The Next Nvidia? AI Bottleneck Stocks to Buy Right Now

Here to talk about the next big opportunity in AI and which stocks could benefit most is Johnny Wu, founder and CEO of xETFs. Johnny, welcome to the desk. Great to have you. Caroline. Thank you for having me. I'm a big fan and I look forward to the discussion. I look forward to it too. And thank you. So Johnny, let's get right into it. You're betting that some of the biggest winners from here in I could be companies solving bottlenecks in the build out. Is that where we find the next Nvidia. Yes. So I always think about following the cash. Right. So we all know about the spending that the hyperscalers are spending in 2026, over $700 billion projected to be over $1 trillion for the next several years. Right. That's because they need compute. They need to build out these AI data centers. So this AI infrastructure buildout, there's a lot of money going into it. And the folks that are going to benefit from that, the companies that are going to make a lot of money from that are the picks and shovels. Now, a lot of those companies, they can't make their products fast enough, right? A lot of their supplies are all sold out not only through 2027, 2028. Every day you see a headline ASML today, for example, right. Sold out. And they're looking to add to their supply out to 2030 right. So we believe that you need to follow the money right. Those companies that are selling out they have pricing power right. They're able to dictate kind of terms. Right. And they have a moat. Right. You, HBM, high bandwidth memory, for example. Right. You know, Intel can't just spin that up and just launch that. That might take 5 to 10 years, right? So we think that owning the bottlenecks are really, really important, right, to benefit from this AI revolution. Okay. So we know that optical is one because we just had ten ETFs on talking about their laser ETF. So we know that that is one bottleneck holding back the AI build out right now. But tell me where are some of the other biggest bottlenecks and who stands to benefit most from those. Sure. So we've seen in the last year, year to date, especially, significant increases in pricing, right for memory, right for D Ram, right for HBM. Right. So the three three companies that dominate in high bandwidth memory are SK Hynix, Samsung Electronics and Micron. And you've seen their stocks appreciate significantly because data centers can't get enough of that memory. And it's even impacting other areas like the iPhone that uses, Dram. Right. And they they've had to increase prices as well too. Right. So that's clearly a bottleneck, right? Jensen Wang from Nvidia says, guys, the real bottleneck here is memory. But away from that, right. As these days, data centers, you know, get built out, you know, a lot of these chips have to talk to each other as well too, right? So photonics, right, is something that is going to be a more significant bottleneck over time, right? Over short distances. That data can talk to each other over the old school kind of copper wires. Right. But the size of these data centers are like football fields. Right. And in order for the communication to happen, it's best done through photonics and light. Right. And so we're seeing photonics being a bottleneck. Again I'm going to quote Jensen Huang as well. Right. So he invested this year for Nvidia in three more than three. But the three companies I'll mention, he put in $1 billion into coherent, $1 billion in momentum and $2 billion into Marvel Technologies because he needs them to be able to supply, right, in order for him to deliver on his racks his systems as I accelerators to these data centers. So away from all that power is, is is a bottleneck as well too, right? Just because you plan to build a data center doesn't mean that you actually can get it done in the time that you need, right? So it's not only, you know, finding the land, for example. Right. Some of these data centers, they required a power of like powering a small city. Right. And you have to like, connect to the, to the grid. Right. And a lot of these towns and cities, there's a there's a queue. Right. You can't just do it. Right. So if you can solve the power, problem and bottlenecks like that could be, fairly lucrative as well. Right. And so there's companies like, Bloom Energy, for example, where it's less dependent on connecting to the grid, where they produce a few cells, for example, for data centers to use. Right. And so we're seeing that like over and over, right. You have to kind of follow the bottlenecks. Right. And bottlenecks. Tom sees that as well too. That's important. So I should mention this is an ETF spotlight. So one of the funds you have to that you're going to be highlighting here. But one of the funds is your bottlenecks ETF ETFs Eye Bottlenecks ETF which is the ticker symbol neck I see I was just saying you look at all of your top holdings. A lot of the names you've mentioned SK Hynix, momentum, Samsung, coherent, micron, ASML, GE, ver, Nova, which area memory optical networking, photonics power has the most upside do you think from here. So that's a great question Caroline. Right. So if you think about the memory stocks, they have run up significantly over the last year to the extent that those top three or trillion dollar market cap companies. But this week