Ticker Symbol YOU: Why Michael Burry Is Wrong About Micron

Hello everybody. Welcome back to the Wolf Financial Podcast. My name is Gav Blackburg. I'm the CEO at Wolf and your host for today. And joining me for another amazing episode is Alex from Tickerol U. He's really my resident tech memory AI expert. So I'm really excited to dive into this conversation with him. We are recording this right now on July 17th and I think that's pertinent because the market has been in a little bit of turmoil when it comes to some of these AI names and also some of the underlying tech below it. And so what I really want to start off discussing here, Alex, is what is an opportunity and what is a trap and how do you decide, hey, the stock has come down. This is something that I want to add to here versus that is a deteriorating sign and not a good one. So let's kick it off there. And also, you know, how you doing today? >> Sure. Yeah, first of all, thanks for having me. It's always a pleasure. I think this is our like fourth or fifth time chatting. You know, I always really enjoy these uh chats. So thanks for having me on again. Um opportunity, right? So I think one of the big things that people miss when they think about opportunities is they think you know the big signal is coming from within the space. So let me explain what I mean by that. you know, hey, memory stocks are doing well because something happens in memory, right? AI stocks are doing well or poorly because something is happening in AI specifically, you know, whether it's a new model release, whether it's a new chip release, etc., right? Um, one of the big things that I try to do on my channel is we try to look at like one level up from that. So, for example, uh, the straight of hormuz recently got reclosed, right? and that sends supply chain shocks all over the place or at least that news sends news of supply chain shocks and so stocks start to drop because of that. Um there's rumors of Nvidia delaying their Kyber rack which is like their next generation of like rack that their next generation of chips sit in. That rumor is unconfirmed and denied by Nvidia, but that's something that could shake the market. Um ASML just had their earnings. TSMC just had their earnings and even though they're growing fast, they talked about some of the constraints in their own production. Basically, what I'm trying to say is sometimes the opportunity and the signal for why AI stocks go down or why the market goes down in general is is like one level higher, right? So, whenever that happens, what we're looking what I look for is, hey, there's bad news. There's a disconnect between the stock, the direction the stock is moving and the direction the business is moving, right? So stock goes down, business goes up. That disconnect is what creates an opportunity, right? When a headline causes stocks to go down, that isn't related to like that stock itself, right? No one's in trouble. The company's doing great. Um, but then the company comes out with an incredible earnings, you know, that to me is what an opportunity is shaped like overall. >> Makes sense to me. Let's talk about it in practice. So, right now, how are you actually allocating capital? And on this, you know, recent pullback, is there anything that's become attractive to you? >> You know, I I hate being the boring guy in the room, but like basically what I try to do is I dollar cost average in and when the market is ripping, what that means to me is like my money goes less far uh per dollar, right? So like stock prices go up, I put in I try to put in X dollars amount a month, you know, I'm getting way less shares for that money. When the opposite happens and the market goes down, I try to dollar cost average in a little more aggressively. Not much more. I'm not like going all in every little dip or anything like that. But when I notice prices are going down, I'll okay, I'll add in a little more money. So the types of stocks, in my opinion, every great portfolio starts with a fund. I always talk about this same one fund on my channel. It's VGT, Vanguard's information technology ETF. Uh that to me, if if that fund is ever down at all, like on a serious dip, I try to load up. So that's that's the fund that I treat as my cash position. And I do that because it's like kind of the perfect blend of performance and diversity for me, right? It's got 300 stock. We we can talk all about the fund, but at a high level, it's got 300 stocks. It's got a lot of the winners at the top. It's not afraid to let its winners ride. Like the biggest positions can be close to 20% of the fund. It's outperforming the the S&P 500. It's outperforming the NASDAQ. It has lower fees than both of them. You know what I mean? So, it's like it just kind of like checks all my boxes. So, when I don't know what to do, which admittedly is like plenty of the time, I'll just throw money in VGT and wait. And then when I see an opportunity for a specific individual stock, I'll take that money out of VGT and I'll rotate it into that stock specifically. >> Yeah, makes sense. And let's talk about these specific stocks. So, I want to kick it off with Micron because this was one of the most talked about stocks of the year and you have a very low cost basis on it. You mentioned it's around $150, I believe, per share right now. That's right. Micron went all the way up and they peaked at about $1,250 per share and since then it has had a bit of a steep retrace about I would say 30% at the moment 35% uh as of you know yesterday and now back up to 30 since it's had a decent sized move here on the day at this point trading at 8 uh67. We talked uh pretty recently and you had mentioned that 750 was a potential ad level for you. So just you know obviously you talked about hey there's a dollar cost average approach here and that makes sense to me is there a secondary addition to that as well of hey if this stock was to come back to a level that I feel comfortable I'd maybe take some that BGT money and put it into it and as you examine Micron and you're you know maybe feel free to go into depth on the memory here and what people could expect moving forward why is this something that you would add to and where would you add to it and then just one other piece within there I think some people saw a post from Michael Bur uh that came out uh pretty recently ly and Bur retweeted actually a post from a friend of mine uh Dr. derivatives. And so in his post he essentially cautioned people against buying uh low uh let me see what he said. He said he said now class repeat after me. I will never buy cyclicals at low pees and record earnings. Say it again. I will never buy cyclicals at low pees and record earnings. The one that he's referring to here are MU, Nvidia, Meta, and Microsoft. Those are the ones that were kind of selected out there. So just curious to get your