# Actually, everything is a wrapper…

## Метаданные

- **Канал:** Varun Mayya
- **YouTube:** https://www.youtube.com/watch?v=hRKlffMezxQ
- **Дата:** 30.05.2025
- **Длительность:** 27:38
- **Просмотры:** 313,691
- **Источник:** https://ekstraktznaniy.ru/video/11985

## Описание

Ever wonder why people say “It’s just a wrapper” when someone builds something new on top of an existing tool? It’s not really their fault, because our education system never taught us how the real world of tech and business works. In this video, we break down what “wrappers” actually are, why they’re not a bad thing, and how a lot of real innovation happens only this way.

We’ll look at how this broken way of thinking started, how to unlearn it, and the best skills you need to learn to give yourself a head start in the post-AI era.

And to show you exactly how powerful those skills can be, we’ll use an AI PC powered by the Intel® Core™ Ultra Processors to build some cool stuff right in front of you. These new AI PCs let you run AI models locally, so you can experiment without relying on cloud services. If you’re someone who wants to build cool new stuff with AI, this is the kind of machine that makes the process smoother, faster, and way more fun.

Learn more about Intel® Core™ Ultra 

## Транскрипт

### Intro []

Okay, so you're told to get a degree, learn to code, memorize facts, and follow the rules. So far, that's supposed to be the formula for success, right? Well, I'm not so sure. You see, while millions of students in India are still being trained the same old way we were trained 30 years ago, I think AI is quietly changing the rules of the game. To understand just how badly we are misaligned, let's talk about coffee.

### Wrappers [0:29]

See, you walk into a Starbucks, you buy a latte, and you drink it. Seems pretty simple, right? But beneath that simplicity lies one of the most intricate, globally coordinated systems in human history. And it's very relevant to how you should think about life and careers. Think about it. A coffee store doesn't really need to grow coffee beans, milk cows, or build coffee machines. It simply needs to wrap everything underneath into a simple product you can buy. That's because coffee stores like Starbucks depend on a large network of suppliers who move these coffee beans from one place to another. Starbucks may make coffee beans today, but many other coffee stores rely on other people and other institutions. Now, these suppliers themselves don't grow coffee beans either. They also depend on something more complex. A massive shipment network that moves goods across countries and continents. Go one step lower and you'll find local factories that manufacture, process, and package these goods before they're even ready to ship. But even these factories don't produce anything from scratch. They depend on something much deeper, the people growing these raw materials. Now, let's call this entire system a stack. Each layer in the stack depends on the one below it. Starbucks can't sell coffee without its supply chain. The supply chain can't function without shipping. Shipping can't exist without factories. And none of it would be possible without the people at the bottom. And yet, when you order coffee, you don't think about any of this. That's because by design, Starbucks has to wrap all the complexity into one simple transaction. And that's exactly what a wrapper is. It takes something complicated, hides all the layers beneath it, and it turns it into something easy to use. And it's also very important for the way you think about life and careers. Now, let's turn to software. Kids these days see something built on top of existing software and start laughing and they say, "Oh, this is just a rapper. " As if it somehow makes it less valuable. As if React, Python or C++ aren't rappers themselves. Learning a specific language or a framework at a very specific abstraction level and believing that you have figured out everything and dismissing everything below or above the abstraction level is not a good sign of deeper understanding. It's actually a surface level take. Think about it. No one cancels Starbucks for not growing its own coffee beans. No one looks at Uber and says, "Oh, that's just a wrapper around car owners. " And yet, when it comes to software, people suddenly forget how value is created. This is the kind of status game that misunderstands how the economy works. Because in reality, value doesn't really come from originality, but how accessible, simple, and useful something is. How much you are abstracting away the complexity is important. But this mindset around cancelelling rappers isn't just an internet opinion. It's a symptom of a much deeper problem. And it didn't just come out of nowhere. It started in our schools. We were told that copying is bad and that being original is everything. For a decade, there were countries we looked at and we said, "Hey, these people only make copies of everything. " Until the point where they started making better things than the people they were copying. The building on top of someone else's idea was somehow cheating or immoral. Even though the entire economy is that. So, of course, now when people see other people building on top of something existing, some existing tool or some existing model, they roll their eyes and dismiss it as just a rapper. And honestly, it's not even their fault. They were systematically trained to think that way. Instead of encouraging creativity by using the things that are available to you and modifying it in many ways, schools told us to memorize facts, follow instructions, chase grades, and never pushed for deeper understanding. Our understanding of understanding is a very viva-like understanding of understanding which means that you go you're able to answer a bunch of questions like a rapid fire or quiz round but you're not actually able to build anything and because of it our generation neither learned to grasp nuance and complexity nor did it even attempt to. Now this mindset would have worked 20 years ago though when you couldn't just Google anything and memorizing did give you an edge. If you couldn't Google something and no one else had that information but it was in your head, then well, you did have an advantage. You knew something other people didn't and that information was important. But the times have changed. Information is freely available on the internet now. Specific information for you from an AI model for the exact problem that you have in a few milliseconds. And therefore, we need to change the mindset of what is valuable and important too. But actually, it's always been like this. Sure, while knowledge has been valuable, I think abstracting, which is moving away complexity beneath the surface and giving one quality product in one single transaction to a customer has always been important. So before we go any further, let's try and understand why it was always important because the system that actually encodes this and that system is

