원문 보기
Nvidia CEO Jensen Huang just said something everyone was silent about. In this video, we'll look at a few different moments where Jensen directly tells us what's changed compared to the past. And I think that's why Anthropic is going public. And you should know about this Anthropic concept in the background.
But think about this . Anthropic in the second quarter of 2026. That is, somewhere in April, May, or June, their annual income increased sharply. Of course? Even if you continue to grow at a slower pace , the ideal time to IPO is before another second quarter rolls around, right? Because if you look at the first quarter in history, it would seem like you grew by a huge percentage, right?
Whereas if you compare the second quarter to the second and postpone the IPO, this period will look like growth slowed down significantly compared to this crazy hype. I do n't mention this to say, "Oh, this is why Anthropic is going public now." It's so obvious. We already know this. The reason I bring this up is because something has changed.
I've always said that I think AI is a really good encyclopedic artificial intelligence; basically, it was just a great encyclopedia that could give me what I needed when I needed it, but it was hallucinating and had problems. And this somehow slowed down the year of agents, which was supposed to be 2024, as Jensen Huang predicted in 2023: 2023 is the year of chatbots.
2024 will be the year of agents. I call the years 24 and 25 a "black hole." AI was announced, everyone was spending a lot of money, and then in '24 and '25 it seemed like nothing much was happening. We finally broke the memory barrier in late 2025, which sent the stocks of memory companies skyrocketing.
And the Claude moment was possible because we were finally able to enlarge the context window and do useful things with significantly less programming hallucination. But something has changed in the last 6 months . Now it's not just the programming that has changed. Now it's video work. Now that's HR's job.
Now that's customer service. Now this is legal work. For example, actual document editing. Previously, when you edited a document using AI, you had to take, say, a 50- page legal contract and work through it one paragraph at a time. Okay, suggest edits to this. Okay, how about this for this? As soon as you tried to give him more than one paragraph at a time, he would start hallucinating , it was so annoying.
He didn't work very effectively. It took almost as long to do the work yourself as it did to work on it with AI. It seemed like you were wasting even more time. But in 2026, everything changed. In all these areas and many others, especially in the field of intellectual work. AI has become really, really good in many categories, better than ever before, including web design.
I mean, you can now go into Claude, even at the intermediate level, right? or even Opus 3.5, and ask it to edit a video for you, create a website, or develop an app. Does this mean that programmers will lose their jobs? Just not good specialists. Someone still has to enter the data. Someone still has to collect data and decide what to build.
Creativity still matters. So people will have jobs, but the implementation of a lot of things has not become incrementally, I would say exponentially better just in the last 6 months, where we thought, " Okay, this is slowly getting better," and now we're saying, "Oh my god, this is getting much better.
" But don't take my word for it; I want you to hear Jensen's comment to understand his position, and here we can watch this video. This is an interview that Ultimator Capital conducted with Jensen Huang. Does the revenue from AI justify the capital expenditure? And a little more talk on this topic. There are a few points here that I want us to listen to together.
So we'll go through this casually. But I am one of the reasons why we have become so bullish over the last 6 months, whether it is buying QQQ on the dip at 700 or getting excited about AMD, Nvidia, SpaceX or even Tesla, right? These are all companies we admire and whose shares we buy. Note, we have stakes in many of these companies, including Tesla.
We recently bought more Tesla shares, especially, you know, this is a little bit of a retreat. In my opinion, if you want more SpaceX exposure, buy Tesla. Okay, I already made a whole video about this. Just type something like " watch before the SpaceX merger" or "how to play the SpaceX and Tesla merger, Meet Kevin" into YouTube and you'll find it.
But listen to what Jensen says right here, and then we'll go through some of the other components. We're going to jump around a bit on this one , but listen carefully. This part is very important, but it's also important that you get to reinvest.com before this Friday, because just like that cup of coffee, this coupon will be gone soon.
And you want to the Reinvest terminal. So, once we enter 2027, there will be no more lifetime access . But anyone who makes it to this point will get it forever. So join us at reinvest one.com anytime . safety and security. Of course, recognizing that this industry has recently gone through a transition, right?