thoughts on that. Yeah, I I honestly I think that's a really smart thing to say and I think the word like the specific word we should all be paying attention to there is cyclicals. So, first of all, I agree with Michael Bur's tweet entirely. What I don't agree with is that Micron is purely a cyclical company anymore, right? Um same with Nvidia, same with Meta, obviously. Uh same with pretty much every semiconductor company participating in the AI revolution. So, let's like break those down, I guess, right? So, Michael Bur saying something very smart, but we should remember that his background is really in real estate. Uh, he's the guy who called the 2008 crisis. You know, shorted it well in advance, eventually made a lot of money, incredibly smart guy, incredibly detail-oriented when it comes to these kind of things overall, but he's not a technologist, right? And he's certainly not an AI technologist. And the thing that I think he's missing with that tweet is more and more memory specifically. Let's just talk about micron since that's how you sort of open the question. Um, a lot more memory is moving from cyclical sales, right, commodity sales to contracted upfront like multi-year bill multi-billion dollar contracts, right? And when a commodity starts being priced on a contract basis and on a forward basis, it loses all the risks of cyclicality. So, let's talk about what that means sort of in detail, right? >> Love it. >> Um, in a cyclical market, let's just use RAM as an example. A new iPhone comes out, a new GPU comes out, a new reason to buy your You may see a cat on my lap here. Sorry. Um, >> good. My pup's been in and out of here. >> Yeah. So, we for for your audience, we have a bad thunderstorm here. I have a couple animals. They're afraid of thunderstorms, so I apologize if they end up like coming to me. But um yeah so memory is cyclical right traditionally and so what happens is something creates demand for a new PC a new phone a new whatever right it's time to refresh your device so you do that you buy your new computer or you build it you know you buy your new phone or you build it you buy a new laptop or your work buys a new laptop right there's a big refresh and so what happens is there's all the sudden a lot of demand for memory right with all those computers and devices I just talked out lots lots of new memory sales. Now, everyone's got their new device and they don't need one for a while, right? So, demand for that memory goes down, right? That's the that's how cyclicality works. There's a spike in demand. RAM sales sell out or, you know, they go up in price because production doesn't like kind of keep up with it. Then demand crashes. Production stays consistent. When there's consistent supply but lower demand, prices go down, right? And so, you get this like oscillating price curve. Go ahead. Sorry. And people could kind of think of that as like, you know, the new iPhone came out or the new MacBook Pro came out. >> 100%. You're spot on. The new Xbox came out, right? And it sells a million units. Each one of those units needs memory, right? So a lot more memory gets sold than like the previous cycle where there was no Xbox, right? >> Um, >> excuse me one sec. >> Sorry buddy. Um, so cyclical memory, right? The key there is memory today is not >> that one. >> Yeah. Did you? Yeah. Um, so memory today isn't working like that. How is it working? So a hyperscaler, Meta, Google, Amazon, Microsoft, they're going to Micron, they're going to Samsung, they're going to skinex and they're saying, "We want all of your memory production as much as we can take for the best price." Right? So Micron all of a sudden goes from having to project supply or sorry, excuse me, project demand and try to make enough supply to meet it and not go over or else they're stuck, you know, sort of holding their excess inventory which comes with its own costs. They've overproduced. They can't sell it all. Now they have to get rid of it at a discount. That's like the other side of the cycle, right? When you see memory go on sale for 20, 30, 40% off, right? Which, by the way, if you've ever built a PC or you pay attention to that market, you haven't seen in a while now, right? So there's your first hint that memory is no longer cyclical. RAM prices are outrageous, right? Why is that? Let's talk about the contracts that I just mentioned, right? So, Meta, Microsoft, Amazon, and Google, the big four hyperscalers, they're signing contracts that say, "We will buy all of your memory capacity, specifically your high bandwidth memory capacity, because we want to put it in our data centers, and we can't get enough." >> Right? That's not cyclical anymore. Every single stick of memory, every single chip of memory that these companies produce will get sold at a premium because if Meta doesn't buy it, Amazon will. If Amazon doesn't buy it, Microsoft will. If Microsoft doesn't buy it, Google will. Right? There's a lot more demand than there is supply. So, these companies not only get to elevate their prices, but they guarantee that every unit they produce will be sold. All of the cyclicality risk goes away. That's what Michael Bur is missing with his tweet, right? It's He's not wrong about cyclicals. He's just calling the wrong companies cyclical. >> Why do you think he thinks that this is still cyclical? >> I I you know, for a long time that these companies were cyclical. I again, I think Michael Bur is a very intelligent guy. I really like when he speaks, I listen. I just don't always agree because I'm seeing things from a technology angle and he's probably seeing them from like a macroeconomic, you know, hey, I know this company makes memory. I know memory is cyclical. What the heck is going on here? The price is going up and the sales don't justify this price. >> What else does that do? What else does that do for you when you remove cyclicality from that industry and now they don't have to worry about, hey, we just need to produce as literally as much as possible and it's all going to get bought and it's going to get bought at a premium. What what does that do for the company? >> Yeah, so I mean it does a few things, right? So obviously when their valuation goes up like this, it actually derisks the company because now when they're selling shares, they're selling fewer shares. They can do they have a lot more leverage with banks. They have a lot more ways to raise money at a reasonable price for them, which starts de-risking like their finances and their books, right? You know, one of the best things that can happen to a company is its market cap goes up. It starts paying people out in shares. It starts doing all this stuff, but it's doing less, sorry, it's using less shares to do that because each share is just worth more money, right? That's one. The big thing it does from a technical perspective is it lowers their