The AI Boom vs Local Backlash: What Investors Need to Know And One Trade for Both

Hello and welcome to being exponential. Today we are doing our macro episode. Now Luke, we spent a lot of time worrying about whether the AI boom runs out of chips, power, or money. But what if the resources were actually running out of is public support? If Americans increasingly don't want these massive data centers in their communities, could politics, not technology, become the real bottleneck for the AI boom? And I want to lead that into my first question for this discussion is what happens to the AI capex thesis if we start getting more regulation in in regard to data centers. >> Dude, have you been listening to our podcast or something? >> Like have you been It sounds like you're really plugged in to the the information flow what we've been saying here because that's exactly I mean right now it's super topical and right now everybody wants to talk about it. Everybody is talking about it. Texas is going after data centers. Pennsylvania is going after data centers. New York's going after data centers. It's a big issue in Ohio. So, right now it's super topical. We're going into midterms. I think there's a reason it's super topical going into midterms. But this is something we've been talking about for I want to say over a year now. The biggest risk we keep coming back. We don't think, you know, power shortages or memory shortages or the fact that AI is not going to show up in terms of ROI and the revenue is not going to show up or it's not going to be profitable or collapsing token prices. We think all of that is uh hocus pocus. It's it's it's a load of crap really. Um all those fears are wrong. They're misguided. They're shortsighted. But the one that is not and the one that we've talked about as kind of the thing that could pop the bubble, the thing that could end the whole trade, the thing that could end the capex buildout is a manufactured slowdown via anti-data center, anti- AI legislation, which is now happening. So what do I think about it? It's the one risk. I'm an AI bull. You know this. Everybody listening knows this. I think it's prudent and responsible as an AI bull to spend most of my time thinking about what could go wrong. I've thought about all the things that could go wrong. I still sleep well after thinking about all of them except for this one. This is the one worry that does keep me up at night. This is the one worry that I cannot fully wrap my head around as okay there's a there's a way to navigate around this long term. I think in the short term there is but in the long term I I'm not sure there's a way to navigate around it. It almost feels like this is the inevitable risk that does eventually uh cause this whole trade to to unwind. Now the good news is I don't think that's going to happen for a while. But it's a very interesting and nuanced argument pro data centers versus anti-data centers. And I I think it's interesting enough and nuanced enough that I can see why a bunch of people are um are protesting them. And I can see why that movement could gain a ton. It's gaining momentum right now, but gain even way more momentum and why it's bipartisan. I mean, Abbott down in Texas is is going in on data centers. So is Kathy in New York, right? Like these are people on different sides of the political spectrum, very opposite one another. and they're, you know, Shapiro in in Pennsylvania and then Dantis down in Florida, right? Like these are people that don't agree on anything except for this. So yeah, it's a it's a risk worth talking about a lot because I think it's the most important thing out there right now. >> 100%. So I I want to in this episode I want to get kind of get into technicals for what an investor should be looking for. So let's uh stick with the hypothesis that the data center projects get delayed and not necessarily shut down. Which part of the supply chain do you think is going to get first that maybe we should get out of and then possibly into another uh sector. >> So let's let's just set like the the basics for what's going on. So the basics for what's going on is people are against data centers for a a variety of reasons. raising power costs, um water usage, uh you know, this kind of all the profits going to to big big tech concerns are misguided and shortsighted. And I do think, you know, history shows us we shouldn't stop technological progress, but rather learn to adapt with it, learn to adjust to it. And if that means and if that means passing legislation that allows uh us to do that, like maybe we have to slow it down a little bit but then still allow us to adjust like that that's fine. What you don't want to have happen is just this kind of cold this stone cold stoppage of of building data centers or there's just you know we're going to stop all building until we can review things further. Like that's not what you want to do. You want to have some adjustments. You want to think about this smartly and it's it's not