About 6 months ago. This technology barely worked. That's right. And so functionality, think about it. You had Jensen pushing AI since 2023, and now he says it barely worked. But six months ago, oh my god. Just let me, this is probably the most important phrase in all of this. Recognizing that this industry has recently gone through a transition, right?
About six months ago. This technology barely worked. That's right. And so functionality, and without functionality and capabilities, what's the point of making it safe and secure? Nobody is even interested in it. And so keep in mind, all this talk about security, safety, consistency and all that now.
And that was just a short excerpt from here. But the whole point is, you know, does AI justify that cost? And in 2023, 24, and 25, you had to go through these three years of spending and, you could say, AI development. Now we're like: oh my god, we can do so much more than we thought before. So if you're skeptical about AI because of your experiences in 23, 24, 25, you really should give it another try.
I don't care if you want to use Muse, Grockbot, or Claude. I don't have favorites and, to be honest, I personally don't care. I think, you know, they all contribute to the development of the same market that we're all looking at. But if you didn't like it, you need to try again, because if you're stuck thinking, " Well, you know, he hallucinated me in the past, so I'm just not using AI." You will be left behind.
And it's very difficult to justify spending on AI based on the past 3 years. But if we look at the last 6 months, we're like, "Wow, this is getting really exciting." Cloud computing is for the next generation, right? It's kind of like AWS for today, you know, and it's not just AWS for the future, right?
And that's about a third of our business overall, you know. I think we're in the early stages of building this new computing architecture, and it would be a big mistake to single out just a few companies and think that they're the only ones funding the infrastructure, and one observation -- the open model is a very important component, and I'm really glad he mentioned that, which is why I wanted to get to that : he says it's a mistake to think that it's just a few people funding it.
I think that's a direct response to what I was saying about us having these two "canaries in the coal mine," right ? So, we have two canaries. The first is the labor market. The job market is now normal. I mean, this morning we got 16,750 jobs created . We are moving at a pace of 60,000 jobs per month.
At some point, that will change. This is perhaps the biggest “canary in the coal mine” that will eventually lead to big trouble in the economy. But that's not a problem now. That could be at least a year away. I would say that we will be able to learn about these changes when we see that laid-off workers stop finding jobs quickly.
We are still seeing the absorption of the workforce now. But the second "canary" is what I called the AI underwriters, that is, OpenAI and Anthropic. Jensen is essentially saying, "Listen, we're just at the beginning of the journey , because it's not just Anthropic or OpenAI." In fact, it's Anthropic, OpenAI, SpaceX, Meta, Microsoft, Amazon, Google.
Is it true ? And then there will be others like Marvell, Broadcom, AVGO, or, how would you say, bigger players like Oracle. So, these will be developers of custom ASIC chips, of course. Then we have AMD, then Nvidia, right? These are some kind of builders. So, here are your chip manufacturers. You can add a memory sector here.
Memory, Micron or someone else. But these are actually underwriters. While these guys are in the business of funding vendors and developing chips or whatever, this theory, based solely on OpenAI and Anthropic, boils down to the fact that we need advanced models. He said that only spending on advanced technologies is supporting this bubble.
I used to think exactly the same: that only cutting-edge developments support these crazy costs. But I'm changing my mind. This is wrong . Yes. It is wrong to say that it is only about advanced technologies. It's not just advanced technology. These are also corporate expenses . So, let's call it corporate spending.
This is exactly what you are observing now. You know, this is one of the reasons why Dell is showing such good results. This is one of the reasons AMD is so successful: because corporations are spending so much on AI, not just through cutting-edge models, but also through open- source models. So, really, because all corporations are spending , all of these companies benefit from open-source models, just like these guys.
Therefore, it is too simplistic to say that only Anthropic and OpenAI are funding all this . This is every business that says, " Wow, we can do so much more with open-source and advanced models." You know, we could be 90% closer to the cutting edge , or maybe we're only 6 months behind the cutting edge with our chips, versus paying $50 for a million initial tokens, which is 100 times more expensive than if you spent it on, say , Qwen—one of Alibaba's newest models, which is impressive.