technical risk. When we there's all different kinds of risk, I think when people think when investors think about risk, you know, on the surface, they're really talking about volatility, right? They're talk they're talking about the risk of the stock going down. And if you're not ready to retire, you know, my thinking is when the stock goes down, it's a good time to buy, right? It's like, you know, when my favorite stores have a sale and things I like go on discounts, I don't think that's a bad thing. That's like when I go to the store, you know what I mean? >> So, do ahead. Sorry to keep cutting in, but you always spark great thoughts. Um, on that point right there, right, and we've been trying to drill into people for so long. Hey, the stock market when it goes down, it's like the store going on sale. But I think some people see it differently because they're saying, "Well, why does the store, you know, put things on sale, they put things on sale because people don't want them. Uh, and so, you know, they're trying to get rid of them. They're trying to get them out the door or something like that." Or they're just trying to create some type of phenomenon, you know, Amazon Prime Day or something like that. I think when it comes to the stock market, what people think is, you know, they think of a concept that they don't think of when it comes to retail. They think of like smart money and dumb money, right? And they go, "Oh, it's going down. Somebody smarter than me is saying it's not worth that. it's it's being sold off, right? Like how can people break out of that mindset? >> Yeah, let's let's tackle it in two ways. I think that's a really great observation actually. So, like let's let's talk about that a little bit and see if we can make our way back to memory, right? Um, one thing that stores do that the stock market doesn't is stores have new inventory. What does that mean? Like, let's just talk about a clothing store because I think that's kind of what everybody thinks of first or like a computer store, whatever. things go on sale because the new thing is coming out, right? Like, hey, the iPhone 15's on sale cuz the iPhone 16 is about to come out. Winter clothes are on sale because spring is about to be here, right? You know what I mean? It's like you're one season behind. The next new thing, the whatever, right? Hey, the the latest Nvidia GPU is on sale because the next one's about to drop, right? Down and down the list. Uh the stock market doesn't have that, right? It's like the the next Micron share that gets produced or you know whatever is is a share of Micron. There's no there's no difference functionally between an old share of Micron and a new share that they just like distributed, right? Except the new one dilutes the old ones, I guess. You know, so that phenomenon that people are worried about in stores doesn't really happen in the market the same way because there's no new thing and old thing. M >> the second thing I want to point out is there's actually no such thing as smart money and dumb money. The I I do think this used to exist. So like back before the internet or when the internet was really young and information wasn't like easily and freely distributed, you can make the argument that there was smart money, the people who were doing the research and sharing that research locally, right? Like at big hedge funds, at big banks, you know, private family offices, blah blah blah. And then there was the dumb money, which is kind of like the retail investor who didn't have access to that information, right? >> Yeah. >> I I get that that concept used to exist. Today, that doesn't really exist. A lot of people do great due diligence on Substack, on Twitter, on YouTube, right? Like there's there's Yeah, I'd like to consider myself among them, but I'm I'm not really talking about myself. I'm talking about like our whole ecosystem, right? Like >> bro's getting half million views on YouTube videos but doesn't think that he's >> a I mean it's it's yeah I'm I'm flattered but in reality right it's the whole community right it's guys like you it's guy like including guys like me but across lots of different surfaces email newsletters right it's like you don't even have to go seek it out you press one button and that information comes to you right is kind of like the overall point I'm making so I think instead of thinking about smart money and dumb money we should be thinking about fast money and slow money, right? What is fast money and slow money? Slow money are the institutions. They don't make moves unless they're sure or unless they're algorithms and their robots and like the things that really move the market, which is largely al algorithmic market cap weighted strategies, right? Like you big funds that rebalance quarterly like based on certain metrics, these kind of things. You know, people are watching that like a hawk and then they're acting accordingly. That's slow money, right? These are pension funds. These are big hedge funds. These are people who minimize volatility while trying to make their ticker like their fund go up regularly, right? They're big institutional slow money. And slow isn't bad and fast isn't good. It's just these guys have a lot more on the line and their performance is directly tied to like their paycheck, their job security, right? Like so they care a lot more about did my stock go up this quarter, right? They're slow money. Fast money is guys like us, right? It's like, hey, when the market dips, I put more in. When I think I'm I found something, I just buy it, right? When I have a better use for that money, I just sell it. I don't need approvals. I don't follow compliances. I just get to do what I want. I'm fast money. I'm agile, right? Um, I think retail investors have a huge edge because they get to be fast. And when you're fast, you get to like look at the wave that's coming. You know, you're a surfer. you see this big wave and you get to ride it. But when you're in a giant boat, you know, you got to decide, am I riding this wave or am I getting out of the way? And you have to decide very early in that wave's life cycle because you're a slowmoving, big moving boat, right? And your job as the captain of that boat is to keep everyone on that boat safe. When you're a surfer, you can be like, I'm a good surfer. I'm going for it. If you're wrong, you get washed out. But if you're right, you're fast. You're riding that wave all the way. You know what I mean? So surfers and cruise ships, not smart and dumb, not you know what I mean? That that's kind of how I think about it. Does that make sense? >> Yeah, that's a good analogy. Now, let's work it back to memory like you were saying. We were talking about how the industry is no longer cyclical. So coming down with some of those pieces, let's kind of revisit that. >> Sure. So, um, Michael Bur is in an interesting