The Art of Becoming Self-Educated with Claude

0:00 If you want to be able to learn any skill, workflow, or concept dangerously fast so that you can be the expert in the room, you have to master self-education. Now, this isn't just for

0:09 people in school. The framework I'll cover is specifically designed for professionals who are ready to fight intellectual obesity, learn new skills, and stay upto-date with AI. So, in this

0:18 video, we're going to use Claw to help us answer six simple questions that will take you from beginner or knowing nothing about a topic to an expert in a specific domain. And I know these work

0:27 because I've used this strategy to teach hundreds of thousands of people on YouTube. I've used it directly with executives ranging from 25 to 60 years old. And it's inspired by the exact

0:36 framework that surgeons use to learn topics when studying about medicine. But before we get to those six questions, we have to first establish why this is so important and define what we mean by

0:46 fast. When I say fast, I don't mean skipping the learning. I mean compressing it. Yes, AI can give you answers instantly, but if you never build a foundational understanding, you

0:55 just become an AI middleman with no value of your own. Andrea Kpathy said this best. You can outsource your thinking, but you can't outsource your understanding.

1:03 >> So, the real competitive advantage isn't removing learning entirely. It's the ability to use AI tools to self-educate faster and immediately apply those learnings. And to master this process,

1:13 it all comes down to asking six simple questions in this exact order. So, let's get into question one. Why are you even learning something? The first mistake I see is people aren't strategic about

1:22 what they choose to learn. They pick a topic because it's trending, not because they need it or because they're even interested in it. So ultimately, who cares how fast you learn something if it

1:31 doesn't provide any real value to your life? We want to work smarter, not harder. So before learning anything, understand your why. Why are you learning it? Are you learning something

1:40 because you're intellectually curious or because it's professionally actionable? Those are the two different buckets that all these topics should fall into. And if you're intellectually curious, great.

1:49 That's enough to make it worth it. You don't have to be a hardo where everything has to be about your career. And if you're unable to bucket a learning into something, it starts

1:57 because you're intellectually curious. And then over time, as you learn the basics, it may become professionally actionable. So let's just take an example of a topic I'm learning,

2:06 optimizing conversion funnels. This is professionally actionable. I'm not genuinely intellectually curious about this. And I identify this because it's a constraint in my business,

2:14 buildpartner.ai. and for us to grow, I need to solve this. So, because it's professionally actionable and it is a high lever opportunity, it makes sense for me to learn it. And a key part of

2:24 this is identifying the constraints in your day-to-day. And if you're not sure what they are, run this prompt, which will go through the process called the five W's. Essentially, what this does is

2:33 it will ask you why about five times to really get to the heart of the problem. And for any topic where you aren't sure how it applies to your life, you're intellectually curious until you've

2:43 realized it can be professionally actionable. This is critical to understand because it determines the level of depth that we want to dive into for that specific topic. Which brings us

2:51 to the second question. What level of understanding do I need? If it's a topic you're just curious about, just go until you feel like you've satisfied that itch. But for professionally actionable

3:00 topics, we have to have the end goal in mind. This is a process called destinationbased learning. You want to know what the end goal is and work backwards from that. And if you're

3:09 working with something that's constraintdriven, that is the end goal. You want to solve that constraint. And so you work backwards from that. Ultimately, the goal is to find a

3:16 Goldilock zone where we learn just the right amount to solve the constraint in front of us, but not too much where we're learning unnecessary information. This Goldilock zone is how we make sure

3:25 that we're moving fast with anything we learn. So no matter what, for both intellectually curious as well as professionally actionable, we have to establish a foundational understanding

3:34 on any topic as quick as possible. So to do that, we want to climb up the foundational learning ladder, which has three steps. The first step is we tell Claude, explain it like I'm 5, 12, 18.

3:44 This will force AI to simplify it so that you can get a foundational understanding that you can then later build on. And along this lines, one thing that I love doing is having AI

3:53 explain something to me in sports references. So on screen, you can actually see I was learning about design interfaces and I said, can you explain this like I'm 16 years old with a

4:03 basketball example. It makes the learning experience fun and it helps me get to that level one understanding. Then once you get that foundation, you want to go to the next level where

4:11 you'll say, "Give me level two analysis on this topic and why it's important to me." It's important to remember the big picture in terms of why you're learning something. Then once you get that level

4:20 two understanding, a little bit deeper about your situation, you then want to go to the next level, which is explain it based on what I'm already working on and what I already understand. Interview

4:29 me if any context needed to answer this is missing. This will take the topic and center it around your specific environment so you can start to see how you'll actually apply this in your

4:37 day-to-day. And the reason that I love these last two steps of this ladder is that you're leveraging AI's competitive advantage when it comes to teaching you. That's that AI is a stateful tutor, not