getting crushed. So it's not a buying opportunity. Yes. So the last seven days a semiconductor is broadly right with the news in China on potentially disrupting the fourth largest maker of Dram. Right. You can never predict what happens in the short term. Right. In the long term. Right? I do think that this is an opportunity that you can't miss, right. Memory. Memory. Right. But I do think optics and power right are not the biggest bottlenecks right now. But they're up and coming. Right. So if memory is solved at some point then you have to like overweight certain other areas. Right. And our ability to monitor over 100 stocks, right. Where we have 20 in our ETF, we're trying to create a one ticket solution where we're monitoring the supply and demand dynamics. Right. Pricing dynamics. Don't forget these things can be disrupted ASML for example right. Extreme ultraviolet machines that can cost 200 and $300 million that companies like TSMC need in order to produce these advanced semiconductors. Right. They can get disrupted. They're the only game in town right now. But there are startups that are trying to create, you know, these these machines, right. That could be disruptive of that. Right. So you have to constantly monitor what goes on in the marketplace. But I do want to talk about that too, because what happens when a bottleneck gets solved. Do you see these companies that are benefiting from the shortages today see their pricing power completely disappear? And that and so do we have to look at these as short term investments. So, I do believe that if bottlenecks get solved right, then you don't have the pricing power anymore. Right? And so then that's why you have to continue to review that. But what I will say though, that is that I think this is a multi-year buildout. Right. And it's not just, you know, one year or two years. Right? We're talking ten years, significant longer. Right. Again, I'll quote Jensen Huang because he's the godfather. I, he's doing a service beyond just selling Nvidia GPUs. Right. So this is about a billion people today that use computers, right? We all need compute. Right. But if you think about a genetic AI, which was in earnest rolled out earlier this year, right now you're going to have machines needing compute, right? So, you know, I'm just starting to use a genetic AI myself personally. Right. But like you can have ten assistants constantly 24 hours a day around the clock, like needing compute. Right. And so that could increase the need for compute by five, ten or even more times. Right. That means a lot more data centers that need to be built. Right. So I think these bottlenecks are going to stay here for a while. Right? So I'm not that concerned about the short term. But to your point, some of these stocks, even despite this week's sell off, have already seen huge runs. So how much of the opportunity is still ahead of us and how much is already priced in at these levels? So I'll give you a statistics. Right. I'm a little bit older. Right. So I live through the.com infrastructure buildout. And I think there's a lot of analogies between today's AI data center buildout and 25 years ago when that happened. Right. So, I'll pick three stocks. Right. So in every infrastructure build up they're going to be winners and losers. Right. Back then, there was a company called Nortel Networks. Right. And they were laying down all the fiber optic cables to connect the U.S. to Europe, for example, for the internet to be able to transmit data. Right. They went bankrupt after the financial crisis. Clearly, there are losers. Cisco Systems was a darling back then. The beating earnings every quarter. And you know, they were really important for all the routers and the networking. Right. 25 years later ending September 30th, 2026. Right. Think just beat the S&P by a little bit. The S&P compounded a 10.3% for 25 years. And Cisco did slightly above 11. Right. And then the clear winner which wasn't obvious back then was Amazon. Right. So Amazon over that 25 year period ending September 30th, 2026 was up 833 times, right? And in 2001, you could have bought Amazon between a low of $0.30 and a high of a dollar oh nine. So, you know, if you had bad timing and bought it a dollar oh nine, well, you would have been down almost 70% in a year, right? But if you look 25 years later, the stock today is $259, right? So yes, you would have loved to got gotten at $0.30 instead of 109. So things you're going to have winners and losers. There's lots of volatility right. If you if you can't accept the volatility right then you know investing in AI is going to be difficult for you. All right. Time for a quick break I love having a trip on the calendar, but obviously the trip has to get paid for too. And I find it a lot easier to save for something when I can actually see how close I am to the goal. That's where today's sponsor monarch comes in with monarch. Or I can set a specific savings goal, whether that's a vacation, a big purchase, or just building up savings, I can connect it to an account and actually track my progress in one place. And because monarch brings your bank accounts, investments, credit cards, loans, and recurring bills into one dashboard, you can get a much clearer picture of where your money's going without bouncing between a bunch of different apps. I also like that