spot, right? Because he is kind of an individual investor, right? like we think of Michael Bur the individual, but he's also got and I apologize. I always blank on the name of his fund, but like he he's a big asset manager, you know what I mean? >> Yeah. I mean, Cassandra Cassandra's his handle and like I know he's got >> Yeah. I apologize. I I didn't mean to. Um but like so he's got a fund that he manages clearly for other people, right? Yeah. >> So he's at the same time, you know, he talks fast but he has to move slow. >> Scion management. >> Scion. Thank you. Scion. management. Yeah. >> Yeah. Um I I follow it on whale watch. I just don't have it up in front of me. Um so he talks fast, but he actually has to move pretty slow because he's managing money for more than just himself, right? So he tries to call these cycles early. He tries to think about them from a big macro picture. He's trying to look for a wave out there, right? Memory is cyclical. That is a long-term truth, right? But in reality, very recently that changed. So, you know, he's just not doing a good job reading that particular wave in my opinion, right? Um why? So, we already touched on the fact that memory is cycl is no longer cyclical because of contracts, right? One of the big things is the hyperscalers are just buying up all of the revenue production, excuse me, all the memory production and fronting all of that revenue so that these memory companies can invest as much as they need to to ramp up production as much as they can to produce as much memory as they are able to knowing that it will all get sold. Right? So that's another tailwind. You talked a little bit, you asked a little bit earlier, what are the ways that this derisks memory, right? Another way this d-risks is they can invest heavy in production. All of the sudden all of their investments, you know, their return on invested capital goes way up because they're like, "Hey, if we build it, they will come. If we build a new factory and we fill it with memory production, it will all get sold out, right? So imagine, you know, you're a car company and you're like, "Hey, if we build this next Ford factory, we will sell every unit we produce." That's insane, right? Like that that almost never happens. It's happening in memory right now. So, um that's sort of my rebuttal to >> Michael Murray, you know. >> And how long do you think that that can go on for? >> Uh years, right? So, it's like I I think one of the mistakes I think another mistake that maybe retail investors make that sort of slow money has queued in on is that like things don't really happen instantaneously. Like people think that like stocks go up and then they go down and then like or like bubbles pop and like one day it's over or you know what I mean? Like they they think about things in terms of like binary the market is good or the market is bad. Stock go up or stock go down you know um the dot bubble for example took three years to pop. Three years like the width of the bubble is like you know how long people spend in college almost right? So my point is like you actually can take some time and really think before you act. You don't have to worry about like missing your window, right? Institutions know that when they whether a stock is cyclical or not, whether demand is up or down like in the moment or whatever, what they're looking for overall is like the overall long-term trend, right? The overall long-term trend for memory and for chips is demand is going to run away relative to supply because as soon as there's a new version of high bandwidth memory, chips will rise to meet it. And as soon as there's a new chip, memory will have to rise to meet it. Right today, we're on HBM4, maybe HBM 4E, depending on which company you're tracking, >> HBM 5 is going to come out and then for a hot minute, memory is not going to be the bottleneck. So people are going to be rearchitecting chips around memory that can feed that chip literally as fast as physically possible, right? The chip will exceed the memory speed, the really the memory's bandwidth, right? Uh and then all of the sudden we'll have to rethink how memory works. You know, maybe there's a new shape that memory should come in that's slightly faster at transferring to the chip or slightly faster at transferring to a server that transfers to a chip or maybe the answer is using memory through light or you know like I'm I'm kind of making this stuff up but you can imagine a company that you know instead of fiber optic networking it's fiber optic connecting memory to the chip or something right um my point is there's like so many avenues and so many angles in which memory can improve and chips can improve around memory and all that stuff. This will be going on until literally like AI saturates every market. Just like the internet saturated every market, computers saturated every market, mobile phones saturated every market, the printing press saturated every market, right? Like we have like historical president is what I'm saying. Like I'm not saying anything that people should be like that I would be that sounds crazy, right? Like we have technologies that saturate every market. >> Yeah. pretty fascinating stuff and not to go too much farther on memory because there's a couple other things I do want to hit but sure we talked Micron for a good amount of this and then obviously SKHix is on the scene now SKH had a big boom off the bat when they did IPO here on the ADR well the ADR on NASDAQ they went from 149 all the way up to 193 they are back to 154 um you know you mentioned to me that this was actually potentially even slightly more attractive to you than Micron just because of some of their unit economics and other pieces, but that of course you do have a very low cost basis on micron, so you're not necessarily trimming it off to buy pieces of that. If SKHX was to drop another 20 30%, does is there a point where it gets even more enticing? >> I I think so. I think the question isn't how far does SKHX drop? It's always like how far does SKHX drop relative to Micron, right? So, I think for me, I think for me personally, like I'm ne I try never to be like do what I say, not what I do guy, right? like that's never a good look. But because if you like go watch my channel over the last few years, we've been talking about Micron. So people who invested alongside me three years ago are in a very different position than those who are watching my channel maybe for the first time today and seeing Micron at like literally 10x the price than when I was talking about it. So, it's like I kind of have two audiences when it comes to Micron and I'm I'm in the like first camp of like, hey, I thought this was a good stock at I think it was like probably 150 to 180 is when I really started like, hey, something is wrong here, blah blah blah, >> right? >> You know, so