The AI Stock Nobody Understands

[music] AI is not running out of GPUs. It's running out of electricity. Even Nvidia's CEO Jensen Wong has explicitly called energy the bottleneck for AI development. And the data centers behind the AI boom aren't only looking for cheap kilowatts. They're looking for a power source that can run 24/7 with no downtime, that can cool itself, that can plug straight into Nvidia's racks at 800 volts. And there's one company claiming it can do all of that. For the last 20 years, they've been quietly building tech that they say is perfect for this exact use case. Most investors still think it's a science project. Uh, some think it's a bubble, and some investors think that their stock could be a 5x runup from here. Today on Dumb Money, we're going to break down the AI energy trade and why Bloom Energy could be the AI stock that nobody understands. This is Dumb [music] Money Live.

Hey there, Dave here, along with Chris and Jordan. We are Dumb Money. Welcome to Dumb Money Live. Where did Chris and Jordan go? There they are.

>> It's been a, it's been a, uh, quite a morning this morning. We're watching the stock market, uh, fall apart. Oh, quick reminder, if you, if you haven't done it already, smash the like button. Like the smash button. The almighty algorithm needs that. Uh, but yes, Chris, Jordan, uh, it is, it is really remarkable that we all made it here this morning. Uh, we'll let, let our audience in on a little behind-the-scenes magic here. The three of us were out together, uh, IRL for the first time in a long time. Uh, we were at Chris's bar, a little reunion of the, uh, varsity team from our first startup company together. Fun times.

>> Uh, Jordan, you were telling me how, how good you're feeling this morning.

>> I had, uh, like five or six old fashions. I think.

>> About four too many.

>> Just kept refilling them. It was, it was just like, one, one is done and the next one comes. Let, let me tell you something, Jordan. I had three Heken double zeros and I feel awesome.

>> Way smarter. I should have done something like that. Ugh. I'm supposed to go to a holiday thing tonight, too. And I'm like, I don't know, man. I don't know if I can.

>> It is the season.

>> Do they make NA wine? Because that's, uh, that's what I'm feeling for tonight.

>> Dude, I'm telling you, man. Like I could have had a drink, but I just, I've gotten so accustomed to drinking this non-alcoholic beer and I love feeling great in the morning.

>> Well, that's the thing. I don't drink like at all during the week normally, but then the holidays come around and then it's like drinking event after drinking event and then.

>> Yeah.

>> Yeah. Wait, did, did.

>> Get on track. People aren't here to hear about our.

>> Follow up after I left on, on my table on our.

>> No, no, that was that was a, uh, a chat that was, uh, going on in our group chat. Uh, they were talking about the last time we got together.

>> Yeah.

>> And he and Miller Spillard.

>> So he didn't do it again. Okay. Because I would have felt bad.

>> He didn't do it again. No, he, he, he replied that he was on his best behavior.

>> Okay.

>> I think, did I Irish goodbye yesterday? Did I just bounce?

>> I know me and Dave did. I, yeah, we.

>> Cuz I was like, I went, I went to the restroom and I'm like, "Oh, I should order a lift." And then it showed up and I'm like, "Well."

>> Yeah, we, we, we looked over and like half the group had kind of, uh, scattered and we're like, "This is, this is our cue. Let's."

>> By the way, let's go.

>> By the way, let me just state something right now. This, that was the first night ever that our group had turned the corner and people started going up. Has that ever happened before? Usually at least twothirds are going all night. That, that, yeah, that's shocking.

>> We were still getting texts at 1 in the morning, so I, I don't know, uh, at what time the, the last, the last dude fell, but.

>> I got home at 11:30 and I was like, I haven't gone to bed at 11:30.

>> In like a decade.

[laughter]

>> Yeah. Well, you got home at 11:30. You, you're a far Uber ride. So.

>> Yeah.

>> Wait, are you saying that, that 11:30 is late for you?

>> Oh, yeah. Oh, yeah. Oh.

>> I go to bed at, I go to bed at 8:45 or 9:00.

>> Holy crap.

>> Are you serious?

>> Dead serious. It's amazing.

>> Oh, I, Oh, I, I'm like 2 a.m. every night.

>> Oh, man.

>> I'm not out partying.

>> Dave, is, are, is the video showing up on X? Like, why is it my, is, is this just an issue with my.

>> It shows it is.

>> Oh, it is. Okay. So, all right. Forget about it. My, my bad. Um.

>> You should, you should, uh, share this on your X. And if you're watching us on X, you should share it, too.

>> I, I just tried to share it. I just didn't want to share something that wasn't actually playing. Um.

>> On your.

>> Yeah, I did it. I just reposted it. All right, here we go.

>> And look, our, our view count is going up as we speak.

>> Bloom Energy. Uh, dude, we've been talking about this for a while. We all know that the bottleneck is not GPUs, right? We all know that the bottleneck is is energy. Like that's a.

>> Everyone is saying it. And, and if you saw that chart that that was floating around on X where it's basically like, here is the US energy production ramp up and then here's China, you know, it's like.

>> We're behind in energy.