4:46 a stateless tutor. To understand what I mean by stateless or stateful, think of a substitute teacher versus a full-time teacher. The substitute teacher doesn't know anything about you, what your

4:55 learning habits are, and it just provides generic information. Whereas the full-time teacher, they know more information about you and how you best learn. AI is stateful. It knows

5:03 information about you. So, it can tailor the information about what you're working on. And if you want to enhance Claude so it becomes a better teacher for any subject, I'd run this prompt

5:12 that interviews you and creates a life story file. And when it interviews you, instead of typing your responses, use the voice mode. Just click the mic button and then speak to it like you're

5:21 getting interviewed. Now, at this point, you've established a foundational understanding of the topic you're learning. So, if this was a topic you are intellectually curious about, great.

5:28 That's pretty much all the advice I have. But for this video, we're going to focus on the professional actionable topics. And for those topics, there are four specific questions that you have to

5:38 answer. But before we get to that, the beauty of AI is that it has removed all of the barriers and the walled gardens to learn any skill or really how to build anything. Which is why when

5:46 Anthropic, the team behind Claude, decided to sponsor this video, I wanted to tell a quick story about how Claude has removed the barriers to build. Now, if you're on this channel, you know how

5:55 much I love Claude Code and I use it every day. But not only has it transformed my life, it has also transformed the lives of the people around me. So rewind about six months

6:02 ago, I started working with my buddy Nick who runs a thriving football media brand. Nick is not an engineer. He is a football influencer, but he had an idea and he wanted to build it. We started

6:12 working together and once we onboarded him to Claude Code, his eyes just lit up. He said, "I used to be unable to build things that I always wanted to, but now I can just build things myself.

6:22 The second he got hands-on, building became second nature. And with the football season fully upon us, there is a lot we need to build to get ready. So Claude has made it super easy. And so

6:31 for one feature, here's exactly how we built it. We told Claude, "Enter interview me to make draft rankings guide for 2026." It goes and asks me questions about the product. And then I

6:40 follow up, "Make sure to follow our brand guidelines." And I then link a design guide Claude had already helped us make. On screen, you can see the design guide and what it looks like.

6:48 This created a V1 product, but visually wasn't what I wanted. So I then say, "Make me three design variations to show inspiration I can choose from." I then like the third, so I choose that. I then

6:58 go back and forth having it adjust the design. And once it's done, I tell Claude, "Launch three sub aents to review the code. One is a senior engineer, the other is a security audit,

7:07 and the third is a user experience expert. Flag any issues." Once they come back, I have a product I'm happy with, and it's time to deploy it. Now, that's a fantasy guide that over 10,000 people

7:15 will use throughout the course of the year. And this process is simple. And so, if you're building something for one person or 10,000 different people, it doesn't really matter. The point isn't

7:24 that AI made it easy. It's that it made it possible for somebody who has never going to touch code before. In my comments the other day, I was going back and forth with a subscriber and I ended

7:33 up saying, "The best project is your first project." So, in sponsorship with Anthropic, which I'm absolutely hyped about. Let me know in the comments what you're building. I'd love to help as

7:42 many as you that I can. Removing all those barriers is so powerful, but when learning any new topic, the fastest way to become an expert is by learning from people who have already solved this

7:51 problem. Which brings us to question three. Who has solved this before? This is all about identifying someone in the real world who has already solved this problem or is an expert in the specific

8:00 domain. In the second question, we got the textbook answer. But in this section, we want to get the answer from a mentor who solved it in the real world. And the reason we're doing this

8:08 goes back to the speed component of this video. We're looking to compress our learning so that we can learn enough to provide value. By having this private coach, it'll streamline the time it

8:17 takes to getting a understanding that is deep enough to provide concrete real world value. That's the differencemaker and why this question is so important. So the process that we're going to do is

8:26 what I call alpha farming. We want to get valuable context from industry experts about the specific subject we're learning about. And the alphaing process can be broken down into two parts. The

8:36 first is you need to curate your resources. Find someone who is an expert in the space biasing towards people who have an online presence. So there's reference data that you can use to train

8:44 AI. Then once you've identified the person, you want to identify resources that are specific to this topic to have AI help you identify a short list of people and then go ahead and find

8:52 resources about the topic. Run this prompt. And if you're looking for a tool that can help you search the internet, I use a tool called Firecrawl which is linked below. The second step once

8:59 you've made that list of people and you've identified the resources is you need to make it super easy to ask questions where you can create a notebook and just bring in every

9:07 resource related to that specific subject from the experts that we identified in the first step. The simplest thing you can do here is use a free tool called Notebook LM where you

9:15 can create a new notebook and then just bring in all of the resources that you just gathered from the previous step. You're essentially creating a hub of information that you can ask questions

9:22 about and it'll only pull from the resources you'd gathered. This is a good way of removing all the other bloat from the AI training data and instead focusing on the context that you

9:32 provided. Our goal here is not getting the textbook answer, instead getting the lived experience answer from the people you brought in. Now, that's the simple approach. Frankly, that isn't what I do.