compare that to today's prices. And of course, people are like, hey, you know, Micron's overvalued. Why are you still talking about it? And it's like, well, because it's a big holding, you know what I mean? like and and a lot of people watch me to understand what to do with this stock, right? Um go ahead. Sorry. >> No, no, finish the thought cuz I want to transition a little bit off memory. >> Yeah. Yeah. So, so for me, if I was a new coming into the channel new, right, SKH does look more attractive. It's just a straight up better valuation. Um, so of course, you know, I want to be cognizant of that and I want to tell people, hey, the better memory play here now that it's available to you, right, on the stock market on the NASDAQ is SKHEX. It's just not the one I bought because I bought three years ago, not today. Right. >> Yeah. Makes sense. >> Yeah. >> All right. Outside of memory, which I know makes up a good chunk of the port, what do you think right now when you look at your portfolio is the stock that is getting the least attention? Like people are not talking about it, but they really should be. I know I put him on the spot. >> I I think the n probably so like it's I don't know what people talk about. So like my gut says the answer is ASML. Like if I had to pick one stock that people should be talking about way more. Uh it's ASML. And it may not be for the reason that people think, right? So when things get supply constrained, so ASML has a few benefits. It's in the Netherlands, right? So it's like pretty far removed from the straight of Hermuz even though like it has a really global supply chain, but like it's pretty resilient to these supply chain shocks. ASML is like first and foremost a supply chain company. It is like mastered the art of creating this really ridiculous machine that it sorts that it sources from like I want to say over a thousand different companies, you know. So like when it comes to supply chains, nobody has it nailed like ASML in my opinion. And so in times like these, that's they have a lot of supply chains resiliency relative to other companies. Does that make sense? Like a lot of other companies are just worse at managing their supply chains through COVID through these supply chain shocks. Go ahead. >> Yeah, just for those that aren't maybe deeply familiar, can you just elaborate on your thesis about them and then go into the >> Sure. Sorry. So, ASML uh is the company that makes EUV lithography machines. EUV stands for extreme ultraviolet. It's just a fancy way of saying they make the oven that the chips cook in, right? Like, think about it that way. So, if TSMC runs an industrial kitchen that makes all these amazing chips, right? Nvidia's GPUs, Apple's processors, AMD's stuff, you name it, right? TSMC makes all the chips, right? ASML makes one of the most important machines in that chipm process which is called the EUV lithography machine. Right? Okay. The special thing about ASML is they're the only company that makes these EUV machines. They like I'm not using the word monopoly negatively here. I'm not trying to call them a monopoly, but they are effectively the only company on planet earth that can make these machines at enough scale to give them to TSMC, to Intel, to Samsung, etc. Right? Like so that they're a single source supplier for a very important machine, especially in the AI era. >> They're also extremely supply chain resilient in an era where we have tariffs, uh supply shocks from COVID, supply shocks from the straight of Hormuz closing, right? like all these other things that are affecting all these companies but that don't come from the company which is how we started this conversation right a supply chain shock over here affects us all even though that's not any one company's fault right like no company did anything wrong that caused this this is a bigger picture macroeconomic thing that's happening is largely removed from that not entirely right because some of their stuff just comes from all over the world but largely right um between so between being extra resilient to a lot of the things that's pushing the market down and having a monopoly on like one of the most important things of why the market is going up. I'm really surprised that ASML just like isn't on way more short lists uh that you know memory gets made on ASML's machines, GPUs, CPUs, AS6, right? Like this is the company like if you delete ASML, you you just delete the stock market at this point, you know. >> Yeah. And up 55% year-to- date and has held up a decent amount better, I would say. Also, even on this pullback potentially because of that monopoly like structure that they have, they're down, you know, 12% in the same time period that an MU is down 30%. >> Yeah. >> Why do you think that they have no real competition? >> Um, I think the problem that they solve is just like incredibly hard to solve. It's like the the only thing they do. I also think it takes a lot of like specialized parts. One of my favorite like little facts about um ASML is so they they work with a mirror company called Zeiss, Zeisss. Um and Zeiss makes the flattest mirrors on planet Earth. And uh if you put so basically the way EUV works at a super high level like I'm not trying to derail the conversation is uh it takes a bunch of ultraviolet light shines it through a bunch of mirrors and then that light is in a pattern that the chip will get like stamped out in so to speak. Um and that light hits the chip and it etches out like the pattern of the chip. Right. >> Right. So, these mirrors need to be super super super flat because there's a lot of them and if they're not flat, they distort the light, which changes the shape of the chip, which results in errors, which is how you get lower yield, right? So, one speck of dust can cost like millions and millions of dollars if it fs up an entire like um >> yeah, >> you know, silicon wafer of Nvidia's GPUs, let's say. Um, so these mirrors are so flat that if you blew one up to the size of the United States, right, the 48 contiguous United States, the sm the biggest bump on it would be about the size of one grain of sand. One grain of sand. It is one of the flattest things that like produced in the universe are these Zeiss mirrors, right? That's one part in uh these EUV machines. So it's not that simple to just like oh we can build it too. There's a lot of like science, engineering integration institutional knowledge, physics like this. I mean I almost hesitate to even use the word physics and we should start using words like black magic, right? Like like these guys are wizards, not engineers. And the fact that this stuff exists at all is mind-blowing even to me and this is my background. So >> have you had a chance to visit any of these factories and checking this stuff out? I would love to if anyone in the industry is listening to this, dude, just just open