>> And so we were trying to figure out, what, what do we do? What, what, what stock might benefit? Well, here's the deal. You can't grow grid energy so fast, right? Um, and I think that that's kind of a concern, especially when you think of how long it takes to start up, like a nuclear power plant. Um, if you want to deliver something somewhat clean, nuclear is really the way to go. But those projects can take decades to pan out. Yeah, Jordan, we've heard, we've heard, we've heard a lot of rumors about, you know, chips that were purchased, uh, that can't be put to use. Data centers are behind. Energy definitely appears to be the primary bottleneck. The thermal load required for an AI data center are absolutely ridiculous. 20 to 30% of the power, um, is just for cooling the racks. Okay? So like, not only do you need power for the GPUs, you need an additional 30% power just then to cool the racks afterwards. It's absolutely insane. And kind of the go-to power right now for a lot of these, uh, a lot of these data centers are these gas turbines, right? But what's the biggest problem with the gas turbine? Like, they go down about one to two hours a week, something like that. Uh, so you, you can't just have enough gas turbines for your data center. You have to have redundancy. You have to have additional turbines, right? And there's a huge like backordering.

>> Yeah, there's that. And they're not clean, right? So they're not giving you clean energy. Um.

>> Um, I, it's, it's really a big issue. But that said, it, it is the go-to power source because it's a known entity. Uh, you know, you know exactly what you're getting. It's not cheap. Uh, it's not ideal, but it works and you're not going to get a whole lot of surprises from it. So then you have Bloom Energy, and this is a fuel cell. I, I think a lot of people don't even understand what the hell a fuel cell is, but it's essentially a way to.

>> I had no idea what a fuel cell is and how you can just like plug a box in and it just starts generating its own electricity. Like.

>> It's b, it's just a chemical reaction. So it's like a chemical reaction to the output is electricity.

>> So.

>> It's the easiest way to go into it without getting into chemistry, but just say it's a chemical reaction. It's just natural.

>> So, so, so Jordan, would you agree that essentially instead of combusting natural gas, it's taking the natural gas and a chemical reaction is turning that into energy?

>> Okay.

>> Correct. Right. And the, the, um, you don't get any of the, um, pollutants that way. Um, these fuel cells last, what, about six years? Um, and you just get a, it's a, it's a much cleaner process and more efficient, by the way, than using a gas turbine.

>> And, and unlike the turbines that go down for a couple hours every day, these go down a couple hours every year and you only have to refresh them, like, put new fuel.

>> I think they count the downtime in minutes. Um.

>> Yeah. So the uptime is 99.999% supposedly. So like, 1 to two hours a year. And the input is essentially the same. It's natural gas and air, right? And the output, and we'll talk about this later, the output is 800 volt DC power. Uh, and one other interesting tidbit about this is you get a byproduct, which is 400° Celsius exhaust heat. And that exhaust heat, I don't know exactly how this works, but I guess the exhaust heat is then used, uh, to help, uh, fuel the absorption chillers, which, which actually helps.

>> Somehow heat is then converted into a cooling system so that the, uh, the data center can use it to actually help cool itself. So it's a power source that generates heat that then can be used to cool.

>> And the cooling effect.

>> Science. The cooling effect, which is a really interesting data point here, um, is 25%, right? So those absorption chillers reduce the cooling overhead, uh, uh, uh, cost by 25%, uh, excuse me, from 25% to five, 5 to 10%. So basically, instead of it, you now, you, you now only need an additional 5 to 10% of energy for cooling as opposed to the close to 30% additional energy for cooling that you needed before with the more traditional gas turbines. So the big rub though on this type of energy on Bloom is that if you just look at it simply, it's, it is more expensive. Uh, this is an, this is an energy product that they've been working on for two decades now. And if you.

>> Yeah. So it's more expensive, but especially, um, from like a capex perspective, because there's a large upfront cost to this.

>> Yeah. So I think the capex, Jordan, on the upfront is two, Bloom Energy box as opposed to like the turbines, it's 20 to 30% more expensive than gas turbines. But that, this is where the misunderstanding comes in. Okay? Because.

>> I think most investors these days, I, I hate to say it, are a little lazy. They just kind of look at the top line, right? And so everyone's like, "Oh, it's, it's."