black or white. It's 50 Shades of Gray just like all of life. Um, so that's that's how I see things. But how I see things doesn't matter. What matters is how the public sees things. And how the public sees things is uh very anti-AII. If you look at any sort of survey, in fact, um I'll send you these charts so you can pull them up on this podcast and that people can see what I'm actually talking about. But there was a recent survey that came out to show kind of the public support for for data centers um for AI for things of that nature and it's it it's terrible. Um if you look at let's say Republican support for data centers, the question asked is this heat map news. The question that was asked was as you may know data centers are facilities that house the servers that power the internet apps and artificial intelligence. Would you support or oppose a data center being built near where you live? A year ago, August 25, 14% of Republicans said, you know, they they would support or uh the the net review on it, the net rating on it was a positive 14%. The net rating on independents was a minus 5%, the net rating from Democrats was a minus 8%. So pretty much hugging the flat r flat line, right? You take the average of plus4 minus5 and minus 8 you're right around zero. Today they asked that question August 2026 net rating for Republicans of data centers minus 43%. So we've gone from +14 to -43. Independents minus5 to minus 65. Democrats -8 to -75. You take the average of those and we're at like basically minus 60. So we've gone from flatline. People don't really care all that much. people really really really don't like this stuff. So that's what matters. Not my opinion on it, not your opinion on it, not you know Elon Musk's opinion on it, not Trump's opinion on the matter really. Uh what matters is the public opinion because the public opinion is what politicians go. This is how populist movements work, right? Populist movements, the people feel something, politicians recognize that and they capitalize on that sentiment. And that's what you're seeing happen. And that's why Texas is doing this. Pennsylvania is doing this and all this all these actions are taking place to slow this movement. Now does that actually slow the AI trade? Right now the answer is no. And the answer is no because the legislation being passed is very peacemeal. It is not consistent across you know a variety of places. So there are nine states in America where most of the data center construction is happening. You got Virginia, you got Texas, you got Georgia, you got Illinois, Arizona, my backyard, Ohio, Indiana, Louisiana, and North Carolina. Those are the big nine. Pretty much all of 90% of the data centers being built in America are being built in those nine states, both existing and the pipeline for future construction. So if we take a look at those nine states, those are all the only states that matter. What is what's happening in those states as far as legislation is concerned? Well, only one of those nine states, Texas, has a statewide construction freeze. Has passed legislation to actually freeze construction of new data centers. Abbott ordered that that Urkott um and the the PUCT to audit every data center project in the interconnection process. No project can move forward until it is audited. That happened in Texas. We haven't seen any statewide legislation in any of the other eight major states where construction of data centers is happening. Virginia has enacted I think it's a a penny basically I think it's 011 cents uh per kilowatt hour a data center electricity tax. So that makes it a little bit more expensive. Uh the county of Lden eliminated by development which is now requires uh public special exemp uh exception approval for data centers. Um a broader uh pause in that county is also being studied I believe but nothing has actually passed. So and then off in uh Georgia I think Atlanta's done something but the rest of uh Georgia hasn't. Uh Chicago's mayor has ordered an enhanced review and and requested a moratorum, but the city council has not yet enacted it. Statewide, Illinois uh has stopped accepting new data center incentive applications, but is still welcoming data centers and they're still building. Uh Arizona froze data center tax incentive qualifications through 2029, but they also are still building and still welcoming. Ohio, there's there's no statewide moratorum. Uh I think a little bit over a dozen maybe like 15 16 uh different municipalities have adopted temporary pauses. Um but not you know again no statewide thing. So it's very peacemeal. It's it's here it's there. It's here. It's there. It's here. It's there but it's not statewide stuff outside of Texas. So from that perspective the legislation being enacted today is more concerning in terms of what it may signal about what may come as opposed to what is happening right now. Um, the reason I'm not concerned over the next two years about this, or at least I'm I'm less concerned, uh, than I would