저자: TickerSymbol:YOU진행자Alex · 외국인
요약 보기
엔비디아 CEO 젠슨 황의 BG2 팟캐스트 발언을 인용해 AI가 직원 생산성을 높여 기업 이익을 키운다고 설명. 직원당 AI 에이전트 투자로 마진이 개선된다는 계산을 제시하고, 글로벌 AI 시장이 2030년까지 크게 성장하며 AI 인프라(GPU, 파운드리, 메모리 등) 수요가 늘어난다고 주장. 소재 과학·특허 증가 등 AI의 파급 효과도 언급. 영상 중간 스폰서 광고(AI 교육 플랫폼) 포함.
다듬은 글
A few days ago, Nvidia CEO Jensen Hong went on the BG2 Pod, which is one of my favorite podcasts for investing and advanced technology. But what I thought would be a pretty cookie cutter interview, ended up being over 90 minutes of Jensen dropping non-stop bombshells and revealing Nvidia's entire AI playbook.
Would I do it? Heartbeat. I I'm doing it across every single person in our company right now. Right. So, in this video, I want to share some of the most important takeaways for investors because AI is about to make a lot of people very rich, including me and you. , this video is a little different from my normal style.
I'm going to show you a few quick clips from Jensen's interview and then add some more context to help us find the biggest winners of the AI revolution. I'll also leave a link to the full interview, which every serious AI investor needs to listen to because it's jam-packed with value, and value is the perfect place to start.
In this first clip, Jensen explains the real value unlocked by AI and why every single company is investing in it so aggressively. Suppose I were to hire a $100,000 employee and I augmented that $100,000 employee with a $10,000 AI. Yes. And that $10,000 AI as a result made the $100,000 employee twice more productive, three times more productive.
Would I do it? Heartbeat. I I'm doing it across every single person in our company right now. Right. Every single co- aents. That's right. Every That's right. Every single software engineer, every single chip designer in our company already has AIS working with them. 100% coverage. Yeah, as a result, we're growing faster as a company.
As a result, we're hiring more people. Our productivity is greater. Our top line's greater. Our profitability is greater. What's not to love about that? Now, apply the NVIDIA story to the world's GDP. Yeah. So, right away, we have a few big takeaways to talk about. First, AI isn't going to replace most jobs.
It's going to augment them. We can think about AI sort of like a utility, like the internet, electricity, and gas. We measure AI in tokens. And the goal is to use those tokens to do something productive, just like with the internet, electricity, or gas. Second, let's quickly break down Jensen's math because the economics of AI are central to the investment thesis for it.
When companies hire somebody, they're hoping to get at least a 2x return on that employees total costs, including things like benefits, training, computers, equipment, and so on. If a company hires somebody for $100,000 a year, after all expenses, they're hoping to make at least $200,000 a year from that person's work.
Usually way more, but we're being super conservative here. An employee that costs $100,000, but generates $200,000 of value has a profit margin of 50%. Jensen said that an AI agent that costs $10,000 a year can make an employee two to three times more productive. But we shouldn't just take his word for it.
I mean, come on, just look at those glasses. So, let's cut his numbers in half and say that an AI agent can make an employee just 50% more productive by doing only the most obvious things we all already know that an AI can do today. Basic research, writing first drafts of documents, formatting emails and reports, taking notes in meetings, answering basic questions about articles that we already don't read.
So, no chip design, and no serious software engineering. Now, with that AI agent, that same employee costs $110,000 a year, but generates $300,000 worth of value instead of $200, which is a profit margin of 63%. Or set another way, the company made an extra $100,000 for a $10,000 investment, even when we use very conservative numbers.
Oh, yeah. And who says that one AI agent can't help two employees instead of one or three or 10? Charlie Munger famously said, "Show me the incentive and I'll show you the outcome." Well, when the incentive is a 10, 20, or even 30x return per AI agent, the outcome is crystal clear. Any company not investing in AI will eventually be disrupted by one that is.
But if you want to hear something really scary, 2026 is less than 100 days away. And one of the best investments you can make to get ahead is understanding AI. That's where Outskll comes in, the sponsor of this video. Outskll is the world's first AI focused education platform. And they're hosting a 2-day AI training to take you from beginner to advanced AI professional this Saturday and Sunday.
And they're giving the first 10,000 people who sign up with my link a free seat. Whether you work in tech or finance, sales, or HR, you'll learn to use 10 of the most powerful AI tools like Gemini and Groc, best practices for prompting AI and using AI agents, and even how to automate workflows without any coding.