>> They see the headliners, it's more expensive. And then they don't dig into the fact that this just, that cooling element alone is a huge savings. Uh, and apparently, you know, this, we, you mentioned the, uh, 8, [clears throat] 800 volt DC power, which is unlike traditional electricity, which is AC, that has to be converted to DC and, you know, it, it basically generates the thing that our chips need. And they didn't intentionally develop this for AI, but it just happens to work out that it's kind of exactly what, what computer data centers need. You're exactly right, uh, Dave. So, basically, because they have this weird direct to rack 800 volt DC system that is completely by chance, uh, they don't have to have any medium voltage transformers. Okay? No UPS. Uh, battery farms, the AC to DC conversion stages, um, add a tremendous amount of cost. So Oracle, which is like their first big hyperscaler customer, which is part of the reason why Bloom Energy is trading down because Oracle's in the dump. So it kind of just trades down with Oracle because that's kind of their big initial hyperscaler. Oracle has come out and basically said that they estimate that they're saving $1.5 million per megawatt in capex once you account for the lack of redundancy that they don't need to have anymore as opposed to gas turbines. The cooling savings, remember, because they're bringing those that cooling extra energy down from 25 or 30% down to 10, [clears throat] to 15%. So they're cutting that in half. The direct to rack at 800 volt is saving all kinds of money with transformers and wiring, and it's just a more compact setup. So they're removing a lot of the energy infrastructure when they install a Bloom box. So when you consider all of that, it's actually 10 to 20% cheaper all in. All in. Okay. So, and that's the thing that's really misunderstood. But then on top of that, and again, this is super misunderstood. Gas turbines have barely changed in 30 years. Okay? All right. So, this is on more of like a technology chip, uh, I, you know, innovation cycle. So the Bloom Energy technology is constantly getting better and more efficient. So every few years, you're, they're bringing additional efficiencies, right? So unlike a gas turbine, if you move out six, seven, eight, nine years, it just continues to get more efficient every couple of years, which is not something that we're accustomed to, uh, with like a gas turbine or conventional energy models. Maybe with solar to some extent, right? We, we've seen some of that. Um, but once you kind of put all that together, uh, it's really not only is Bloom Energy, and we didn't talk about this, the big reason, uh, the big attraction for a Bloom Energy Box is its speed to compute, right? So, if you get a Bloom Energy box, you can speed up the buildout of your data center by many, many, many months. It could be 6 months, 12 months time savings. So they already have the speed to compute. They have more capex in some ways on the front end, but ultimately they end up saving you money in the long term. So, and it essentially, as much as they could produce of these Bloom boxes over the next few years, you could theoretically put out a thesis that says that their growth should be near infinite, only capped by their ability to actually manufacture boxes, which I think realistically they could double their capacity more or less every year. Um, you know, going out the next four, five, six years. So, we're going to present a thesis here that is not our thesis. It's actually was one of our friends, Xianu, that a lot of our our initial Bloom research was sourced by. We're not financial advisors, right? We've said that many times. We're just sharing something with you that was shared with us over the past few months. We've taken many meetings, done a lot of research on this ourselves, and I personally feel that this is one of the most interesting misunderstood AI plays, even though the stock is up 5x, right, since the beginning of the summer, which is where a lot of the people in our group started investing in Bloom Energy. But by the way, that 5x move was not really encompassing all of the theoretical growth in AI data centers for Bloom. It was really more of a, hey, this company was kind of in the dumps and now you have this AI wave and you have some proof, uh, that their technology is viable and working with some of the early deals that they secured. That's kind of why this stock went up 5x. Uh, it, in my opinion, by no means measures out, uh, what is possible for Bloom Energy over the next two to five years. So there is a thesis that Bloom Energy could, what, 5x, [clears throat] maybe even 6x, uh, over the next few years purely based on conventional multiples and expected earnings growth. If Bloom Energy simply executes, uh, on their plan over the next few years. Now, analysts are not openly putting those numbers out there, right? Because they just seem outlandish. Also, Bloom Energy is a company that has disappointed for 20 years. This has been like a developmental stage company that has continued to throw out.

>> Company that investors and analysts love to hate.

>> Yes. Absolutely. And, and probably for good reason, right? I mean, it's been a long time for Bloom Energy to just lock their way into a use case. I mean, and that's the crazy thing.

>> Well, the other, yeah, the other thing about, I was thinking about is that, you know, I think, and people have mentioned, well, what about, you know, on-site nuclear? And look, that's a great option. I love, um, the SMR, um, small modular reactor, um, concept, but that stuff takes time, right? It takes regulations. It takes, uh, you know, there's not, they're not operationally ready the way that Bloom Energy can just come in and deploy. And so.

>> Jordan, Jordan, look how long fuel cells have taken with Bloom. Okay? We're talking.

>> You're talking like a 25-year story.

>> Yes.

>> For Bloom Energy, right? So, you know, it just takes time, right? And the, and the good thing about Bloom Energy is that they can work side by side with grid power or, I'm sure, with, you know, small modular nuclear reactors. So, um, look, I mean, are they, are they going to be everything for data centers? Um, maybe not, but they'll definitely be in the mix of power that you use. And again, I think we skipped over this part of the story, but for anyone following Nvidia and their next generation, um, architecture, it's entirely based on output voltage, which is 800 DC, and it just happens to be exactly what Bloom Energy can deliver, right? Like, right out of the box. So, this is just pure circumstantial luck. Uh, and that is really the cornerstone of the entire narrative the next few years in that. And I personally think there's a high likelihood that at some point in the next year, we see some interesting partnership, maybe even an investment coming from Nvidia. So far, because listen, it only helps Nvidia, right? Because they are perfectly geared to power up their next generation racks. So far, the biggest investment that was made recently in Bloom Energy was by Oracle. I don't, do we have the nump, the stats here, uh, on how much Oracle invested? I think it was a conversion at $18 a share, but it was a mega, mega large investment by Oracle into Bloom Energy. And it just shows how much faith, uh, Oracle has in the technology. And they've been the most aggressive in terms of being the first hyperscaler to say, "Hey, we're going with these Bloom boxes."

>> None of the, you know, the, the feedback that we're getting from institutional Wall Street is that the other hyperscalers are all interested, but nobody wants to be first. So they're like, let's see, let's see how this goes with Oracle. Um, and if this goes well, uh, also OpenAI. So it's Oracle and essentially OpenAI. Uh, if it goes well, I believe that Google, uh, is powering their Mountain View facility with, uh, Bloom Energy Boxes. I believe so. I also believe that there is an Nvidia, uh, multiple Nvidia, uh, offices that are powered by, uh, Bloom boxes. So there are numerous, uh, potential future events that could propel the validation of Bloom in the marketplace that I think are likely to unfold over the course of the next year. But why don't we just talk about the actual math, okay, for a minute. Um, if we talk about the math, we have global new natural gas generation is about 100 gigawatts a year. Right? So let's say about 50% of that becomes AI driven, that's 50 gigawatts a year. If Bloom Energy wins 10 to 20% of that, that's about 5 to 10 gigawatts a year. Okay? So that said, um, let's look at the revenue for earnings math. Bloom Energy earns roughly three and a half to four million dollars of revenue per megawatt installed. Plus, you have a 10% annual revenue service contract that will be locked in for decades for everything that they install as part of their servicing. So they get an additional 10% over that annualized. If they're able to hit the 10 gigawatt annual install by 2030, which I believe they can, the conservative model puts them at about $20 earnings per share. The aggressive model, um, which includes some non-AI markets, could put them anywhere between 30 and $40 earnings per share. Now, a lot of the mission critical infrastructure will trade at 25 to 30x earnings. And that would basically put the stock at five to $600. Okay? At a $30 per share, uh, earnings multiple, uh, you're looking at maybe even $800 plus for Bloom. So, that, that is the bull case on Bloom. There is a bull case here that ranges from $5 to $800. And all it takes is for them to actually execute on the plan and meet the demand that is out there for energy with Bloom boxes. Okay. So by no means do they need to dominate the sector. They just need to be a participant in the sector for that to happen. Um, but, but why is the market not not not valuing them? There's a bare, there's a lot of bare cases. What do you guys want to talk about some of the bear?