be otherwise, is the federal government is still highly supportive of of data centers, right? Trump has come out and said, you know, uh, this an data center movement is dumb, stupid, communities should have them. They produce a lot of tax revenues, etc., etc. So long, so you have this this tug war between federal support and state kind of rejection of them. Uh so long as that tugof-war exists, then data center uh operators, the hyperscalers can navigate, right? If Texas blocks them, okay, screw Texas. We're going to go to Arizona. If Arizona does something, we're going to go to Wyoming, whatever it may be, right? Like, we can reroute. Hyperscalers can reroute so long as the federal support is is there. That changes if the federal support backdrop changes. And that's why to me, this is a 2028 risk, not a 2026 risks. Right now, everyone's talking about it in the midterms or coming into the midterms because it's just such a topical issue, but there's not going to be a change of power at the federal level, right? Trump's going to stay in the White House. Republicans going to own the the executive branch. Uh still pretty much own the Supreme Court, if you will. You might get a a flip of the House. You might get a flip of of uh the Senate, though betting markets are kind of 5050 on that. But even if if there's a blue wave, you know, the the executive branch still stands as as purely Republican. And so from that perspective, I don't think that that's enough power shifting hands for the federal backdrop to meaningfully change. That's a 2028 risk. And so that's how I'm looking at things right now. But if that federal backdrop does change and this populist movement stays uh very anti-AII and the states continue to enact legislation like Texas is the tip of the spear and Illinois does something similar and Arizona does something similar and Ohio does something similar, Indiana does something similar and all those dominoes fall over the next you know 12 to 18 months and then two years from now we elect a different president that has a different viewpoint on uh data centers. Then 2029 is set up for a year where everything kind of crashes. So, that is the worry that keeps me up at night. Sorry for speaking a million miles an hour. I've I've looked at this a lot over the last uh week or so to really dig into the details and the numbers and what's going on. >> Yeah, 100%. I'd just like to add on there. It it does appear that New York does have a statewide moratorum, although that isn't all data centers. It is a a very specific section. And then you don't Yeah, New York. There are other states that have done things and those states don't matter, right? The states that matter are nine the nine states where the construction's happening. Like if Virginia did that, that'd be a pretty big deal. If Ohio did that, that'd be a pretty big deal. If Indiana did that, that'd be a pretty big deal. If Arizona did that, it'd be a pretty big deal. Which is why when Texas did it, so New York's had that thing forever and ever and ever and ever. and the market didn't really give a crap about it because >> but Texas all of a sudden I think was it last week or two weeks ago was when Governor Abbott came in and and did his thing and all of a sudden the market did pay attention to that. The market did care about that. So what you have to focus on are those nine states that I talked about and again those nine states are Texas, Virginia, Atlanta, Illinois, Arizona, Ohio, Indiana, Louisiana and North Carolina. Um, and only one of those states so far, Texas, has passed statewide legislation that actually bans new data centers from being built until, you know, they they are approved. >> Yeah. And the unfortunately the largest state there. >> Yes. Which again, which is why it's it's a very real um it's a very real risk. Uh but again, I don't think that this changes the calculus for the AI buildout, the AI infrastructure buildout today, here and now because of the things I just said. It's it's a tug war between federal and state. Only one out of nine of the critical states have enacted anything. The rest is kind of like, you know, city by city, county by county, and the big tech companies know how to reroute around them. And then also can't forget that these guys have so much they can to make sure these things do not you know legislation like legislation Texas does not become >> more common more widespread and so you just you gota you know my friends who live in DC always tell me you got to follow the like it's like Wall Street you got to follow the money unfortunately in DC you also have to follow the money and the money isn't uh in the anti-data center movement the money is in in the big tech world. And so I think until there's a change of power at the federal level, like I said, that's a 2028 risk. And until that happens, I think that big tech keeps building, not, you know, without any red tape. There obviously is red tape coming in in places like Texas. So that does mean higher cost. That