### Capitalism [5:12]

capitalism. So my viewpoint around this is that capitalism is just a stack of rappers. See, capitalism isn't actually about hard work or making money. I know you're going to scream at me. You're going to be like, "No, but capitalism is all about money. " No, not really. Think about your phone. Every time you send a text, you're not worrying about signals bouncing between cell towers or data packets traveling through optical fibers across continents. You just type and hit send. That's because your phone itself is a wrapper. Not just around a bunch of technology from different places, but also around the entire telecommunications infrastructure. Not just your phone. Everything around you is a rapper. Your car is a wrapper around engines. Netflix broadcasting networks. And even Google search is a wrapper around the internet. Now, each of these products has a stack of its own. So, let's zoom out. Imagine not just one but millions of such stacks all interconnected and relying on each other to make trade, markets, and entire industries possible. That is what capitalism really is. When you make a cup of coffee, when you use, you know, one of those coffee mixes, you have no idea who the people are who actually grew the beans, and they don't know who you are, a lot of parts in the shipping chain either. But they all do their jobs without sometimes knowing who the end customer is going to be. But schools never explained it this way. Instead, they made it sound like this grand abstract machine that runs on its own. It actually runs on all of us. We run it. It's human nature and behavior and our ability to coordinate that run it. But in reality, it's just rappers all the way down. And the truth is, capitalism wouldn't exist without such rappers. But I think a lot of people don't understand how the real world works. So they oversimplify. They see billionaires and they think it's either pure luck or exploitation. But what they don't see is the invisible architecture. They don't see that people who build billion-dollar businesses aren't the ones memorizing facts or doing the foundational work. They're actually the ones building the biggest rappers because there's actually no such thing as foundational work. Everything, even the foundational work, depends on other foundations. And as a result, the world rewards them massively. And it only makes sense. See, these people who you think are making a lot of money, actually the people on top, the rapper most test of rappers, right? What's the difference between them and the layers below? They are more exposed to the general consumer. The players on top of the stack are B2C businesses, which is they work with consumers directly. They take the most risk. They handle consumers saying good things about them, bad things about them, talking about them on Reddit, all that stuff. They absorb all of this. They wrap the complexity of the layers below and turn it into something easy, usable, and accessible to the end consumer. And because of it, they create the most value for themselves and capture the most money. But they also absorb the most risk. Because if you play one level down, let's say you're a provider for one of these consumer businesses, you're a B2B business. Let's say you're AWS and you're just providing cloud services. It is a much lower risk business. You may not capture as much value as like the biggest mobile phone business, but you're still critical and you're protected from a lot of the risk of going straight to consumer. The lower down the rapper you go, let's say some person who's shipping uh you know actual hardware devices from one place to another that the servers can use, that person is most protected from the consumer. They only have to keep one customer or a bunch of customers happy. But the topmost rappers don't just create the most value or capture the most money. They also keep the stack alive. Think about it. Consumers pay Starbucks. Starbucks pays suppliers. Suppliers pay shippers. Shippers pay manufacturers. And manufacturers pay farmers. At every step, money trickles down the stack, incentivizing every layer to do what they do best, but also reducing the risk. The higher up you go, there's risk. The lower lesser risk. And by risk, I mean exposure to direct consumers. There's a lot of business owners that say this, right? Which is exposure to the markets directly is like exposure to psychopathy. You have no idea what they're going to wake up and think because it's all driven by vibes and emotions. Unlike a farmer who simply grows coffee or a shipper who simply ships goods, Starbucks has much more at stake. One bad quarter and all of the news will be talking about how their stock has dipped. A failed product launch and lots of employees will quit. But this isn't anything new. This is how progress has always been. Every time technology has advanced, people have stepped up, taken risks, and found a way to build something on top of existing technologies and created the most value. And the reason I'm telling you all of this is because after AI, all of this becomes even more