it separates fixed expenses from flexible spending, because that's where I can actually see what I have room to change. And the AI powered insights help explain what's changing in your finances based on your own data. Monarch is private and ad free. You're the customer, not the product. It syncs with more than 13,000 financial institutions and works across both web and mobile. So whether you're saving for your next vacation or another big goal, you can actually see how close you're getting. Check out monarch using our link below and promo code Street 50. That's promo code Street 50. How do you establish or distinguish between a company that's essential to the I build out, and one that's actually a good investment? That's that's a good question. So folks, always, you know, the two sides of the stories are right. This I build up is going to go forever. Right. And that's the bull case. Right. And then there's the, the the bubble argument, right where stocks have gotten ahead of themselves. Right. Even some of the focus of companies earlier this year just, significantly increased. And they have since sold off. Right. But the market, is a voting machine, right? You get you get data every day. Right? And then you calibrate to what that data is and the market, you know, I think it's short term inefficient, long term efficient. Right. So follow where the cash goes. Who's making money from it. Right. But I'll give you an example. Earlier this year, the hedge fund situational awareness. Right. They had to put out their positions. Right. They were overleveraged. They couldn't meet their margin calls. Right. And Citadel swooped in at the end of July to take down the positions. Some of those stocks, including SK Hynix, including Bloom Energy from their peak, were down more than 50%, but close to that, like 55 or 60%. Now, some of the stocks have recovered, right. But if you go down 50, you have to go back up 100 just to get back there. Right. And they haven't gotten back 100. Some have gone up like 30, 40, 50% or so. Right. But there's there's talks about how do you really change by that much in a matter of day or so. So there's other reasons why. Right. You know, valuations can fluctuate. Right. And every time you get new data right. The other day, open AI right, there is speculation of, oh, we heard that their IRR was $70 billion. Right. And then and then different investors actually it's $50 billion. And the entire market sold off an eye. Right. After the close, right. Momentums. A CEO said, oh, wait, we're sold out through the, you know, 20, 29. Right? And then the next day, lumen come rally like 5%. Right. And so, you know, you have to follow the news, right? And no one knows exactly. And a single point in time what the right valuation is. Right. But over the long term, I think I think you do figure that out. Right? So for me, I think it's really important to, invest in the critical companies, right, that are bottlenecks. And you can identify that by looking at, you know, their supply, you know, where they're they're sold out, how far they are sold out if they're increasing, like, pricing as well. Mike, trying to take an example. Right. A quarter last year ago, I think they made slightly over $11 billion in income, right? The most recent quarter a year later, over $50 billion. Right. So that's significant earnings. Right. But the companies are realizing. But going back to your comparison to the internet build out which I infrastructure stocks today remind you of companies that got ahead of themselves during that time which will appear in one of your ETFs. So so that's very difficult to say. So I think there are there are free lunches in finance. Not many. Right. Diversification is one of those. And then the eighth wonder of the world compounding. Right. And so you know if you had a basket of Nortel, Amazon and Cisco, like the three examples that you use where one went bankrupt, one slightly beat the S&P and the other one crushed the S&P. Right. As a portfolio that actually works over the long term. Right. And so that's why when we think about designing ETFs right. You know, both the two I thematics that you had mentioned, our AI bottlenecks ETF and our AI Korea Semiconductor ETF, they both have around 20 stocks. Right. Because we're not smart enough to pick an individual winner. Right. And so what we want to do is give a portfolio and some diversification about who the winners could be and continue to monitor it. Okay. So shifting gears from the bottlenecks to the, the next fund that you mentioned, the ETFs Korea AI semiconductor ETF, KSM, h why should U.S. investors be looking at Korean chip stocks right now instead of just investing in U.S. companies like micron. Right. So there's a lot of history to Korea, right. You know so so after kind of World War two, after the, you know, the Korean War, etc., the country was just very, very poor. Right? And it was a conscious decision. Right. And Korea wasn't a democracy back then, right by the government. Right to basically invest in certain industries. Right. And they needed to do that, but they also needed to what are called chaebol, the Korean family's right to actually, make those investments. And they were able to do that by giving them government subsidies, government contracts, and then also protect the industries from foreign companies