the door for me and I'm in. Clean rooms tend to be pretty close hold. These are like where institutional and trade secrets like really get, you know, baked into the product itself. I've asked TSMC once, they were like, hey, we actually even know who you are. No. And I was like, okay. You know, so you you ask all these companies and they usually like uh that's very close hold. You know, I've seen a few I've seen a few content creators get in and I'm always like, "How what black magic did you pull to get into this?" You know what I mean? >> You were ready to take the next flight out to Taiwan. >> Dude, I'm there. I'm there. If they ever let me in, I'm there. They're like, "We'll give you a tour of our museum and you can see some." And I was like, "Ah." Like, "Sorry, that's not what I'm looking for, you know?" >> No, we need I'm not cool enough. [snorts] >> Yeah. Uh this is a little bit of a broad question, but what do you think is the coolest use or use case of AI that you've seen recently? >> What is the coolest use of AI I've seen recently? >> Yeah, it could be anything. >> I think like one of the things so when people Sorry, this may be a long-winded answer. Um >> okay, >> when people think of AI, right, they think of large language models. They think of like, you know, I'm putting in a prompt and then I get back a maybe even not text, but I'm getting back a website. I'm getting back a video game. I'm getting back whatever. One of the cool things that I've I've started like paying more attention to lately is uh there's two. One is just like uh metamaterials and material design. So people who are like using AI to find like new combinations of like chemistries, you know, ceramics, like all these things to do things that like humans haven't really thought of like ways to combine these elements to do cool stuff. So like the next generation of like shielding on spacecrafts, the next materials for cooling in space, right, which change phases to like dissipate heat and like all this weird stuff. You can find a lot of that that you wouldn't have thought of by using AI to like brute force through this design space and say like, oh, this material, this combination of elements and things here, like this chemistry is like largely untouched by man today. What if there's something cool in that space? Right? And by the same token like medicine I think is going to be another really big uh impactful area of AI because like protein folding right deep mind has um alpha fold 3 which is like doing all this crazy protein folding and looking at all these crazy things that like humans it would take humans forever to do this and the AI can just sit there and crunch numbers and be like oh hey this compound could work this compound could cure this disease you know this thing this thing and it's like so AI is just doing a lot more than just you you know, building websites, answering prompts, customer Q&A, right? Like, you know, there's this whole space where it's just like doing crazy design work that you would have never expected. >> And then for you personally, as someone who has such a such a depth of knowledge of AI, how are you integrating it into day-to-day life? Are you using agents? Are you excited to get a robot? What what are you thinking about? >> I I am probably like the best vibe coder you've ever met, it turns out. So like my so my background is in electrical engineering and I I'm honestly not a great coder like straight up you know when it comes to any particular language I'm mid at best right but I I tend to have like really good ideas and I tend to be really good at like architecting the overall system like hey it should work like this here's the basic flowchart I'm looking for here's what I expect the outputs to be the biggest errors to be here you know like I'm really good at defining sort of the requirements right that's that's kind of where my heads at. And it turns out that when you supply that to AI, AI is incredibly good at building to those requirements. So, I've been able to build systems that like take my video production time way down without sacrificing quality, factcheck every single sentence independently and then larger paragraphs put together, you know, triple check numbers, update numbers without like changing the script, you know, like. So, there's like all these little tricks you can do with AI that make the final product better for my audience without driving me insane, having to go chase down like every individual like number and fact. And then what I get back is a script and a like big research paper where I can read the research, verify all the numbers, do all the things myself, but it got handed to me instead of me having to put it together myself, which is like I mean it used to take me 40 hours to make a single video and I I might be down to like 12 or 16. You know what I mean? So so much of that is just purely mechanical like I need to go find this number. I need to go download this presentation. I need to go find the point in the keynote where this quote got said. You know, it's like AI is great at delivering those things. So, >> makes sense. Are you using more codeex or anthropic? >> I'm on anthropic. So, I'm using cloud code for everything. >> Okay. >> And sorry, go back to you. One of the things you asked was about agents, right? So, like the cloud code is special because that's the part of the cloud ecosystem where you tell it what you want and it's like, oh, hey, guess what? I'm going to spin up a thousand sub agents to go do this thing. You know what I mean? So I don't like use agents and sub aents but the environment that I work in can invoke agents and sub agents. So like when necessary cloud will be like I think I should solve this with agents and it'll go do it. So >> yeah that makes sense. Just one or two other stocks I just want to get quick thoughts on because they're very active in the ecosystem. There was actually a news story today on Nebius that they were uh seeing a bump positive because they raised 775 million in first secured debt financing to accelerate the global buildout. That's a lot of money, right? And I think it goes to your point of this is not a short-term thing. They're not just going to spend almost a billion dollars right away. They're going to get that money and this is a a multi-year buildout. So Nebius and maybe even Iron and some of these other Neo clouds, just how you think about them and if you have exposure as well. Uh, I definitely have exposure. I buy all three. I came out with a video earlier this year talking about all three. It got almost a million views. And I'm only saying that to say like, you know, I I'm very known and very public about my investment in Neoclouds, right? Like, you know, I'm on the record in a big way saying Neoclouds are like a core part of the next generation of infrastructure. And it's not a stretch of