### AI [9:36]

important. Now, remember I told you that every layer is dependent on the layer below it. But actually, every layer is also dependent on the number of people you can hire and all of their knowledge intelligence and all of their experience. Most people think AI is some unstoppable force that will take over everything. But really, it's just another layer in the stack that wraps complexity and makes things easier to use. AI itself is a rapper. If you think about how AI works, it's only replacing people who are not very good at what they do. But the people who are reasonably good at what they do are actually using AI to do a lot more than they could traditionally do. You can do programming, you can do design, you can do writing. You're better at everything as a go. Instead of now hiring five people at your layer and being able to provide some service to somebody, you're now able to use just these tools. It's kind of changing how capitalism works. See, a lot of people feel that using AI for something is cheating because it feels like you're using somebody else's work. It feels like using a system that's making all of the things, all the craft that you the effort that you spend putting into something, it kind of makes it useless. It feels like you you're not really doing anything. But if you've noticed the world around you a little bit, you realize that society has always worked that way. In fact, society has only rewarded people who built something useful. Whether it was Facebook copying MySpace and Friendster or Google building its search engine on top of the open web, society has only rewarded them. Use what society gives you. Add more value and make that convenient for the end consumer to transact with. We always had a finance system. Why did UPI work? It just made it so convenient. No, it wrapped on top of such a complex system that we already had. If you're a software engineer and you think the specific part of the stack that you work on is the most important stack and anyone building on top is a just a wrapper builder. Let me give you an example. Imagine I dropped you on an island and asked you to build an app from scratch. Let's say you love using React. So, first you need React. But React can't do anything since it's just a wrapper around JavaScript. So, you need JavaScript first. But JavaScript itself runs on engines like V8 which are written on top of C++. And C++ is built on top of C. To make matters worse, C doesn't run by itself. It's compiled down to assembly. And assembly is also just a wrapper around raw machine code. Let's assume you somehow figured out on that island how to code in binary. Well, wait. You don't even have a computer. So now you need to build one. And for that you need a processor, which means you need to extract silicon from sand. But you don't even have electricity or a factory. So maybe you need to start by rubbing two stones together and creating fire first. Let's also somehow assume you managed to build a computer from scratch. What about the operating system? You need to write an operating system from scratch. But how are you going to write that when you don't even have a keyboard or a screen? Think about what an awesome world we live in that all these abstractions exist. Nobody starts from scratch. Everyone is standing on thousands of years of human invention. And funny enough, nobody writes assembly today. But I wrote it in college. But even then I was relying on a computer built by somebody else. And that's the thing, every modern tool exists because of countless layers of inventions that came before it. Not just physical inventions, but a lot of knowledge that's in people's brains. While the inventions already exist in the real world, and now we can buy them for cheap because commoditization has happened. And you can buy a computer for really cheap compared to what it cost 30 years ago. The knowledge is available for free with AI or nearly free with AI. And that's what people miss. The underlying work doesn't disappear. It's still there. It's just that rappers take care of the lower level stuff so you can focus on the higher level creative stuff. And if you feel AI has taken over some level or some layer, there are layers on top. The transaction is the final layer. Every time a new rapper has emerged, human effort has shifted towards using the entire stack to build something new. And it's the same thing with AI. At least right now, at least in the next 5 to 10 years. Instead of starting everything from scratch, you can just tell AI what you need and it does it for you. Instead of writing each line of code by hand, you just describe what you want and AI writes it for you. It's not really cheating. It's a technology that you now have access to that a lot of entrepreneurs have had access to for the last 10, 20 years. If an entrepreneur wants to build something, he can always hire somebody and make the thing happen. Now you also have the ability to do that. So you are now competing with that entrepreneur with much less