trying to enter. Right. So, you know, into the 60s, into the 70s as they began to invest in technology, right? You know, they borrowed many of the pages from Japan, for example. Right. And investing in like semis, etc.. Right now, if you flash forward that, you know, 50 years or so, right to today, right, you know, the best time to plant a tree. It was 20 years ago, right? So you can't just start up and say, hey, I want to be a PM. Like like, you know, manufacturer. You just can't do it, right? Korea has been doing it for, you know, the better part of 30, 40 years. It's like investing in tech, investing in SMEs when it was unsexy too right. Because there's booms and busts. Right. So this is just a logistics thing. It's not necessarily a evaluations thing. Right. Well ships are better. They're better. So you know I think Korea today right. It's it's it's the golden age. And I'm stealing another quote from Jansen Hwang. Right. Because, you know, the high bandwidth memory is dominated by both SK Hynix and Samsung with micron like 30. Right. But beyond that, those three companies, $1 trillion company. That's right. But in order for those companies to actually deliver on HBM, there's a there's a supply chain behind them. Right. And so when we were thinking about designing this ETF, a lot of volatility. Right. We looked at the two better performing markets in the last year. One is semiconductors right. And the semiconductor index has performed amazing because of the AI infrastructure below. Right. And then two is South Korea. Right. So the broad base, you know, equity index linked to South Korea has been one of the most, best performing geographic markets. Right. So we saw the intersection of those two. Right. Again, volatile. Right. But it gave you more pinpoint, exposure to Korea and what they're really, really strong in AI semiconductors. And so that's why we launched that product. Right. And since we launched it there has been a lot a lot of volatility. Right. You know, at one point it was up over 20%. It's given that back in the last seven days. Right. You know, still up right. 4% or so depending. Right. But you're going to have that right. If you can't stomach volatility then investing in AI stocks is not for you. I know you're going to say, well, you don't want to pick one because you like to own a basket. But if you could only own one memory stock right now, would it be SK Hynix, Micron or Samsung? So of those three, which are all trillion dollar companies, that those are not necessarily my favorite. But I think but I think that they're really, really important in terms of, you know, being an anchor for the portfolio. And they are in both of our bottlenecks, ETF and also our Korea Eye semiconductor. Right. I'm particularly fond of Samsung, right. Because Samsung does so much more than just memory. Right. They're a foundry, right. You know, if you think about your iPhone. Right. Every time Apple sells one of these, these, these folding iPhones, you actually have to pay a fee to two. Samsung. Right. Because, you know, they're deleted behind, kind of like folding technology around. Right. If you think about televisions, right. Like Samsung for many, many years has been number one. Yeah. Right. And their smartphones in market share the latest quarter. They slightly edge Apple in terms of the total number of phones that they're selling. Right. So Samsung to me stands for innovation. Right. And and you know I think I think they their, their ability to kind of like continue to adapt and and and grow their business is is fairly significant. But not your favorite. What's your favorite. Look. That is my oh that is that's my Samsung is my favorite okay. I thought you said those three aren't your favorite. Oh those three aren't my. So if you ask me those three Samsung is my favorite of those three. If the world is your oyster. Yeah. So, look. What's your favorite? We want to know Johnny Woo's top stock right now. Well, memory is not just, HBM, right. But my reason why they're not my favorite is because of, they've run up so much. Right? I still think they're going to be market leaders and, going to sell a lot of HBM, etc.. Right? So if you look at, you know, some of the other memory makers outside of HBM, like Nand for example, right. He Oaks here, right, based in Japan, for example, right now their stock has run up. Right. But they're still on a relative basis from a market cap perspective. Lot a lot smaller to me is it's math right. So I'll give you, you know, statistics. 25 years ago the top ten companies in market cap included GE for example AIG, right, ExxonMobil, Microsoft, 25 years later, nine of those ten companies underperformed. The S&P. The only one that beat it was Microsoft and Microsoft. Was up 15% compounded for those 25 years ending September 30th of 2026, S&P was up 10.3%, but S&P would be 11.6 times your money right over 25 years. Microsoft would be, like 33 times your money. Right. And so why I tell you that is it's hard to stay on top right. And large companies for them, the double, triple, etc. it's just harder, right. Versus the three companies, in the Mac seven right back 25 years ago, three of them were IPO yet, right? Like Google and Meta etc.. Right. But four of them were around. Right. The three companies that did the best, Nvidia, Apple right. And Amazon, all three