>> There's a lot of bare cases. Yeah. And, and we've seen, we've seen a huge run up in the stock and it's pulled back recently and it's down big today, but it was up big yesterday. So it's like, it's one of those very volatile stocks. It's not a huge market cap. Uh, and it, it, it moves and it's, you know, it's off of its highs, but if you look at that runup, it's, it was a dramatic runup since summer. So bare case number one, it's the big one and it [laughter] it's the bare case for all the AI stocks. AI is in a.

>> It's the, yeah, it's the, it's the bare case that isn't just this stock. It's, it's what Michael Bur is worrying about. It's, you know, we're in an AI bubble, right?

>> Yeah. So, so.

>> And I don't think we are. I think, I think that, you know, we are as, as fast as we can build and deliver AI, both on the chip GPU side and on the power generation, and we'll use all of it we can get. We talked about this in our last episode, we talked about delevering, right? I currently do not own any options.

>> Period. I own not a single option.

>> That said, I did own some about an hour before the show, but I sold them all. So as of right now, I have no [laughter] I, I have no options. Okay. Um, and I do have leverage, but even my leverage is hovering around, let's call it 50%. I think it's around right around 50%. You know, 50% leverage for me is low. Uh, 100% is high, 70 to 80% is norm. So I am, have no options and only 50% margin leverage on my account. The reason for that is we are in this period of time where we're going to have a lot of volatility and it's really hard to kind of get any sort of conviction around how investors are going to react day in day out to the news cycle. There's a lot of nervousness out there. So, I think it's, this is the time to kind of pick your longer term winners and not be super concerned about trying to predict a short term, uh, movement in price of any of those stocks. [clears throat] But AI as a bubble is not something that I'm concerned about. You know, you guys know that the AI trade is going to be my trade. It has been for two and a half years. It will be for the next two and a half years. So I'm just going to like check that box.

>> On both the GPU side and the power side. There's, there's plenty of demand and that demand I think will only increase.

>> Well, these guys are spending. I mean, they're building data centers, right? And they have to be powered.

>> You nailed it.

>> That brings us to, uh, bare case number two, though, that Bloom. Yeah.

>> Chris.

>> Yeah. So, so, so that Bloom has no real demand, uh, because it's a company that has never, they, they basically never meet what they say is going to happen, right? So they, they put out these.

>> Bad history of missing expectations.

>> Yeah. And, and, and that, that's a real concern. So, uh, we do have people that are friends of ours that have dug into the supply chain, uh, in China, and we have at least our intel is telling us that Bloom Energy has already booked, uh, all of their supply chain for 1.3 gigawatts in 2026. So there is a degree of confidence that I have that they have the supply chain sewn up and are going to deliver 1.3 gigawatts in 2026. Okay. Uh, hyperscalers are privately confirming that they, that Bloom is Costco competitive when you factor in all of the advantages. Okay. But again, they're all kind of waiting on Oracle and OpenAI to be the first adopters. But I like that. I like that because it means they're, Bloom is conversing with all the hyperscalers, right? So it's like these other deals are just ready to, to, to get announced. Uh, we don't know when they're coming, but I think there will be other hyperscaler deals announced over the course of the next year. Um, so I like that not all the good news is currently, uh, visible to the market at large. Bare case three. This is a, this is a fun one. Uh, dude, these short sellers, man. [laughter] Just, they will say anything.

>> They will find, they will find something to put out their short report and then play the stock on the way down.

>> Yeah. So, there's a company, there's a short, uh, fund out of like Puerto Rico. Is it Puerto Rico? I think, uh, that says that Scandian, there's a Scandian shortage and that's going to kill their ability to scale at Bloom. Okay? So the Bloom fuel cell requires a rare earth called Scandian that essentially.

>> Only comes out of China. Initial, like for the last, so many years, you can only get out of China. Jordan, do you know anything about this? Scandium is how you say it. But.

>> Scandium. Okay. Scandium. [laughter]

>> Sounds like a scamium to me.

>> So the short report says, and the short report actually has like a low, medium, high prediction for Voloom. And the low, I think, is zero. That the company's going to zero. I love this stuff though. This is great. You want to have, there's two sides to a market and unless you have this side, there's no upside, right? Because you have to have, you just have to have a balance in in terms of how the market's looking at a, at a, at a company. So they're basically saying that they have calculated the amount of scandium that is needed that Bloom has historically used. Yeah. For each and how much in order to build one fuel cell, they, they tried to, you know, back do the math to figure out how much of this scandium was needed. It's like 15 kg.

>> Yeah, like 15 kg per fuel cell. And then they have calculated Bloom's, uh, projections on growth and they called the scandium, uh, sourcing company in China and said, can you increase the amount of shipments of scandium, uh, by this much to support.

>> Can you double it every year?