9:42 I prefer creating an LLM knowledge base directly in clawed code where instead of adding the files to notebook LM, I just take those files and I add it to my claw project. On screen is a prompt that will

9:52 help you make that LLM knowledge base. But if you want to dive into that, I have a video on my channel with over 800,000 views that I'll link below as well. Now, Notebook LM or the LM

10:01 knowledge base will be fine for what we're doing in this video. But if you'd rather not go down this rabbit hole of farming experts and setting up your whole system, my build partner.ai AI

10:09 plugin has a skill/expert advice which you can reference and it'll essentially skip this entire step because it's pre-trained with expert data. At this point, we know we're learning the right

10:19 thing and we have a foundational understanding with experts who will help guide us in the right direction. All of that is great, but it's worthless unless we answer this next question. How do I

10:29 make it concrete? And that's question number four. So, my brother's a plastic surgeon, and I'm always asking him, how does he learn so many things? And yes, he's a surgeon. I am a YouTuber. So very

10:39 different careers. And so in learning there's a philosophy. You want to see one, do one, and then teach one. See one. That is the basics. That's the foundational understanding about the

10:48 topic. And we've already done that in this video. But that is not enough to have a concrete understanding of a subject. You have to do one and then you have to teach one. That's how you bridge

10:57 abstract concepts that you've learned to concrete outcomes that will engrave these concepts into your brain. So let's start by doing one. How can you apply whatever you just learned immediately to

11:07 your situation? Because the fastest learners are the ones who shorten their feedback cycles and can immediately apply what they learned. So to do this, go ask Claude, "Wherever you are

11:16 learning about the new topic, how can I apply this directly in my project and what is the simplest V1 I could make?" The key word I want to address there is simplest. You want to walk before you

11:26 run so you can build up momentum in the specific domain you're learning. Now, if there is a clear constraint that you're solving, you can say, "I'm looking to solve this constraint. How can I

11:34 immediately apply this to my situation? So, this will give you a guide for how to apply it and just start doing things. Build the landing page, make the email marketing copy. It's not that serious.

11:44 Just build and learn. That is the most important thing in this whole video. And I know the natural next question because this is what most people ask. Why don't I just skip to this step? Claude can

11:53 just do the thing for me. I actually had this exact conversation with two of my buddies the other day. One is a lead engineer at a fintech and the other is a firefighter who wants to learn about the

12:02 subject. And the firefighter goes, "Why is coding even valuable if you can just have prompts do things?" And our answer was, "Prompting is trivial, understanding isn't. When AI creates a

12:11 game plan or does something for you, you have to be the one that says, "Oh, that's good." Or, "Oh, that's not good." You have to be the one that's directing it. And if you skip what we covered in

12:19 steps two and three in this video, you don't have the foundational understanding to give it the direction that you want it to go. So, if you skip right to the step, you're just more or

12:27 less becoming the AI middleman, and you're not providing any real value. So remember, you want to see one, do one, teach one. So make sure you see one. Then we just covered doing it. What

12:36 about teaching one? So why even do this? The goal here is to identify any gaps in your knowledge. And this will help us strengthen our understanding, which is needed to answer the last two questions

12:46 that I'll cover in this video. So ideally, if you could teach the topic to a human, that's great. But the reality is most of you won't be able to do that. So what I tell people is run the mic

12:54 test. Turn the mic on in your Claude setup and then explain the concept out loud like you're teaching it to someone. Then you can have Claude ask you questions about what you just explained.

13:01 And so on screen you'll see a prompt that will kickstart the mic test and at the bottom you can add the voice note. I know everyone watching this has had that experience where you're sharing someone

13:10 information or you're making a proposal and a person asks a fairly simple question that you never thought to ask yourself and you feel a bit blindsided. So what this will do is help you get

13:19 ahead of that and get a deeper understanding. These are all processes for streamlining the learning process. But for a professionally actionable skill, in order for you to make a real

13:27 impact, you likely need to integrate and work with other people or other systems within your organization. That's how you take these lessons and then you scale them to make a realworld impact. And

13:36 we're going to cover that in the next part of this video. But if this is your first video of mine, welcome to the channel. But if it is your second or more, you know the drill. This is our

13:44 anti-slap agreement. The visuals, everything that's going on here, this was made by humans for humans. So, all that I ask as part of this agreement is you subscribe to help this content reach