of them were under $6 billion in market cap. Right. And and if you had $1 million in each one of those, each one of those would have gotten you, multiple times over. Right? Amazon 833 million. And both Apple and India would be over $1 billion, right. So like, you know, so I do believe that you have to invest in some of the smaller companies, right. That are bottlenecks. Right. That can increase supply. And they can just, you know, accelerate from there. And we can get exposure to those smaller companies through both NEC and KSM. But if I'm considering both, what am I getting from one but not the other? Sure. So so the bottlenecks is basically looking at not just memory, right. But photonics power companies. Right. Semi and compute. And we have like the, the the monopolies. Right. We have TSMC and we have ASML. Right. And we have Broadcom right in the eye bottlenecks. Right. So it's a more diversified portfolio across the various bottlenecks. Right. And and we think that's important right. Versus the when you if you considering Korea if you believe Korea is in its golden age right. And it's really really important to the whole memory trade. Right. You want the companies that actually are in the supply chain in addition to the 2 to 2, kind of like, you know, a pillars of, of of memory, right, of SMEs in Korea. Right. So obviously you will have Samsung Electronics and SK Hynix. Right. But but the smaller cap companies, right. 15 of the 20 that we have, right, are under $10 billion in market cap. Right. And so a lot of these companies supply what both SK Hynix and Samsung need in order to deliver on memory. HBO okay, I'm sure that they're we're almost out of time, but I have two questions. And I want to get to where this are that I'm sure that someone, a viewer, will comment. But the expense ratio, why should I pay a fee? Why don't I just look at your holdings and just buy them myself? Sure. Yes. Look, if if you're an investor and you can do the homework, you should pick the stocks on your own and not pay anyone fees. Right? And if you don't have time to do that, right. Then I would argue you should just buy an index, for example, right? You can't go wrong. Well, I shouldn't say you can't. I can't predict the future. Right. But if you look over the last five, ten, 15, 20, 50 years, the S&P 500, it's really, really hard to beat. Right? I gave you the example nine out of ten companies, the largest market that couldn't beat the S&P 500. But if you want to tilt right to I for example I'll give you another statistic. The last ten years has been amazing for the equity markets. Right. Instead of 10.3 S&P has compounded a 15.3. Right. The Nasdaq has compounded at 21.2 right. But semiconductors has compounded at 34.5. Right. So you're talking about in ten years. Yeah a little over 4 million for the S&P. 6.8 million for the Nasdaq. Right 19 million for semiconductors. Will that continue. Who knows right. Personally I'm a believer of the I trade and the I build out. So I do think you want exposure to you know companies that benefit from this build out. So whether that's bottlenecks or semi's right. Or a higher data version Korea I said meetings right. And so we can't predict the future. But we do think that we want to give investors that tool and that portfolio. And we have to monitor all that. Right. We do the research and we do charge, in the case of Csam age, 65 basis points in a case of I bottleneck, 75 basis points. Right. But you know, we do the research right. We hope that there is value add to that. Right. And it's just a one picture easy solution for investors. Okay. We I kind of started things off by saying is that where we find the next in Nvidia and you had a whole bunch of Jensen Wong quotes. So he's still very relevant, but I don't see it at least in the top holdings. Nvidia obviously, you know, or at least not in the bottleneck. Right. One the hour. Nvidia's best days behind it is still relevant as we think about significant upside from here. Sure. So the last ten years Nvidia has compounded over 60% for a year. Right. You know, back then, you know, it's funny, when I got my son into investing, he bought a stock called Nvidia. And I said, well, what is that? Right. And he did it. Okay. Okay. Right. And it was like the. Why we talked about Nvidia. Back in the day. Yes. Yeah. And so but now it's like a $5.5 trillion company. But I think the importance of Nvidia is that there are market leader not just, you know for selling their chips, but for the overall industry what they're doing. And in terms of creating systems, AI accelerators, right. Racks. Right. To make it easy, right, is that they're building the ecosystems and all the suppliers that are needed to deliver. Right. And so I think they're going to continue to be super important going forward. Right. But here today. Right. And and I I've talked to folks they're like, oh I'm supposed to I have Nvidia you're dating Nvidia is up like 20%. Yeah right. Some of the names in our basket right where they're partners who are Nvidia. Right. You're creating these I you know accelerators etc.. Right. The momentum to coherence to Marvell is right. Some of them are up over 200% for the year. Right. So if you think you're getting exposure to I by just owning in media. Right. You know, there's just a lot more to it. Right? And you either have to