>> And they said absolutely not. They can't do it, right? They can't do it. So the short report's like, this is a scandal.

>> Bloom Energyium.

>> But if you look at the footnote of where they got that information, where was it? It was like some like unreliable.

>> Well, well, hold on. Our research and, and, and our partners' research, uh, is that it appears that that 15 kilogram per fuel cell, it appears that that is sourced from a debunked 2010 student estimate. That, that, that's our research. Okay.

>> Yeah. Now, Bloom Energy has also come out and said that their usage is 20 to 40x lower than that. Why? Because that was a long time ago. First, 2010, when they were just initially developing the technology, they needed more scandium and they've become more efficient. But nobody watching the stock has like done a revision to what they, you know, the whole calculus of scandium in fuel cell out.

>> Yes. And that they're, and that they're, it only represents like 5, 100ths of 1% of the cost of goods sold. Okay. Is the scandium. Uh, they've been, uh, that they've been procuring Scandium for a long time. Um, and that they have like.

>> Some really large runway. I don't know what the exact number is.

>> Yes. So, so they have been supposedly stockpiling scandium for the past decade because their use requirement of scandium is meaningfully lower than this short report claims it is. Therefore, rather than using the scandium every year, they've just been stockpiling it. In addition to stockpiling the scandium and having a lower requirement for scandium by a factor of 20 to 40x, they also claim to be getting new supply of scandium now out of Australia and New Zealand beginning in 2028. So there's a lot more to the scandium story than is being reported by this short seller. Uh, if you were to do your own research and then also listen to Bloom Energy. So, but again, it's part of the short, it's one of the many short thesis on Bloom, which which results in the volatility that we're constantly seeing on Bloom Energy. Um, okay. Now, how about new tech? This, this is a good one. And there's, there's part of this story just emerged this week, but like, I think the new tech that people worried about is like linear generators, right? And, and, and those are generally used as peers, not base load, uh, right for energy on these, on these hyperscalers. But I think again, the thing that no one appreciates is the hyperscalers are not going to just adopt a brand new unproven technology for a data center that costs billions, tens of billions of dollars, right? That needs to run 24 hours a day.

>> I heard a, I heard a quote today. I was listening to, um, uh, I don't remember what his role is at Bloom, but he's a higher up at Bloom and he said that data center people that design data centers love new technology as long as it's a decade old.

>> Um, and that's about when they started getting traction in the data center.

>> You, Jordan, you just, you just nailed it. You just nailed it. And, and, and that's, that's something that really nobody really appreciates is that even with Bloom Energy, and they've been around for a few decades and they actually have their Bloom boxes out there and working, right? Like at this level of scale, they're still concerned with fuel cells, even as a multi-decade old technology that's been improving. So when you're talking about a brand new technology, people are just totally getting ahead of themselves. And on that note, let's talk about space data centers, okay? Cuz like this came out of nowhere this week. This is like, for me, this was the craziest thing. First, you see a news story about, um, about SpaceX potentially IPOing, right, in the next year. Within a matter of days of that story coming out, there's now all these stories coming out about space data centers. This is the future. It's freaking insane. Will there potentially be data centers in space one day? Yeah, absolutely. Probably. I would say there will probably be data centers in space one day.

>> Yeah, but anytime, like this is something that's a pie in the sky. There, there's no designs currently available that can dissipate heat the way that you would need to. Um, it, it just doesn't exist. So, we're, when, when I, I just made the joke about that data centers love technology that's a decade old. There's zero technology that's any years old to be able to run a data center in space right now.

>> So, I'll just say this, when it comes.

>> It sounds like a, it sounds like a perfect solution because you have unlimited solar and automatic cooling and like.

>> You don't get automatic cooling. You have zero cooling, Dave.

>> Yeah, but we also have to like ship it out on a spaceship to even.

>> But the cooling is a big problem because you can't, you have no, um, atoms to bounce things off of. So you can't, you literally can't cool anything. Well, well, you could radiate heat.

>> To infrared is probably the only way. So you'd have to have some system to then convert.

>> You, you have to have a large radiate, like radiated cooling, right? And, and so like there will be a solution.

>> Yeah. So they will engineer a solution at some point, but that solution hasn't been engineered yet.

>> Okay. So, so here's the big thing about.

>> Just give Elon a decade. He'll figure it out. But until then.

>> And then another decade and then. [laughter]

>> Here's the thing. I, I, I think everybody can agree that it's likely that we'll, we'll see data centers in space at some point. The big question, it always comes down to one thing, guys. Timeline. Timeline. Full scalable data centers are not designed for space today at, at, at full scale. There are numerous questions, numerous issues. They're all solvable on paper. It's really interesting if you can solve some of these problems. Uh, you can have a hyper efficient data center in space. But, but the people that I've spoken to and people that I trust have spoken to, the consensus is that you will see initial deployments in the 2030s. But the realistic time frame for hyperscale clusters that are economically competitive with those on Earth in space is likely to be 2040s. Okay. You need the launch costs to fall significantly today for Starship. Okay. To get that done. You will likely also need to have some type of space robotics, maybe humanoids in space, uh, for maintenance of the data centers. Uh, the chips, cooling systems, electrical systems, supply chain, all need to get built, tested, and optimized better for space. Okay? Um, all that stuff has been built and optimized for Earth over the past 70 years. So, you know, there's a thesis that unless we're really running out of power here, it, yes, it makes sense to work on space data centers. Uh, there will be a need for it, but I think the timelines are totally skewed, like everything else that comes out of Elon's, right? [laughter] Um.