So, I find this very exciting, and I agree with Jensen that we're probably still in the early stages. In fact, it may seem strange, but I think it's more likely that the labor market will collapse before corporations stop spending money. I think corporate spending is just getting started, which is pretty wild.
And the reason for this is that we work with them —depending on the size of the client—for at least a year, and for the largest ones — with a forecast of two years. So we already have a very large order portfolio. And think about that for a moment. He tells you that they are already taking orders for 2028.
And 2029, because they look two full years ahead, so the 27th, the 28th, and that rolls over to 2029. So, I guess, a full year of 2728, and then a rollover to 2029 for large check orders. This is incredible visibility that Jensen has. That's how we get these forecasts, and what's striking about them—Altimeter is pointing this out—is this graph right here, where, essentially, at the beginning of the year, people were expecting Nvidia's earnings to be this gray line , and you see 2026 earnings at $2.53 .
It is now $3.83. Part of the reason for this was a turning point in functionality. AI is becoming qualitative. AI. I know it sounds old- fashioned, but it's 50 % higher, and by 2029 it's three times higher. So this boom is just beginning. You know, people are angry with me. They say: Kevin, you keep advising to buy on the dip at 700 on QQQ.
Yes, I advise. And here we are again at 720, but I'll be frank. Yes. Do I consider our forecast today, well, a short-term daily forecast...We made the forecast said: look, AMD is the option for today, along with Bloom Energy. You see, most of this happened during the day after our morning call, a 5% increase, which is great.
Bloom also grew significantly throughout the day, which is very exciting. at. This is... a long-term perspective for the next two to three years. Until the job market crashes , this QQQ chart, I'll show it in a moment... in my opinion, it's just going to look like a long- term opportunity. Maybe I'm too optimistic, huh?
Maybe I'm getting too excited, but I think these AI stocks have gotten too cheap, and we're near the 700 level, somewhere where my cursor is. Do you understand ? Take a look, just look at the graph. How many times have we bounced off 700? Again, again, again and again. If we get a deal with Iran, I think we'll never see 700 again.
But even without a deal with Iran, I'm very excited that now is a great time to get into stocks. These could be private companies working in the field of AI. I think we're just getting started. These can also be public actions. I personally think so, and this is just my opinion. Well, let's get back to Jensen.
Uh, I think that... The technological improvements that we're seeing in AI capabilities, especially in open scales, will make the products that we can provide in the Reinvest terminal—both in equities and real estate—much better than we could have even imagined 6 months ago. Every day there is a feeling: " Wow, we could do this, and this, and this." And what we are building is very exciting.
But let's focus on Jensen for a moment. Let 's get back to this order book review. And I want to go specifically to the Vera Rubin outlook because Vera Rubin is turning into an Ultra order. And listen to what Jensen has to say about it. Let me give you an example. Let's just do a thought experiment right now.
If you had to hand someone a check for a hundred billion dollars to place an order right now for the next generation. Yes. If you want to get in line for Vera Rubin Ultra, the time to place your order is approaching. Wow. Without seeing the product. Yes. True. But they have confidence. So, let's say you need to invest $100 billion in someone.
Yes. First, how do you know it's going to be something worth considering? True. But since Nvidia is building it, you know it's going to be better than Vera Rubin, right? And Vera Rubin—that's cool, is n't it? And the reason why Vera Rubin ... the reason why everyone had confidence in Vera Rubin is because Grace Blackwell was wonderful .
Yes. So, essentially, Nvidia's presentation is building up about how their reputation led them to this. And I understand that Jensen will be biased. We know this. But I completely agree with his prediction that something has changed in the last 6 months, and it's no longer just OpenAI and Anthropic that are the cutting-edge companies that support everything.
Anthropic is actually less interesting now than it was 6 months ago. This has nothing to do with their IPO documents or anything like that. This, you know, I'll make a separate video about it. This is due to the fact that we are expanding, and the situation is actually even changing. Even at Anthropic, they don't necessarily have to be just a cutting-edge lab.