### Ultimate Wrapper [13:55]

cost. Now I'm going to tell you about the ultimate rapper. You think the ultimate rapper is some SAS page or some product that people can buy and then you know people just make a transaction and they get it and that's the final rapper the finished product. I think there's a bigger rapper. I think you're missing the bigger picture. Something that I only started seeing 2 or 3 years ago right I had an idea that it would be important but in the last 2 3 years I became very convinced about it and it's worked for us. College fails to teach you the one skill that matters most at the highest levels. It's coordination. They glorify the idea of the solo genius who builds everything from scratch, but that is a total myth. I've now met some of the best entrepreneurs and founders in the world. That's a myth. That's not how the real world works. The higher you go, it's less about doing everything yourself and more about orchestrating people, systems, and resources and making sure things get done. And that's the real game. The more rappers exist, the more pieces there are to coordinate. And the higher you go in the stack, the more valuable coordination becomes. Take my example. What do I do? If you ask some communities, they'll say, "Well, Vun does a lot of stuff in video. " If you ask some other communities, they'll be like, "What does a lot of stuff from Genai? " If you ask a very different set of communities, they like, "He does some stuff in gamedev. " If you ask a very different set of community, he's an entrepreneur. He runs a company with x number of people. So, it's just everyone has their own view of what I do. And a lot of people do not believe it's possible to do four or five things at once. But that's because they see me on screen every day and they see one person. They do not realize that AOS as a company is now 300 people and at AOS we're probably doing like 40 or 50 projects at once. So we may put out X project tomorrow but Y project day after. It could be a different team working on these projects. Now because I'm the platform and the channels we have are the platform and it may go out as a real on my channel. People think ah Vun's doing thing number 55. But the truth is it's actually a team behind it and this is what every big company eventually becomes. But the core value is because I have this audience, I'm able to coordinate how these products and services go. Take the example of Google and Facebook. It's so funny that Google and Facebook have sort of rebranded themselves as these really high-tech research companies when a lot of their revenue actually and early success came from advertising. It actually came from pure coordination. A person wants to sell a flower or has a flower or a boutique store for flowers. They want to reach customers. They go on Google. they put out adwords. The biggest rapper where the most amount of money is made in the ecosystem in the economy is in coordination. That's why content creators these days are starting to kind of make the kind of wealth that a lot of entrepreneurs are, right? It seems strange and people don't understand why. They're like they're just talking. Yes, they are, but the good ones are also spending a lot of time coordinating from talent to projects to teams to actual products to advertising services for people who want to advertise things. They are coordinators. The simple truth is you don't make money by knowing things. You make money by applying what you know and fixing problems. But India is a vivadriven culture where it's all about the quiz this thing of his knowledge knowledge. Can you answer this question and then there's an answer that doesn't work right and it's never worked historically which is also why I have so much respect for the sales and marketing people. India as a country thinks sales and marketing is like ah saleserson marketing person it's not real jobs or look down upon it saying sales people don't build anything and think that only writing code designing products or manufacturing something is real work while sales and marketing are secondary but if you think about it sales and marketing isn't just coordination between products and customers it's actually the only difference between a product that changes the world versus a product that gets forgotten people ask me why I hype things so much without that it would all die. Not just here, but in almost every company in the world. If the company stops advertising, marketing, or going out there and getting in front of people's faces, all the people who work at the company will probably lose their jobs. People assume that if a product is good, it should sell itself. But reality is different. You can have the most advanced tech, but if no one understands it, trusts it, or want it, it simply doesn't matter. Take Intel for example. It was Marcian Ted Hoff, an engineer at Intel, that invented the architecture for the first commercial microprocessor, the Intel 4004. But then it was Andy Grove, the company CEO at that time, who turned it into a product that the world could understand, trust, and want. I often say this, right? If you tomorrow came up with a cure for cancer, your biggest challenge is most people wouldn't believe you. You need somebody to package and advertise it to the rest of the world properly first. But if coordination and rappers are so important, why do many people look down on AI rappers today? Why is there so much hate, so much push back, so much judgment? Why do people not like sales and marketing even though it's like the DNA of the entire economy? If every big invention is just another layer in the stack, why does everyone have a problem with AI but not with any other tech? Well, two