does mean projects probably get delayed. That does mean the economics deteriorate a little bit, but not enough to derail the trade. What it means going forward is the AI trade over the next two years probably slows and promptly becomes a bit more selective but it doesn't mean it stops. Gotcha. Yeah. I mean it is it is really easy to see from a populist perspective why this could be getting more attention. One of the big stories that, you know, is popping up recently is this idea of anthropic, I believe, is buying tons and tons of books to cut the spines off of them, scan the books, and then throw the books away. And then it just kind of like screams back to like uh Fahrenheit 451, right? This dystopian future where big tech is destroying the the written language or, you know, making a worse world for people. So I I think it's easy to see why uh people are becoming more and more anti-data center. >> Okay. So I nothing is I used the word inevitable earlier but I I probably should track that nothing is inevitable. Nothing is guaranteed to happen. Things can always change. People change their minds. The media changes its focus. Uh politicians change their incentives. Right? Like things change. People change, attitudes change. But if you were to come into my office right now, put a gun to my head, and say, "Well, this movement killed the AI trade by the end of 2028," I would say yes. >> Because the cat's out of the bag. Like you kind of just mentioned, there are definitely reasons why people are angry about the construction of data centers. Yeah, there are. And it's such the reasons are multiaceted because and so when you look at a populist movement like if you look at sort of the the rise of Trump I think in in 2016 right um there were there were multiple tentacles of of that appeal that allowed a bunch of people who were not unified to feel unified behind a single move or a single cause. And that's what what led to the whole MAGA ideology, MAGA platform. I see strong similarities with the anti-AII, anti-data center movement. There are so many tentacles to it because you're drawing in the environmentally conscious people who were like these things are destroying um our environment. They're using so much energy that we're repurposing, you know, federal land where animals are living to build these data centers. So you you're drawing in that crowd. You're drawing in the costconscious crowd because in communities where these data centers are being built, you know, energy costs are going up. >> You're bringing >> Yeah. You're bringing in the conservationist crowd. That's, you know, you know, why are we building so much and we don't have enough resources to to build or sustain these things? Um, you're bringing in the uh the anti- tech crowd, which is a pretty big crowd right now because the leaders of AI, we've talked about this in this podcast before. They they they messed up, man. Like, Dario Amade can't come out here and say, "What I'm building is going to take all your jobs," and not expect the public to not like it, right? Like if I create something >> and go out and say every single day, every chance I get, this is going to take everyone's job, then everyone's not going to like that. >> Like, whoa, go figure. People don't like AI because you told them it's going to completely replace their work. And like I said, Elon Musk tries to frame it as like, oh, this postwork utopia, no one has to work and it's going to be wonderful. Nobody believes that, right? Like maybe you want to believe that. Maybe there's an optimist in you that wants to believe that, but deep down you're like, "How's that gonna how's that gonna work? That doesn't make sense to me. That kind of sounds like a version of like significant communism, right?" Like, so it all of a sudden it's kind of like there's just uh they messed up. They fumbled the bag. They they have this great technology and in order to in why do they do it? To inflate their valuations, right? The more that Daario got people to believe that AI is going to substitute human labor, replace human labor, the more he could justify a higher valuation on anthropic. Same with Elon Musk and SpaceX. Hadn't SpaceX become a$ two trillion dollar company. It was a rocket launch company with a less than trillion dollar valuation, but then it merged with XAI and then we launched orbital data centers or talked about AI satellites and all this stuff. And it's like, well, you know, Elon talked about labor replacement as a big part of that story. And if you know we have these orbital compute uh uh capabilities, we can create AI that's going to have these humanoid robots and these self-driving taxis and all this stuff all around the world that's going to play so much labor and so much productivity gains a lot of deal. these guys did it to inflate the valuations and I actually don't think it's inflating just to boost the valuations because I think their their fears are actually true uh to boost the valuations of of their companies and