They should become the platform you turn to for everyday use, which may even include less expensive models that are no longer "cutting edge ." The ones that Claude has, you know, Sonnet or Haiku or Opus, right? Okay, then let me do the math. Let me see if I can take... Okay. Okay, great. Let's evaluate Nvidia.
This is the last thing I wanted to do here. Have you or a loved one been harmed by social media? Morgan&Morgan can fight to help you. If you or your child spent years growing up on social media and experienceat communicatmp; Morgan is stepping in rebates. potentially get over $1,000 in refunds . Video.
This is the last thing I wanted to do here. And I think everyone gets excited when we talk about grades. So, let's take Nvidia. Nvidia. What we're going to do is I'm going to write this out, because sometimes it's not entirely obvious how these numbers come out. I like to use price-to- earnings multiples.
You should already know this. I don't think a company should be valued solely by its P/E ratio because you forget that there are expectations for the company's growth. And comparing growth expectations helps us understand how this company compares to another. For example , Kroger will grow 4% per year, 2% per year, yes, with net margins of 1%.
Nvidia or AMD will grow by 55% per year. Nvidia can grow by 30- something percent per year. It all depends on the circumstances. But let's look at this. OK. So for Nvidia I have growth projections for 27th year of 69.51%, then 34.82%, then 18.32%, then 15.13%, dividing by 4 I have 34.4%. Growth, okay, so we'll say growth, and we'll take 34.43%.
for four years. Okay, let's make this simple and clear . The stock is currently trading at $218. I think it's a good deal. I also own a stake in Nvidia and continue to buy more Nvidia shares. 23.37. PE PE January 27 is 23. 27. I can also show AMD figures , because we did AMD analysis this morning. Let's take AMD's numbers, because they will also be interesting.
So , AMD expects significantly more growth. The growth rate for AMD is 55.88. I'm really excited about what they're doing with Ryzen Pro. That's a topic for another video, but let's put it this way: I don't think Wall Street is paying attention to what's happening with Ryzen Pro because it's " hidden" in the gaming sub-segment and it takes a little more effort to see real growth.
AMD's price-to-earnings ratio is now 71. Okay. Now with this we can get what is called the PEG ratio. And all we're going to do is divide the price-to-earnings ratio by the growth rate. So, 23.27 divided by 34.43. This gives us 675 for Nvidia, but it should actually be around 2.4. And for AMD we will have 55.88.
This is the growth rate, sorry, 71 divided by 55.88. They have it at 1.27. So, Nvidia is now half the price of AMD. This is crazy from a valuation perspective. OK. Both of these companies should probably have a ratio of 2.4. At a PEG ratio of 2.4 divided by 675, Nvidia is probably worth a look, let's take that.
Let's take a different color here. 775 dollars. So, a fair valuation for Nvidia might be $775. In my opinion, at some point during the next hardware supercycle that is approaching, the 227 price point for Nvidia will be a thing of the past. I know, that sounds crazy, because then it would be a company with a market capitalization of over 10 trillion.
But it's not without reason, given this transformation, which has really gained momentum over the past six months. I really think it's just the beginning. I agree that we are in the early stages. Maybe I'm wrong. OK. And here, it's about double for Nvidia or for AMD. So, without getting into the math , that would give us about 1080.
So let's write that down. About 1080. So you definitely have more space here. But the thing is, in terms of dynamics, AMD moves faster because it's a smaller company. But they are both quite cheap. So I'm thrilled. I don't know, maybe I'm too optimistic, but I think a lot of the bearish predictions about AI are based on old AI, on that old garbage, on what was there 6 months ago , on hallucinogenic stuff.
This was worse than "magic mushrooms." I've never tasted " magic mushrooms," so I don't know, but I've heard about it. In short, this is my opinion. And I have an important reminder: go to reinvest one.com. Use your coupon. Congratulations to those who joined right now while I was filming this. I'm getting notifications here .
Thank you for joining. Here, Ryan just joined. There are other names here. This is cool. But it's time for me to go. Thank you very much for being here, friends. See you in the next video. Go to reinvest one.com, see you there. So, let's begin. another take your example. Kevin Pratt, financial analyst and blogger. Meet Kevin. Always happy to hear your opinion.
Comments 0