You'll also get free access to Outskll's exclusive community and learning dashboard to keep growing alongside other AI professionals. No wonder over 10 million people from all over the world have already attended, and the limited slots for this one are filling up faster than ever. This is a great way to gain a competitive advantage with AI, get a serious head start for 2026, and understand the science behind the stocks.
So, make sure to register for your free seat with my link below today. All right. So, companies have a huge economic incentive to adopt AI because it makes their employees much more productive, which translates directly to much higher profits, even when we use conservative numbers. But at the end of that clip, Jensen said the same math could be applied to the whole world's GDP.
And if we do that, we can see how big the global AI market can get over the next few years, which would tell us how fast it could grow from its size today and what kind of returns we could expect on AI stocks as a result. So, here's the next part of that clip. Now, apply the Nvidia story to the world's GDP.
Yeah. And so, what's likely to happen is that that $50 trillion is augmented by, let's pick a number, $10 trillion. that 10 trillion dollars needs to run on a machine. Now the reason that AI is different than it in the past in a way software was written a priori and then it runs on a CPU and it doesn't it runs it a person would operate it
in the future of course AI is generating tokens but a machine has to generate the tokens and it's thinking so that software is running all the time whereas in the past the software was written once now the software is in fact writing all the time it's thinking in order for the AI to think it needs a factory.
And so let's say that that 10 trillion of token generated 50% gross margins and 5 trillion of it needs a factory needs an AI infrastructure. So if you told me that on an annual basis the capex of the world was about $5 trillion, I would say the math seems to make sense. Yeah. So Jensen says that global AI capital expenditures could reach $5 trillion per year by the end of the decade.
But again, we don't just want to trust him. So let's be ultra-conservative and cut that number in half. The global artificial intelligence market would still roughly 5x 2030 and almost 20x 2034, which would be a compound annual growth rate of close to 40% for the next 9 years straight. That means companies building AI hardware, software or services should grow roughly that fast, even if their market shares and margins stay the same.
One nice thing about these AI markets is there are only a few major companies in each area. For example, holding Nvidia, AMD, Broadcom, and TSMC means holding almost 100% of the entire AI accelerator market in just four stocks. Nvidia and AMD design the world's GPUs, while Broadcom designs many of the world's major data center AS6 like Google's Tensor processing units, the Metatraining and Inference chips, and OpenAI's upcoming Titan XPUs.
While TSMC manufactures those chips and many more. But AI data centers need much more than just processors. They need power, cooling, memory, and storage. That's why companies like Verdivive Holdings, Micron Technology, Arista Networks, Dell, Oracle, and Cororeweave come in. Each of these companies makes critical components or provide serious services that help other AI companies scale up.
And because these companies all do very different things, they don't really compete with each other. So any portfolio holding them gets a lot of exposure to AI while still being fairly diversified. But the second point that Jensen made in that clip is even more important. Traditional software gets written once, runs on CPUs, and is mostly operated by people looking to achieve a specific outcome.
But AI doesn't work the same way. It has context and memory and it improves over time. So it's essentially constantly getting rewritten. And instead of running on CPUs, AI runs on many distributed GPUs in parallel which requires a very different kind of network and hardware infrastructure. But the biggest difference comes in the form of reasoning.
In traditional software, the human does the thinking and the application is the tool. But in AI, the model does the thinking. It reasons through problems step by step. It uses tools and calls APIs. It renders graphics and compiles code. All of which take exponentially more tokens today than when chat GPT first came out.
That means the same prompt today could take a thousand, a million, or even a billion times more tokens in the future as AI keeps evolving. The one that that was probably most profound for me was you pounding the table. And you said inference isn't going to 100x, a thousandx, it's going to 1 billion x.
Mhm. Which brings us to where we are today. I underestimated. Let me just go on record. I estimated. And I don't just mean that AI will get better at answering text prompts. I mean it'll get better at generating audio and analyzing HD video. It'll run physics models and robot simulations. It'll synthesize drugs and write software applications.
It'll even be able to call other AI models and AI agents. But now, trying to predict the future of AI. Everything we've talked about this entire video has nothing to do with the future of AI applications at all. It's all about using AI to make existing applications more profitable and existing employees more efficient, just like we discussed earlier.
That's why Microsoft, Amazon, Google, and Meta Platforms are all incredible investments today, even though they're some of the biggest companies in the world. Each of these companies has hundreds of thousands of employees that will become much more productive thanks to AI, and that will skyrocket these companies margins.