now they're looking at the the back side of that the underbelly the dark side of it and it's it's pretty pretty scary and I think another interesting story that popped up this week is maybe you saw so Stanley Ducken Miller is a very famous investor >> and he wrote this piece in the Wall Street Journal he used to be Scottison 's boss. He wrote this piece in the Wall Street Journal attacking Treasury Secretary Scott Bent for his uh increasing of buybacks of long-term Treasury yields, calling it, you know, artificial and basically it's going to cause a crisis li. The point of the article is not important. The point I'm trying to make here is that the article, as it turns out, was written by AI. So, this piece was super well read. Uh, it was widely read. I mean, it was very uh people reposted it across the internet and praised it as being super thoughtful and then 24 hours later, it turns out was written by AI. So, just let that let that sink in because that's a very important data point in this whole conversation. Yes, AI is a risk of human labor. The public is waking up to that and now they're pushing back against AI. That is a total natural and human reaction. And that's why, like I said, gun to my head today, I do think that this this kills the rally in in two years. But it won't kill the rally until the federal power changes. And Wall Street is greedy as can be. That's something we also can all agree on. And so they're going to push this game to the very last second. They're not going to get out early. They're going to push it to the last second. So this bull market will live on to 28 for sure in my opinion. But where I'm concerned is is what happens after. Gotcha. I I want to ask kind of an inverse question here. Uh something we talked about in the last podcast is uh investors aren't necessarily short-sighted, but they have a short memory. So, a lot of things from the average person, average citizens perspective are happening higher than normal gas prices. We have all of these recalls happening. You can't, you know, buy eggs or blueberries. So, is it potentially just a case of everything's going wrong and one thing I can um pit my vengeance to is data centers, right? Cuz another I don't want to say it's similar, but fossil fuels have been around for decades and we've had reports of global warming and all of these things for decades and our fossil fuel usage hasn't necessarily changed. So, do you see a case of just this is bad press for the moment? >> Absolutely. Um, society has I don't I don't think it's an American thing. I I think it's it's a global thing. Society has lost its sense of accountability. Like individuals and companies, they don't want to have accountability. They just really don't. And so when something goes wrong in a person's life, I don't want to say nine times out of 10, maybe seven times out of 10, they look for somebody else to blame, right? >> As opposed to like they didn't work hard enough or they didn't, you know, study long enough or they whatever it may be, right? They look for something else to blame. Um the blame game is such a such a big thing. And so yeah, finally now we have this big kind of amorphous thing, artificial intelligence that is somewhat tangible but also very intangible at the same time. So it makes it pretty easy to blame. Yeah, screw those data centers. Screw AI. You're not even blame person at that point. You know what I mean? You're not even blaming a person. You're blaming an idea. You're blaming a thing. And by removing the person from it, you kind of allow people to get, you know what, I'm not even, you know, screw these machines. It's man versus the machine. >> And so, yeah, I do think that what is contributing to this movement, absolutely, is the fact that consumers are struggling. Americans and households are are struggling. What's really interesting is that that same heat map news, and I'll send you this chart, too. I don't have it off the top of on one of my screens right now, but that same survey found that American support for data centers is the lowest in the developed world. So, it's not just that people hate data centers. It's that Americans in particular hate data centers more than Germans, than Brits, than uh Chinese, than Japanese, than South Koreans. South Korea actually had one of the highest favorable ratings uh of data centers in AI. Germany actually had a really high one. There's another big European country that had a high one if I can remember, but America was the worst, >> right? So Americans hate everybody hates data centers. Like by and large data centers having negative and AI has a negative approval rating by the public, but Americans have the the worst view of them. And I think that again that speaks to the situation in in America that I do think people are are are struggling in terms I mean consumer confidence the conference board uh consumer confidence survey just came out and it hit a seven-month low for August. Um, we're seeing that people's expectations about