the imagination why, right? So like if we rewind the clock to the internet era cl normal cloud computing companies today which we take for granted Google cloud Microsoft Azure uh Amazon web services right like they really are the foundation on which in my opinion society is built right like let's be honest the internet runs through these three companies they power most of the businesses on the planet uh most of the digital businesses I should say right like if you have a website or an online store chances are you're on one of these platforms etc. Right? AI needs a different kind of that kind of infrastructure, right? Not just regular cloud compute that's traditionally CPUheavy, but AI focused compute that's traditionally GPU heavy, right? And so these neo clouds as well as the cloud providers of course, right? As well as Amazon, Google and Microsoft. Um they're investing a lot in AI infrastructure, right? Because every company wants to use AI, but not every company wants to build the infrastructure themselves to run it, right? So they want the Amazon cloud instance that's best for running the AI that they want to run just like they want the Amazon cloud instance that's best for running their website right like or serving their software or what have you right so Neoclouds are just trying to do that natively right instead of taking all of their other infrastructure and all the other projects they do they're built from the ground up to be AI infrastructure providers right so I have a lot of exposure to them I you know when they go on deep discounts. I buy them. And one thing we should probably emphasize here is like long-term investing doesn't mean it's going to go up a lot today. I think that's like when people hear long-term, they think like small company like big upside, right? And it's like that's true, but the thing that's always missing from that conversation is long time horizon, right? So I acknowledge that the stocks are probably down 30% over the last month, let's say, right? 10, 15, 20. Sorry, I'm looking at Micron stock when I say that, but like, you know, >> pick your tell you. I could tell you for Nebius, you know, as of today where we're looking at it right now, we recently hit uh all-time highs, I believe, right at around 300 and we're sitting at 177. So, it's larger than >> Yeah. So, like 40% down, right? >> Uh after after today, it's basically exactly 40. It was down about 45. >> Perfect. So it's like two there's two things right first long time horizon and second things don't go straight up right oh throughout that time horizon you're going to experience volatility and if you consider volatility a risk like we started this conversation with of course that sounds bad that sounds really risky right but like like I said in the beginning of this conversation right after that when I go to the store and there's a 40% discount I'm not upset right so if I'm still looking at these companies and I their best days are in front of them and it's at a 40% discount to what it was a month ago. I I don't know, man. I don't see the reason to be upset. If I was retiring, I I would get it, right? It's like, hey, I put this money in the stock to retire, but it's like, why did you put your money in Nebius if that was your retirement fund, right? Like, you know, and you're planning to retire soon, I should say. You know, like it's the problem isn't that it's in your retirement fund. My retirement fund doesn't kick in for another 30 years, right? So, but if your retirement fund kicks in three, you know, it's probably you're probably more of an index investor hopefully, right? >> Yeah, that makes sense to me. And an adjacent question in the same area and I I just keep getting answer from people, so I feel like you're probably a good person to address this. People look at Neoclouds and talk and you know, New York just passed a rule, right? They're not going to allow these AI data centers specifically to be built even in New York. Other communities are like, we don't necessarily want these by us. How how do you think about the I guess the ethical side, right, of the AI buildout? >> That that is a good question, right? Like this is one of those like I want it but not in my backyard situations, right? So it's like I I get it, right? The market is one thing, you know, you want to put your money where it's going to grow, right? And right now that's AI, but you don't necessarily want like all this AI infrastructure affecting your day-to-day life, right? You know, they build a big data center. you know, it's hogging all the power. Your personal power grid becomes less reliable. I I get it and I don't know where that line is. And it's probably a state-by-st state, municipality by municipality thing. For me, honestly, maybe this is a little naive, but I would be fine with a data center around me. Like, I think they're super cool. I know they create tons of jobs. you know, it's a way to get the tech community because like, you know, all the nerds building all the cool stuff definitely want to live close to a data center because they're going to have high internet speeds. They're gonna have the best services. You know, like there are a lot of positives by being next to a data center. I don't know if you know this, but for a while the edge that the hedge fund would get would be how close it was to its closest data center uh on Wall Street, right? So, it's like if you're if you were a little bit closer, your trades would execute fractions of a second faster and you would get the price advantage over the guy literally further away from that data center than you, right? So, like I know that's largely gone today, right? Like I know there's probably some safeguards against that kind of stuff. Yeah. >> But I [clears throat] think it would be totally cool. So, but I I get it. I Sorry. All that to say, I don't have a good answer. For me personally, I think they're cool, but I absolutely understand why some people are like, "Hey, I don't want that in my community." >> Well, one thing that you could do just as an explanation, because I also think you'd be a good person for this, is maybe dispel some of the myths around water usage and some of the other resource usage that I've seen because I think that that is something that really scares people as this is happening. But, you know, with some of these Nvidia racks and other pieces, it seems like that's that's really improved. >> Yeah, I mean, it's improving on two fronts. First, the chips are getting way more power efficient, which just means yes, like their power usage is going up. But one thing to consider, you know, it's it's not really about like how many gallons of water are being burned off or being it's like what are you getting for that burnoff, right? It's like if you're getting 10 times more compute for 50% more water usage, you should be happy, right? Just