>> He's a visionary and it takes time to get to that vision.

>> Yeah. Yeah. No, exactly. And, and listen, he hasn't put a timeline on this. He's, he's saying this is possible. This is, this is going to happen, but we'll, we'll just have to wait and see how long it takes.

>> Yeah. And anytime we see a, you know, kind of a short thesis, usually one of the short thesis is always about competition and emerging technology. Like we hear about it for, yes, there will always be emerging technology that will displace the thing that we're talking about, obviously. And if we do get space data centers and all data, AI moves into space at some point, yeah, solar will likely be powering most of that and/or nuclear up in space. But that's not a problem for today and it's not something that I'm concerned about today as a Bloom Energy investor. Um, the, the, the next, uh, bare case is, we already kind of discussed this. Bloom has a terrible 25-year track record. Totally agree.

>> There's no denying that that is a big problem.

>> I, but again, this is.

>> You finally found the right use, right? So let's just, you know.

>> Yeah.

>> Yeah. And, and if even if you look at their website, their, their existing use cases are things like, uh, they, they, you know, a Walmart, uh, with batteries, not batteries, but fuel cells, uh, on it. Like the, there was not really a need for a Walmart to have batteries. I mean, it, it's great for redundancy, maybe, but they're already on the grid, right?

>> Yeah. They're just on the grid, right? I mean, there's no reason to have that to add to the capex of, uh, building one of these, you know, Walmarts when.

>> Or plug it into the grid.

>> Again, I, if, if you're Bloom Energy and you could draft out your dream scenario to fall in your lap,

>> It would be the AI data center, right? Like this just came out of nowhere and, and it, [clears throat] it's absolutely wild. Now the.

>> Fact that GPUs are moving towards higher voltage. Right now it's all like our existing is what, a 48 volt system and it just, they're, they're moving to this higher volt system that, uh, Bloom has had been developing anyway. It's not like they knew what they were doing. They just happened to be developing that as the, uh, the output.

>> Yeah. Um, the last bare case, they, we discussed it already. The stock's run 5x. The thesis here, the counter narrative is that most of that move was just a repricing based on solvency and legitimacy, not being dominant, uh, as part of future, uh, the future AI, uh, energy super cycle. Uh, so the option volatility is still extraordinarily high on Bloom Energy, which tells us that the market is unsure about this company. They are not euphoric. Okay? So this is still a company that has probably an equal to larger number of bears and bare case narratives going around it than bullish narratives.

>> Yeah, I think just tell us the options market just tell us that there's going to be, there's going to continue to be large moves in this stock. Um.

>> Yeah, but again, I, but this is, this is the type of investment where you want to dig in. The reason why we're sharing all of this with you today is this is an idea that we want you guys to dig in on because we are in a super cycle. A lot of these stocks have run. Uh, we've just laid out a thesis that could theoretically pave a way for a five to 6x on this stock over the next few years. But it's not our job to poke holes in the thesis. No, I like, I really want everyone in the community to do their own homework on Bloom. The one thing that I will say with very high conviction is that this is a company that is worthy of the research. It's worth your time to teach yourself about the risk factors and what is real. Where is the ground truth in Bloom Energy? Because if you could get conviction on what the ground truth is, this could be, I think, moving forward the next couple of years, one of the most interesting AI trades. Again, we did our homework. We're, we're totally fine taking on the risk, but you should not until you do your own homework on this stock. There is a lot to learn when it comes to fuel cell technology. But because there's so many questions, because there's so many doubters, that's where the opportunity is, is figuring out, do they actually have a good bare case or is it just fear and misunderstanding? Because if it's misunderstanding and just fear and just bias based on a company that's made a lot of mistakes in the past, if that's true, then this is a massive opportunity. But if it's, if it's correct, then maybe not.

>> Um.

>> Go ahead.

>> There was another, there was another, um, negative, um, and the G, the ticker symbols GEV, GE Verona announced that they were going to, um, create data center, uh, fuel cells the other day and the stock, and the stock went down like 8% that day. Um.

>> Then it went up 8, now it's down 10. Um, but the, if you read a little closer, uh, you know, it was just a, they've got plans, right? And, and the plans call for maybe like a 2028 or later, um, launch date on some of those products, um, and they're not going to work the exact same way that Bloom Energy's, uh, product does. So.

>> That, that's fine. That's totally fine. To the compet, more, more comp.

>> I think what it does is it it validates, um.

>> Yes.

>> The use case for Bloom Energy, right? Because we know that Bloom Energy is not going to be 100%, um, of the mix, right? But they're going to be in the mix of.