their income over the next 12 months versus inflation over the next 12 months, that gap is getting ugly. Inflation is outpacing income growth right now. And based on the survey work, people expect that to worsen over the next several months, uh, the next year. So yeah, I mean I do think that the the current economic conditions and we've talked about it the whole thesis that we have predicated our trades on our model portfolios on we've called it the AI bifurcation. That's what we've called it. You have AI winners on one side and everything else losers on the other side. And that has been the reality expressed in the markets over the last three years. Finally, the other side is like screw this and and and and fighting back. So that's I mean that's what's going on. And so it's a very human and natural response. And I'm actually not even mad at the I think it's again I think it's misguided. I think it's shortsighted. But I'm not mad at the people for feeling the way they feel because the reality is over the last three years, one group of people in the economy has won and the other 500 groups have lost and and the the market is has been that way too. The market has been carried entirely by AI stocks. The market has been carried entirely by data center stocks. It's been carried entirely by semiconductor stocks, the hyperscalers, the neoclouds and all that stuff. And what's what's been dragged down is Lululemon. What's been dragged down is Nike. What's been dragged down is Dix. Dix was down 31% uh in a single day this week on on a big earnings miss. Uh what's been down is into it. What's been down is Service Now. What's been down is Atlasian. What's been down is is everything else besides the AI blast radius. So again, I think it's misguided and shortsighted but I get it. >> Yeah, I get it. Nike is down what, 80%, Walmart is down and then Elon Musk becomes a trillionaire. So, it's really easy to pinpoint your frustration at, you know, one guy in this case. [snorts] >> Well, I think yeah, you I mean, so I don't know if it's an it's not an anti- Elon movement at all. I I think it's an anti- AI movement. Yeah, when you remove the person from it and it becomes amorphous is when you start to get both sides of the aisle on it. Um I mean you're you're seeing a lot of Republicans of again that survey the Republican supports minus what 44 and Democrats are minus 35. So Democrats obviously dislike it more but Republicans dislike it a whole bunch too and independents dislike it a whole bunch too. So it's not just the left or right thing. Uh it's it's bipartisan and that's that's the scary part if you're an AI bull. politicians can't agree on diddly squat right now and they haven't been able to agree on diddly squat since COVID and here they are agreeing on this. So that should scare you. Um but again I don't think this is a risk that manifests into an actual reason to sell AI stocks in 2026 or in 2027 or for most of 2028. And I actually think the party in AI stocks continues. The AI bifurcation gets wider and that's what causes some big populous swell in the 208 elections. Um I also think again the timing here is we're what a month or two months away from midterms >> or a month [snorts] month and a half. Yeah. Right. Like >> now we're all a sudden passing legislation to ban. It's like come on now guys. It's all it's all smok and mirrors too, right? It's also politicians playing their games where it's like, "Oh, now I got six weeks. We're in the home stretch. What do people hate? Look at that survey work. They hate data centers. Let's go after let's win some votes." Right? So, I'm pretty sure a lot of this calms down after the midterms, too. Uh, long story short, I think it's a buying opportunity. The AI trade has been stuck for three and a half months. Whether you look at AIQ, the AI ETF, whether you look at the NASDAQ, triple Q's, the semiconductor ETF, SMH, uh, or socks, whatever, whatever index benchmark you want to look at for the AI trade, they're pretty much all where they were, uh, in the first two weeks of May. So, we've been stuck in neutral for three and a half months. Some people are calling it, you know, broken. That's what Jim Kramer called the data center trade broken. I think that wrong. Um, inverse Kramer for the win. No, I I respect him, but I think I think he's wrong on that. Um, and I do think that this is just a fatiguing and eventually we're going to we're going to wake up and this trade is going to come back to life. And I just found actually in the last week a single ETF which in my opinion best encapsulates this trade and that ETF is AIS. So, AIS is this, it's the Vista Shares Artificial Intelligence Super Cycle ETF. And the name kind of says it all. If you are a believer in the AI super cycle, if you are a believer that the capital expenditures uh for artificial intelligence are going to continue. So right now I think the numbers are the hyperscalers are going to spend $850 billion on um capital expenditures for AI compute this year. That's