like, you know, your iPhone gets a little more expensive, but it gets a lot more capable. you shouldn't be complaining about the cost because as soon as you say, "Oh, but per dollar I'm getting a lot more phone, you know, all of a sudden it's a lot better, right?" You know, one maybe a more direct example is like if I doubled your internet bill, but I 10xed your internet speed, >> are you like happy or unhappy? Right? >> So, it's like I would probably be happy. I'd be like, you know what? Like, if it's if it's noticeable, like I'm probably more happy about the speed than I am upset about the cost, right? >> Yeah. that's happening on a big scale in data centers right now. So back to your actual question. Um so first the chips are getting more efficient but second the cooling is also getting more efficient. The cooling and the power draw and the power systems themselves. Yes, this is a huge problem today because the grid actually isn't ready for like this increased load. Right. One of the big well-known statistics I think is like the American power grid I think it its average age is like several decades old. like we're way behind on infrastructure and our infrastructure is growing fairly slowly per year, right? So like the number of gigawatts we add to the grid each year is fairly low, right? Like single digits or maybe like very low double digits, right? Like 10 gigawatts a year. Whereas these data centers are increasing the demand for that power much much much much faster, right? So that leads to power scarcity which means either places are losing power altogether. It's becoming less reliable, right? Which is bad, right? like we don't want hospitals to lose power, emergency services, like of course there are all these things we don't want to lose power or power becomes much more expensive, which hurts everyone, right? Your your electric bill doubles for something that's out of your control. You're pissed. Totally. Right. >> So I I don't know what the answer is. The what strikes me Oh, go ahead. Sorry. >> It's nuclear in my opinion. >> Sure. Sure. But here's the problem with nuclear, right? If you start that project today, it's not up for 10 years. like the lead time on nuclear is long, right? I'm not saying there's a problem with nuclear. I'm saying the time to invest with in nuclear was 10 years ago, right? To make sure we have it today, right? [snorts] >> So, what I I am not the guy with the solutions. I'm just really good at pointing out the problems, unfortunately. You know what I mean? Like, I I think it's going to be some combination of nuclear, renewables, a big overhaul to American grid infrastructure, right? A lot of investment in it, a lot of pushes by data centers to provide their own power. So, like I know Microsoft was buying their own micro micro nuclear reactors, right, to start being able to supply their own power because like the grid for what they wanted just wasn't capable enough. So, is it up to private industry? Is it up to the government? Is it up to, you know, voters? Like, yes, we're probably all going to have to work together to solve this like nationwide problem of, you know, power provision. >> It's well said. It's well said. I've just had a lot of people that have questions on that and I've kind of wanted a clip that I could send them to be like, "All right, here's some more elaborate." >> I wish I had a more specific answer. Sorry man. >> I think it was good. It was good coverage and also good explanation of the increase in the potential, you know, of these applications as the price goes up and as you know, water usage things like those lines. All great thoughts, really well covered. For those that, you know, listen throughout, appreciate you for being here. We really went into depth on the memory uh side of things and cyclical or not. And I think that that's huge for people to understand right now because that is what is forming potentially an asymmetric trade, right, of you have people that don't understand that or or are convinced that it's not true and you need that to have a good market here? And then we also covered ASML. I thought there's some great thoughts on that and a few other pieces here as well with the Neoclouds. Any final thoughts from yourself before we wrap up? >> Uh I I think the if I had one final thought, it would be this, right? It's volatility is not a risk. It's not bad unless you know your mindset is that it is you know basically what you should be doing in my opinion not financial advice is you should be thinking about the future and how you can invest in what you think will be the future today right like I think it's largely AIdriven so that's what I focus on other people think it's going to be completely different and that's what they focus on and that's great but you know if you let price action sort of dictate your emotions and your mood and your lifestyle um you're going to have a bad time because we are only going to get more and more volatile from here. And so the answer is to like put yourself in a frame of mind where volatility isn't a bad thing. >> Yeah, really, really well said. I encourage everybody to follow Alex, ticker symbol, all spelled out both on X as well as YouTube. You could check out his channel. He makes great videos. How often are you putting out videos now? >> Oh my god, like it should be like four or five times a month. I think it's been like one or two recently just because of all the travel and all the other stuff going on, but I'm getting back into it. All good. More to come. Make sure you're following his content. Alex, thanks so much for coming on today. Really a wealth of knowledge and you just have a a way with words to be honest where you're able to explain things in a way that actually makes sense to people who are not necessarily experts in this industry. So I appreciate all that you do. >> Hey man, thanks so much. I appreciate you having me again and I look forward to the next one. Thank you. >> Me as well. Thank you to the audience for watching today. When we close out, this live stream will turn into a recording. So if you miss any of the beginning part of it, you can just go back and watch that now. Make sure to give it a like and a share. We'll see you on the next one. Take care everybody. Hey there, it's Gav Blackburg, CEO at Wolf, and I'm so excited to see you at the Wolf Summit NYC on August 3rd. We've got the best names coming to this, ranging from Peter Toughman from the New York Stock Exchange to Val [music] from Stock Talk Weekly, and so many other amazing traders, investors, educators throughout the day. It's going to be in Manhattan. We've got food planned for everyone, breakout sessions, and some exciting activities, plus plenty of merch and giveaways. Grab your ticket now. They're just $250 with a public account and you can be one of the few to secure your spot. They're moving fast.