>> Power solutions that you use for a data center. Um, I think it's really important to assess how institutional Wall Street is looking at a stock and, you know, in speaking, uh, with our friends in the space, the intel that we got back is that there's an analyst gap on Bloom Energy. And what we mean by that is most of the tech analyst, uh, on Wall Street are ignoring Bloom Energy because historically it's an industrial company and industrial analysts generally don't buy the AI demand forecasts. So the type of analysts that would cover Bloom aren't generally ultra long on AI and AI analyses. So the tech analysts don't really understand it and they don't cover it because it's an industrial. There's really no incentive for sellside shops to cover midcap power tech because the way that Wall Street works when you're a sellside shop is you get compensated based on the volume of interest, the volume of calls that you get, right? And there's very low commission revenue with midcap power tech because generally the buy side doesn't care that much about it, right? Uh, you also have institutional PTSD from decades of overpromising. You have crazy high volatility on this stock going up and down 10% that scares away all the risk managed funds. So like there's a lot going on there, which is why you haven't really seen Wall Street kind of go all in on Bloom Energy. It makes a lot of sense. That said, uh, I guess the IR head, uh, who's an ex Citadel portfolio manager, is been, I, I think rumor is he's been buying stock. Uh, I guess in the '90s, Oracle bought $480 million of Bloom Energy recently via a strategic investment at $108 conversion price. But Oracle's in the dump and that doesn't prove to be a point of validity because it's Oracle and everybody hates Oracle right now. So again, this is, this all makes sense. The stock is sits in like a no man's land where retail, institutional, and analysts all misunderstand the thesis. But that's what you want as an investor. You want a stock where there's misunderstanding. You want a stock where there are a handful of short bare thesis that are circulating to to install doubt in the investor base because if you're able to disprove them to yourself, then that, that, that's alpha, right? That's edge. So yeah, I think that's, uh, I think that's pretty much it. All right. So, so what is the bottom line? We have a lot of information. What is, what is the trade? How are you playing this? Again, not financial advice for anyone. You do, you do your own research, but after all of the research you've done, what are you, what are you looking at, Chris?

>> Yeah. So, so I have a large position in Bloom Energy and it's an equity position. I, I was invested in options in the past. I, I don't currently hold options in Bloom because I just, day in and day out, there could be news that destroys Bloom, right? And so if, if Bloom gets hit meaningfully and the, the underlying information that I care about hasn't changed, I'm likely to increase the size of my position. Right now, it's probably a top five position for me.

>> Really? Top five.

>> And would it grow? I, I bought a position, too, but it's, it's top 10.

>> Yeah. So, you know, that, that, that's where I sit. I think that this is a, is like a, a, a 12-month trade, maybe longer. Uh, I think a lot of the information kind of like when I talked about Robin Hood a year ago and Palantir a year ago, uh, they were really misunderstood stocks. And I, I could be totally off here, but I think a lot of the misunderstanding in Bloom will hopefully get worked out over the next year. And I think if that happens, we should see a repricing in this stock to more fully appreciate what appears to be, you know, fairly obvious earnings growth over the next two to three years. That is substantial.

>> Yeah. So, but there could be bumps. If there, if there's huge bumps in the AI, uh, cycle, Bloom could get hit. Okay? I could see this stock continue to get hit and, and, and fall.

>> And this is a super volatile stock anyway. Even when there's no news at all, the stock swings 10 or 20%. Uh, it's, it's not, it's not for the faint of heart. Like, I woke up this morning and, and my position was down 10%.

>> No.

>> But it was up, it was up 8% yesterday. It's like, it's one of those stocks that you don't, for me, I'm not going to look at it every day.

>> Could, could I lose 50% in my Bloom position? Absolutely. Uh, could I make 4x in my Bloom position? I think I'm more likely to make a 4x on this than to lose 50%. And.

That's why I'm invested. Uh, it's a risk-reward play. I've done a lot of homework. We have a close friend who has done probably 50-plus hours of deep research on Bloom. It's like his number one position. Uh, we took a meeting with him.

By the way, that, that was so impressive. We, we had a Zoom call with him and the amount of information he gave us in an hour was hopefully as much information as we've given you in this past hour, but it was, it was kind of amazing to see the way his mind works on this stock. Guys, did he send us an updated research brief that we can share with the community on Bloom? I, I felt like he texted us one. Yeah. How, how can we share that?

So, what we need to do is put it in our Discord server. Dumbmoney.tv/discord is the way to get an invite to that. It's absolutely free. If anyone ever asks you for any money saying that they're us, it's not us. Don't ever send any money to anyone on the internet. Never. But.tv/discord. Completely free. Yeah. We don't charge one penny for anything and we don't manage your money. And if anyone ever asks you to wire them money or sign up for a course or anything, anything other than a Dumb Money sweatshirt, which could be us, [laughter] that could be, that's the only proactively ask you. Yeah, we won't ask you. We just put it on our, we put it on our website. Um, it's not us.

So, yeah, man. I, you have to ask yourself, do you think AI power demand is real? Do you think it's overhyped? I'm going to say it's real. Do we all agree on that? Like, the power demand for AI is not overhyped. It is absolutely real and is not going away. Correct.

Um, and then I, I think if you can wrap your head around that question, this becomes a really interesting piece of the story. So, yeah, man. That's it. That's Bloom Energy. That's Bloom Energy.

What else is going on? Ticker symbol BE for the people who are asking in the chat. Yeah. I actually, I haven't, don't even have our show pulled up here. Any, any other topics we want to hit, guys? Any questions?

People are asking us. A lot of people are, uh, screaming, uh, TAC at us. TAC? What's going on with TAC? Let's see. Uh, well, it's been down flat over the past few weeks. Yeah, I mean, they're, they're at the $14. This is, this is what I like. This is what I want. Um, I, I have, I have theoretically some more money coming to me in the next few weeks and, you know, TAC is one of the companies I would deploy it to. Um, if it's down here in the $14 area when I get that cash. So, yeah. Yeah, nothing's changed. Nothing at all has changed on our TAC thesis. We think data centers are going to Canada. We think TAC is a primary beneficiary of that. It's a junky company. Uh, maybe they'll screw it up. They seem to screw up everything, but hopefully, hopefully these little power companies seem to have like a bad track record that we're hoping for the best this time around. Maybe they can fix it.

Uh, I'm going to give TAC a few months. They delayed the announcement of their partner for Canada. I think that news is coming. I would imagine that over the course of the next few months, we'll get more visibility into TAC. And I just like the risk-reward. I think there's more there there's more likelihood that the news flow, uh, basically tells us what we've been thinking, which is that they're going to have a multi-year run of demand based on AI centers in, uh, Canada. So, it's, it's a very simple thesis. That's all there is to it.