going to march towards 1.5 trillion or so next year. Uh and we're probably going to be stabilizing around there. If you're a believer in that idea and that trade, this is the ETF for you. And I'm a huge believer. I think that's the heart of the trade. It'll remain the heart of the trade. That's the trade that'll work until 2028. Um, and if you want, you know, there's a lot of stocks. You talk about a lot of different ways to play this. A lot of people sometimes just want that one. Give me the one thing I should do, the one name I should buy. Set it and forget it. If you want that, for me, I think it's AIS, the Vista Shares, Artificial Intelligence, Super Cycle ETF. It holds all of the companies that are the capex takers, the ones that are on the receiving end of these massive uh compute checks. Whoops, my mic. Sorry about that. Uh, Micron's a big holding. So they're they're top 10 holdings. Let's go. Um SKH Highix, you know, a big memory play. Micron, AMD, Taiwan Semiconductor, Silicon Motion, Veritative Marvel Intel G Vernova Nvidia, and then that's the that's 43% and they have 69 stocks or 66 stocks in the whole thing. And they got uh you know 60% to to other AI infrastructure plays. That's an ETF that, you know, if this is indeed a buying opportunity, and I think it is, this is, and you want one way to play it, AIS to me is kind of that one diversified big high octane way to play the super cycle. >> I like that. It seems like it's pretty weighted in energy. Uh, is that safe to say? >> Uh, it's mostly weighted in memory. I mean, Micron and SK are 16% of the portfolio by themselves or 15% of the portfolio by themselves. Then you throw in like you know Seagates in there and Western Digital's in there uh Broadcom's in like so if you kind of add up all of these kind of memory network or memory and storage plays I think it's like 20% of it. Um but the way they create the ETF uh is they look at the AI stack and they look at the economics of the AI stack and they kind of decide okay if there is a dollar going towards AI compute 30 cents is going towards this sector 15 cents is going towards that sector 11 cents is going towards that sector and they build the ETF with similar weights and so that's why memory has a really high waiting here because memory is a a big dollar cost for um for the AI stack for the A infrastructure buildout. So anyways, AIS to me, I just discovered the ETF. I can't believe I didn't know about it before. Shame on me. But I think it's an awesome ETF if you're a believer in the super cycle, which uh you know, that's kind of what our podcast is all about right now. >> 100%. Yeah, we we believe it's going to ramp up to over a trillion by uh by next year right? >> Yes. And I think I think we're going to stay above a trillion into 20 well into 2028 and then we got to see what happens in the 2028 election cycle. But if there is no populist movement that that slows it then I think we stay over a trillion dollars in annual capex uh for AI infrastructure peranom for you know for a long time for the next 10 years probably. Is there a uh a positive or you know like a goldilock scenario you've thought of where uh the politicians legislation starts working with data centers and data centers become a you know a means of redistribution. >> Yes. the the Goldilock scenario here is this populist movement gains enough momentum more states pass uh well 2026 the midterms kind of switch some things and some new people come into power and those people come into power and pass you know whether it's local or statewide legislation similar to what Texas has done and that movement gets large enough to where that side works with the federal government works with the pro data center tech side they create maybe new taxes or something that allow the masses to participate in the wealth created by artificial intelligence because right now all that wealth is flowing upward. It's not downward flowing. It's flowing upward to the creators of the AI, the shareholders of the AI companies, the venture capitalists behind the frontier AI labs. Wealth of AI is flowing upward. The goal scenario is that we work together, private and public sectors work together to figure out a way to get that wealth to flow downward. And then you have this tax and then it allows the boom to continue and the bands get removed and the populace backlash is satisfied and the buildout can continue at higher unit cost because maybe there's a 10% tax. It gets redistributed somewhere or whatever. But nonetheless, um it continues and there's no sudden stopping or stalling and everything works out in the end. That is the Goldilock scenario that I hope happens, >> but that requires a lot of faith in people working together. And right now, a lot of people don't want to work together. So, we'll we'll see what happens. I I'm hopeful that'll happen. Um, which is why I'm not, you know, factoring in this risk into my actual investment or recommendation decisions right now. But um it is a goldilock