do the homework yourself, right? Or you invest in a manager, right? Or you invest in an ETF that's actively managed. Or you can do some, an ETF that is passively, indexed as well too. Right? Remember every passive index, there's a committee it actually selects. It has a methodology. So people are doing work behind that. But as an investor would you still buy Nvidia here. Personally I own Nvidia. Nvidia. Right. I you look I think Nvidia is still a market leader. Right. But I think it's just math again. Right. It's harder for them to compound in the future. Where it did in the past. I'm not saying that it can and I hope it doesn't. Right. But, but I do think that some of the other companies, just have, the potential to have more upside. Okay. I think this is a great time to pivot to our rapid fire. Right. If this or that. Quick questions, quick answers, no hedging if you can help it. Are you ready? Yes. Micron or Nvidia. Micron. Micron or SK Hynix. SK Hynix. Broadcom or Nvidia. Broadcom. Loose momentum or coherent. Momentum. TSMC or ASML. TSMC. Applied Materials or Lam Research. Lam Research. Memory stocks or GPU stocks for 2027. Memory stocks South. Korean chip stocks are U.S. chip stocks. South Korean chip stocks. I infrastructure early innings or late cycle. Super early innings. I CapEx in 2027. Accelerating or peaking? Accelerating. Biggest I bottleneck memory or power? Memory. More upside in 2027. Chipmakers or equipment suppliers. Chipmakers. Memory stocks after this recent pullback. Buying opportunity or warning sign. O buying opportunity. AI stocks from here chase the rally or wait for a pullback. Well I I believe in dollar cost averaging right so you can never time it. And so the market gives you opportunities to buy. And so like like I said 25 years ago Amazon $0.30 or dollar oh nine. If you continue to buy it and accumulate it that's a good trade. To buy a little at a time here. Yeah. Finish these sentences. The semiconductor stock with the most upside from here is. Wow that's a that's a good that's a good question. I'm a bit stumped by that because I think there's just so many that have like high potential. Right. That's why I would argue for a basket like the Korea semiconductor basket. Okay. The most overlooked AI infrastructure stock is. So a lot of people don't consider GE ver Nova. Right. And I thought right. But they're supplying all the gas turbines at all these data centers need. Right. You know I think companies like that are really, really important. Right. And the, the investing public will realize that longer. Okay. The I stock I'd avoid at these valuations is. Oh my what's up with my mouth? I'm, I'm, I'm such a bull on I in general. Right. That, I don't know if I would say it in I stock today. The biggest risk to the AI infrastructure boom is funding. Right. If they don't get funding to actually build out all these AI data centers. Right. No company can do it alone. Right? Nvidia is taking a leadership role bringing in private equity firms, investment banks, right, to lend money. Right. But if they don't get the funding to actually complete the day eye data centers. Right. That could that could be a ripple effect, right? Yeah. Okay. And then if I could only own one semiconductor stock for 2027, it would be micron. And just quickly, one thing I didn't ask is anything about software. Obviously you're not necessarily they don't fit into the funds. But are software stocks an opportunity here. So earlier this year you had to assess properly. Yeah. And, and you know, everyone was, thinking that, okay, I was going to like, you know, disrupt software so significantly. I think software is really important to the eye. Right? So in fact, a subsector of software is AI software, right? The companies that are basically re-engineering their business to take advantage of AI. Right? So companies like Palantir, right. Like just amazing. Right. Like, you know, and it's not just them. Right. Like, you know, Salesforce like sold off hard and and has since rallied back. Right. But software is really, really important going forward as well to. Give us one that could be eaten by I. O the well the ones that that don't evolve. Right. It's hard to say as well. Okay. Palantir Salesforce. What's another what's a third that you would say will still be around despite AI evolution? Just trying to squeeze one more, pick from you. Well, my favorite is Palantir, right. Because the government is behind it. Right. You know, they're CEOs, you know, a little bit eccentric, right? He you know, he just goes on TV and and he always has an opinion. Right. But they're just so critical, right, to folks that are adapting to AI. Right. And you know, if I think about the future and I think about like curing cancer, right. And I think about AI playing an important role to that. Right? It's going to need software companies that actually can like take the AI and actually make that progress. Okay. Yeah, that's a conversation for another day. So we hope you'll come back. We really appreciate your insights a wealth of of knowledge and Jensen Wong folks. So thank you so much. Thank you. That's Johnny Wu founder and CEO of X ETFs. If you enjoyed this ETF spotlight, check out our full interview with Yuri Khodjamirian. He talks more about the AI bottlenecks and also how investors can get access to anthropic pre IPO.