But yeah, nothing has really changed. In fact, honestly, nothing has really changed with any of my AI trades. For the most part, I don't think anything's changed. I just, I guys, I don't even keep up with the news cycle. I refuse to spend more than maybe 15 to 20 minutes a day kind of reviewing the latest AI FUD and hype and news. It's just, it just seems like it's rotating.

Yeah. Every few days, people are overthinking everything right now. You watch that one dude on YouTube that gives an 8-minute summary every day of everything you need to know about, uh, AI's, uh, you know, what has changed in the past 24 hours. That's all you need to do. That's it. Like, you don't need to spend 3 hours debating people on X as to whether the AI cycle is about to completely fall apart if Bur is right or Bur is wrong. I mean, guys, if that's what you're doing with your time, you are completely wasting your life, first of all, but it's not going to contribute to any alpha in the market. It's just not. That's not what you should be doing.

Stop arguing with people on the internet. Yeah. Or, or even paying attention to other people's arguments because truthfully, nobody knows exactly how this is going to play out. And why do you, in the short term, like, why do you even care? Um, so I'm not obsessing over that stuff like I was the last year when there were actually meaningful opportunities to kind of trade short-term. Uh, these AI stocks, I think, I think they become too volatile and too unpredictable right now to trade them short-term.

Yeah. So, but by the way, I love that because now that I don't have to think about that anymore, I'm a lot more relaxed. I have more time now to do real things. Real research. Um, by the way, I'm really behind on my, uh, social arb retail research. I have a few guys from the community helping me with it and I hope to have some really good takeaways from the holiday shopping season next week. So maybe we can do like an old-school here retail brands and here's what we're seeing in the social arb data, positive and negative, and whether there's any just kind of a laundry list of, of retail, uh, and what we're seeing in the data.

Yeah. I mean, I think we should do that next week. Did you see Lulu yesterday? No, I didn't. Ah, they popped like 20%, man. What? Yeah. Yeah, dude. I was too early on Lulu. I was too early. Oh, wow. I was way too early. Yeah. Well, oh my gosh, 10% today. What? You, you were definitely, uh, early. If you, if you look at the one-year chart, you're, you've still lost a lot of money.

Hey, well, no. I, I was, I was, I'm out of Lulu. Like, I was even, I was in it. In and out. Yeah. Because of the data issues. The darn data issues. I couldn't gain conviction around it. Um, so like, I'm not winning anything today on Lulu. I don't think I have Lulu even in my retirement account. So, like, but whatever. Whatever, man. Uh, I have done so much research on, uh, quarter-zips, the quarter-zip trend. And so we'll talk about the quarter-zip trend next week and if there are actually any holiday season beneficiaries of it or not. I already know which thumbnail we're using for that one. In the dressing room doing a selfie.

Oh my god. Jordan, did you see what Dave did to myself? You know, I removed the selfie. I think, uh, from Twitter. I saved it and, uh, sent it through the Grock video creation engine and it, it created the funniest. You need to, you need to check out. If you deleted it. Maybe I, I think I was just a reply to yours. So I, I don't know if it's still. No. Does that mean even yours is gone? Wait. Reply to yours? I think so. Wait, maybe. I'll find it. It's still on my camera roll. We'll, we'll show it during the show. No, no, it, it's there. It's, it's there. I didn't delete it. I thought I was contemplating. Why? I just don't like focus. What? I'm not in focus. Like my auto-focus is not. You were, you were out too late last night. Everything that, this is, this is what real life looks like to you. It's like, you can't see my eyes. I don't know what's what's going on, Dave. What's wrong with my camera? Do I? Um, I'm guessing. Let me, let me just, uh, pull up your full screen here.

Oh, also guys, Spear, man. Spear, dude, did you see it? $98 a share. Jordan, it thinks that that globe behind you is your head. You need to put a hat on the globe. Yeah, there you go. Got me. $91 a share. Uh, Spear today for everybody in the community that is still in Spear. I sold, I sold out at like $78. I, I, um, I think I think Graham got out too around the same price that I did, or maybe he kept some because he was the one that got me in this thing. I owe all of this, all of it to him. Um, but dude, it's at $91 a share. And ironically, my wife just texted me like before the show with some bad news that the tickets to Backstreet Boys at Spear that we're buying are like $1,300 a ticket. I'm like, $1,300 a ticket? Are you freaking kidding me?

But you're still doing it. Who do they think they are? Taylor Swift. What's going on here? Oh, my for Backstreet Boys. That's absolutely insane. You will have a great time anyway. You won't even care about that money once you see. So, like, I'm so upset I got out. I can think of a Factory Boy song to sing to you right now. But.

So, wait, the next rumor on Spear is that they might get Star Wars. Okay. Star Wars Spear. Yeah. Dude, that would be insane. That would be insane. It could go to $150 if they get Star Wars. And Spear. I don't know why I sold. I don't know why I sold. I, I just, I needed some liquidity. I felt guilty about having too much leverage in my account and I started pulling back on everything and I'm like, dude, I crushed this Spear trade so hard. I, like, I just need to take my winnings and go because I didn't have time to like re-research what was going on. Figure it out. Yeah, you, you're just happy you got out. But.

All right. Anyway. Uh, what else? We'll plan a show for next week and it will, it should be good. All right. Retail, holiday retail show next week, guys. All right, we'll see you then. We're Dumb Money and, uh, have a, have a good weekend. [music] That's [music]