### Problems [18:53]

problems. Status and fear is the first one. See, most people grew up thinking software engineering was the ultimate career path. I remember in 2021, fresh grads were making 20 to 30 lakhs, which is between $25,000 and $37,000. If you're watching this from outside of India, these were salaries at top tech firms. But now that AI is flooding the market with clever new automations, these companies are questioning whether they even need as many engineers in the first place. If you look at the hiring patterns of some of the larger companies in India who have generally mass-hired software engineers their hiring activity has grinded to a halt. It may not be as much about AI it may be a lot more to do with you know in general the demand for software is saturating like there's a lot of software already built for the same kind of use cases. So people are not being creative about new use cases but also it's the fact that the top software engineers have gone wild. They're on steroids. The thought of why hire 10 developers when one AI powered software engineer with a tool can generate code, debug errors, optimize performance, really think through what should be built in the first place in a fraction of the time. And that's a hard pill to swallow. Many engineers grew up building their entire identities around their technical skills, whether it be writing code or being efficient or solving theoretically complex problems. But now that this technology is here and it's able to solve some of those problems, the world has changed. And I keep saying this that the real test of intelligence is not a specific skill that you've learned. That actually comes from repetition. The more you repeat something, the more times you do something, the easier it gets to do it the n plus time plus you're generalizing some patterns. But the real test of intelligence is your ability to adapt. If you're feeling threatened, which a lot of or see me included, sometimes when I use a tool, I feel threatened. I'm like, "Oh my god. " But you shouldn't be in a place where you don't know what else to do. And it is understandable that the only way to move forward is to defend their identity they grew up building. You can't expect to change this overnight unless you've trained for it. As an entrepreneur in my career, I've had to pivot many times as the market changed. See, as an entrepreneur, it's different from being in a job. In a job, you can do the same thing for 10 years, it'll be fine because if the company dies, you can move to the next company. If you don't like a specific job, you can move to the next job. You can be in the same field. But as an entrepreneur, if you do something and then a Facebook or Google do the same thing, you really have no choice but to pivot, right? So you have adaptability has to be part of your brain. Even when people sometimes shout at you for adapting, you need to do it. It's kind of like now what an entrepreneur has gone through for maybe 10 20 years of their life of being able to navigate the market as the market itself changes. Now everybody needs to do that. Turns out most people are still attached to how things used to be done. But as I said, the times have changed. Plus, it's not really new. We've seen this before. We've read countless stories of the past. In the 1400s, people fought against the printing press. Then in the 1800s they fought against factory machines. Then in the late 1900s they mocked the internet and thought it would fail. Even the newspapers were calling the internet a scam. Now it's AI's turn. People dismiss AI rappers as not real engineering or claim that true skill is in knowing the fundamentals. But if you look at history, just knowing the basics has never been enough. It's the ability, like we've mentioned throughout this video, it's the ability to put things together that create value. And that's where the second problem comes in. The education system itself. See, it's like we're just not taught how to articulate what we want. One of the biggest problems with our education system is that it doesn't teach us how to think in systems and forces us to memorize concepts instead. And as a result, people struggle to understand complexities and fail to articulate ideas clearly. What does articulation mean? We spoke about coordination, right? What's the key skill in coordination? It's being able to articulate. In Gen Z speak, it is yapping. And it is yapping to get the computer to do exactly what you want, not just yapping random things. Articulation is the ability to explain your thoughts and ideas clearly. So not just other people can understand it, so can the computer. In this case, the other it's AI. I don't know if you know, but over 27% of Indians, that's over 400 million people can't comprehend a single sentence. When people say, "Oh, why do you make things so simple these days? " Or why do you try to explain overexlain things in simple fashion? It's because we have now realized how wide India is and how few people can finish a sentence or even finish like a video. Even if it's a short form video, a lot of people don't finish it. They can read the words, but for some reason they struggle with deeper understanding. They're reading the words, but they're not able to make sense of the sentence. And that's the problem. See, when you use AI, you're not just pressing a button. That's not how AI works. You're articulating and coordinating different ideas, models, prompts, files, apps, tools, copying from one another, putting your own context in there, pasting something from somewhere to another, tweaking something here, adjusting something there. You don't like the output, you want a different output. Sometimes you're manually going in and fixing it. It's not about knowing what each idea is individually, but knowing how to use them together to get something done. And that's why clear articulation and coordination are now the most important skills.