scenario that that I think could could manifest in 2728 >> 100%. Now so hypothetically we get a blue wave. Uh is there anything you would factor in? Is there anything in the supply chain that we we'd be more concerned about that we want to sell off >> or anything we want to buy into? >> Yeah. say that a name that could really benefit or names that could really benefit from this movement until there is and maybe even after there's some compromise and because one of the big things here is energy uses, right? Yeah. >> And so obviously we know data centers are these power hungry monsters. Something Trump has proposed and something that I think should become the template across the United States is bring your own power. BYOP. We all know Bob B. bring your own beer. Well, now it's BOP. Bring your own power, right? If you're going to build a data center, you have to supply the power yourself to power that data center. So, I think be bring your own power plays could really do well. Now, what is a bring your own power play? Well, the best one is one we've talked about a million times on this podcast. Bloom Energy. Group energy creates fuel cells that companies like Meta, like Alphabet, like Oracle can buy to plop on their data center to provide power. Why does that not become a more common thing? Well, I think it will become a ubiquitous thing as people get pissed about the power issues of AI and legislation gets passed, it forces these companies to bring their own power. And once they're forced to bring their own power, they're going to go to the de facto leader in bring your own power solutions for AI data centers, which is Gloom Energy. So, I think companies in that space could do really really well with Bloom being uh being my favorite. And then, funnily enough, um I think you know the Nancy Pelosi trade tracker or whatever the heck that's called these days. I think I just read a headline that Pelosi or her husband or somebody disclosed a big purchase of Bloom Energy recently. So, go figure. Go fish. You know, >> who would have thought >> that I mean that's could really benefit here. >> Yeah, definitely. That's really one of the best positive indicators. uh if Pelosi is buying into the uh AI trade. Um so speaking of power uh a sector that I I think has been dead for a long time, nuclear. Do we see anything reigniting there if we kind of >> Yeah. >> No, I I think nuclear could also have its day in the sun because of a lot of these things, right? Like again power to me is going to be like the biggest one because power is we're short on power. The grid gets constrained. No one wants to be blacked out. No one wants our energy bills to go up. No one wants to, you know, basically fund the the energy of an AI data center. So power is going to be like, I think the center point of a lot of these uh potential actions going forward. And so anything that can help supply more power, I think wins as a result of this movement. So Bloom Energy again is is a way to do that. And I think the whole nuclear complex is way that you're right, these stocks have been hit pretty hard. They were winners and then they weren't winners. And I think now they can turn to being winners. Whether we're talking Kamico, whether we're talking Oakla, whether we're talking SMR, New Scale, um I think these are names that can do really well um in this, let's call it, new era of regulated data center construction. >> Definitely. Okay. Uh I'm I'm wondering if I'm missing anything here uh in in regards to this conversation, Luke. >> I mean, we we could go on and on for hours [laughter] on this topic is deep. This is the topic that if you get me at a bar, I'll talk to you for the entire night about this thing. I am passionate about this. I try to be as informed as I can be about this. Um, and I think it's a very important thing for people to know about because it's turning into a political movement with very big potentially societal consequences, uh, global societal consequences. And so we people need to be be aware of this. And I know this episode has run longer than what we've been trying to we trying to keep it down to like 15 or 20 minutes or something. I don't know. It's an algorithm thing. you told me. Um, but I think that this is something that we need to talk about for more than 15 to 20 minutes. We've done that here. We can talk about it more in future episodes. Uh, but I think we've we've covered what needs to be covered for a good base understanding of the situation and its implications uh at least on the investment community for now. >> Agreed. And uh if you know anybody that's watching if you have any additional thoughts, any additional questions, this is going to be an ongoing topic we talk about until you know we see what happens in 2028 2029. So we will further this discussion in the future. So please, you know, like, comment, subscribe, do all the things. Uh but with that in mind, uh that's it for today's episode. We'll see you on the next one. Take care. [music] Hey. Hey. Hey.