But they also run software and services that billions of people use every month, making them the perfect proving grounds for AI at massive scales when accelerated computing and GPUs are doing the biggest jobs done by generalpurpose computing and CPUs like search, shopping, and recommending content, but at a fraction of the cost.
Every single company on Earth will have to make the switch. That's not in the future. It's already happening right now. Here, take a look. Let me give you three points to think through and and and these three points uh it'll help you um hopefully uh be more comfortable with Nvidia in this future. So the first point and this is the laws of physics point.
This is the most important point that general general purpose computing is over and the future is accelerated computing and AI computing. That's the first point. And so the way to think about that is there's how much how many trillions of dollars of computing infrastructures in the world that has to be refreshed.
Right. Right. And when it gets refreshed, it's going to be accelerated computing. That's right. Nobody disputes that. Everybody goes, "Yeah, we completely agree with that. General purpose computing is over. Mo's law is dead." People say these things. And so what does that mean? So general purpose computing is going to go to accelerated computing.
Our partnership with Intel is recognizing that general purpose computing needs to be fused with accelerated computing to create opportunities for them. Is that right? And so one general purpose computing is shifting to accelerated computing and AI. Two the first use case of AI is actually already everywhere
it's in search recommener engines isn't that right? in shopping the basic hypers scale computing infrastructure used to be CPUs doing recommenders right is now going to GPUs doing AI right so you just take classical computing it's going to accelerated computing AI you take hypers scale computing is going from CPUs to accelerated computing and AI and then now that's the second point just feeding the metas the Google's the bite dances the Amazon's and take their classical traditional way of doing hypers scaling and moving into AI
that's hundreds of billions of dollars and and because that may be four billion people on the planet today if you take Tik Tok meta into account Google into account who are already demanding workloads that are driven by accelerated comput that's exactly right and so there a s without even thinking about AI creating new opportunities is about AI shifting how you used to do something to the way new way of doing something.
Okay? And then now let's talk about the future. I just so far I've only spoken kind of largely about just mundane stuff. Just mundane stuff. The old way is now wrong. You're going to go you're no longer going to use uh uh fuel light lanterns. You're going to go to electricity. That's all. Right. Okay.
And you no longer, you know, prop planes. You're going to go to jets. That's all. So, you know, so far, you know, that's all I've talked about. And then now that the incredible thing is when you go to AI, when you go to accelerated computing, then what happens? And now we've come full circle. I started this video by walking you through how making an employee 50% more efficient at their existing job actually makes them twice as profitable.
The same thing can be said about the way that existing search engines like Google work today. the way that recommener systems like Netflix and Meta Platforms and YouTube work today and the way that shopping systems like Amazon work today. This whole video is about using AI to upgrade the way we do existing work today.
And which companies do the most work, which means they have the most to gain from AI? Microsoft, Meta, Amazon, Google, the same software and services already being used by billions of people every month. But let me point out one more important thing that Jensen said about the future of AI at the end of that clip.
AI. The future of AI is about much more than just moving today's work from CPUs to GPUs and from traditional software to AI models. That's just the first step. Just like transitioning from gas lamps to electricity, from propeller-powered planes to jet engines, from horsedrawn carriages to cars, and from renting VHS tapes to streaming, the real value comes from the new things that these transitions enable that couldn't be done before.
Here's a slide from a GTC panel that I attended earlier this year from Jonathan Godwin, the CEO of Orbital Materials and the former lead for AI and materials at Google DeepMind. Each x-axis is time in months and the y-axis is basically the impact of generative AI on the discovery of new materials in chart one, the resulting increase in patent filings in chart 2 and the rise in new product prototypes in chart 3.
The first 12 months give us a baseline for the average amount of progress in each area and the next 20 months show us the impacts of generative AI and the impact has been massive. 6 months after adopting AI, the scientific community started discovering a record number of new advanced materials for transistor-based chips as well as quantum computers for solar cells and solid-state batteries and even lightweight composits and high temperature ceramics for huge industries like aerospace and robotics.
The same exact industries we invest in. Just 8 months after adopting generative AI, material scientists were filing a record number of new patents. And by month 17, they were testing a record number of new product prototypes. And that's just for material science. There's a chart like this for gene sequencing and drug discovery, signal processing and communications, computer aided drafting and product design, physics- based modeling and simulation, movies and music and video games.
Almost every industry is seeing a rise in research papers, patents and prototypes thanks to AI. This is what I mean when I say get in early. This is why the world is spending hundreds of billions of dollars on Nvidia's GPUs. if you want to see mor.
Comments 0