### Articulation & Coordination [23:45]

Now, I'll give you a very specific example of what this means because this new skill of being able to prompt, engineer, think, articulate, have some context about a specific field. I'm just going to show you how it works. Okay? Now, we've got this AI enabled Intel PC right here and we're going to show you how we are going to be able to do some cool stuff with an LLM. Let's create something practical. So, let's say I'm a teacher and I'm going to make an educational app that visually and intuitively explains how Einstein's space-time continuum actually works. Now, you may have heard about this concept. It might feel confusing, but let's show you. So, we're going to use this tool called cursor to build our app. It's an ID and there's an AI there, and I'm going to tell the AI what I want. So, we're going to say something like create an interactive 3D website, showing a slightly tilted green space-time fabric on a black background. When users click a button, objects appear and warp the fabric between them beneath them, simulating gravity-like curvatures with realtime physics. So now what this is going to do is install all the necessary packages and start the project for us. As you can see, this AI tool just created a functional and interactive website to play with. I just click add planet and it adds a planet for me onto the fabric which I can move around and see the space-time distortions in real time. Not just this, I can now add multiple planets and see them interact too. Now this is where articulation becomes absolutely critical. When we first asked it to create a curvature beneath the planet, it got it slightly wrong. The AI placed the curvature based on the planet's initial drop position, not its current center, which made the distortion look off. The warping effect didn't align visually because the center of the curvature wasn't directly beneath planet. But here's the thing, it wasn't really AI's fault. It was ours. We weren't precise enough in our instruction. Thought we were clear, but we weren't really that articulate. And that difference matters. The moment we told the AI exactly what we wanted, that the center of the curvature should always track the center of the planet in real time, it gave us the correct code instantly. And this wasn't the only time articulation proved to be very important. There was another moment. This time, the AI gave me the visuals just fine, but when I moved an object with my cursor, it kept passing through the fabric, which obviously shouldn't happen. Again, not the AI's fault. It was mine. I hadn't told it exactly what I wanted. What I should have said was, lock the X-axis. So the objects stay on the fabric and don't pass through it. And the moment I said that, it worked. The AI fixed it instantly. And that's the real shift here. It's not knowing how to code every single thing. It's about communicating your ideas with clarity and if something goes wrong, being able to drag it in into the way you want and fix it. It almost like air bending from Avatar show, right? Like where you're bending this air to do what you want. You're not really creating air or fire or any of the uh, you know, elements. You're actually moving them in the direction you want. It's like being a sailor in a boat where you don't really control the storm or the water, but you can use this or that you have to direct where it goes. And it's certainly a new skill. I just want to quickly call out the partner of this video. As you just saw, this Intel AIP PC makes it easier than ever to build powerful applications within hours. All you need is the ability to articulate ideas, orchestrate those ideas, and this Intel PC. And the best part is this PC doesn't even need an internet connection to use AI models since its hardware is powerful enough to run LLMs locally. That means you can build, create, and learn anytime and use many different kinds of local models. And for parents who want their children to really have a good tool in the post AI age, this is the perfect way to give them a head start and make sure they feel confident in their careers. Cuz education today, as you all know, isn't really about learning very specific concepts in isolation. It's about integration. It's about learning how all of these ideas connect, how the rappers work, how the stack works, how the layers are, you know, in tandem with each other, and knowing that knowledge is just one part of the ecosystem. What you need is application. So, good luck and bye.