romantic fantasy. And once you enjoy the art of reading, you can of course move on to something more difficult, like Charles Dickens, classics, non-fiction, or whatever. If you focus on enjoying the process, then learning will come naturally. Reading will also come naturally, and your life in general will become much better. If you enjoyed this video and want to dive deeper into how to use Claude Code, here is my beginner's guide, which many have said has changed their lives and helped them

get started coding with AI, especially if they didn't know how to code before or were afraid of the technical aspects, considering themselves not to be programmers. So this video is right here. Thank you very much for watching, see you there. See you later.