13:53 more people so I can keep making videos like this. We're about 14K subs away from 100K. So, first off, shout out everyone who's already subscribing. I may not be a surgeon like my brother,

14:03 but this makes my mom proud. Also, every video I give away a CloudMax subscription as a thank you. So, this week's winner is Holly Jeff, who's building a tool to help teach producers

14:13 how to use AI. Absolutely goated. Now, to enter the next giveaway, comment below with whatever you're learning or a recent topic that you learned about. It doesn't have to be related AI. Just let

14:22 me know what you're learning. Getting to question five, how can you apply this to your actual situation? We covered C1, do one, teach one in the last section. So, at this point, you've created a simple

14:31 V1 of whatever you're learning. But if you want to provide that real impact in your organization, you have to understand how other people are doing things and how it's currently set up

14:41 because rarely you're starting from scratch. That's just the reality of the situation. and that information your AI setup doesn't necessarily have. So, for example, from my business, when I link

14:50 something to my products in the description of my YouTube videos, I don't use normal tracking URLs with UTMs. Instead, I route it differently because I think the UTMs are ugly

14:59 visually. On screen on the left, you can see how it's typically done, and then on the right is how I do it. So, if you were hired by me to help with marketing and you were like, I'm going to learn

15:08 about UTMs and tracking, you would throw all of that learning out the window because my business does it differently. generic UTM tracking advice for URLs is wrong for my business, but right for

15:17 everyone else. And now, this is just one example, but there are thousands of examples like this wherever you're working or whatever you're working on. And the best people are the ones who can

15:25 dissect how things currently work and apply improvements to that situation. So, how do you actually do this? Well, the key is understanding the baseline of how it's done today. One of my favorite

15:35 ways to do this is have Claude draw a mermaid diagram, which is a flowchart, to map out the current lay of the land. This is ideal for team settings where you have to get an understanding of how

15:44 something's currently set up, but it also works if it's just you because it helps you better understand the whole situation. Especially with AI, a lot of the times it's not exactly what you

15:53 think it is. So, here is a prompt that will help you map out a process into a mermaid diagram and it'll help you map out an entire process end to end. And then once you diagram, challenge every

16:03 step. Does this agree with what you thought it was doing? Is this actually how it works today? This is something I do with some of my clients who run their own businesses and they're amazed when

16:11 they realize they don't know how their own systems work. And this process is establishing a baseline so we can better understand how to apply our new learning. Then you can take that

16:20 finalized diagram and share it with Claude to see what can be done today to solve the constraint. Using that baseline you've established and your depth of understanding that you created

16:28 in the first four questions that we covered, you're now able to take what you learned, adjust it, and then apply it in a way that actually works. One of my biggest pet peeves is when people

16:38 start suggesting improvements without knowing the current baseline. That's how you instantly lose credibility. You're a bull in a china shop. Instead of doing that, be a Navy Seal. Be precise and

16:50 exact. Now, even when you do all that, the reality is you won't become an expert overnight. We focused on compressing the time it takes to learn a specific topic, but it doesn't remove

16:59 time. When you follow all the steps here, you'll get to a level you initially needed to solve a problem. But often you'll realize that you actually have to go deeper. And question six is

17:08 about how do I keep learning? The more you learn about a topic, you'll realize the less you actually know. Let's take email marketing for example. You learn the basics and then you realize subject

17:16 lines, the body, the images within it. There's so many different verticals that you have to learn. And so the reality is that each layer of depth that you want to go, that opens up a new learning loop

17:25 where you can go through the first five questions in this video. Is this worth learning? Why are you learning it? Maybe the level of understanding is good enough. If it is worth learning, what

17:33 level of understanding do I now need to get to? Who can help me learn this quicker? Who are the experts in the field? How do I make these learnings concrete? Applying the see one, do one,

17:42 and teach one method. How do I then take that and apply it to my actual situation? If you do these steps and then repeat, that is the learning loop. Now, AI has compressed the time, but it

17:52 can't remove the work. And you've likely heard the concept of 10,000 hours by Malcolm Gladwell. essentially takes 10,000 hours to become an expert. But I think that's outdated. AI when used

18:02 properly has compressed the time it takes to become an expert because you're able to learn the right things at the right time. Practice does not make perfect. Perfect practice makes perfect.

18:12 So what we learned today is the way to direct and orient ourselves so we're learning the right things at the right time at the right level of depth. Now everything that I covered here gives you

18:22 the exact playbook to learn anything quickly. But you may ask, "What should I even learn?" Well, to solve that problem, I break down the four clawed projects you should build instead of

18:32 watching any more tutorials. A key part in this video is about doing, and those four are curated subjects that if you got this far in this video, I know that you're going to get a ton of value from

18:42 that video. So, I'll see you over there. Peace.