Transcript
Use with ChatGPT / Claude Auto-scroll A
The discussion opens with Google's sharp stock drop and whether it signals trouble for the AI boom. Rob and Joel argue the selloff is driven by headline reactions, not a real crack in Google's business or the broader AI story.
400 billion dollars lost overnight.Is this crack in Google the first sign of a crack in the AI bubble?Joining us today is Rob Safavi and Joel Litman with Altimetry Research to look at the real accounting behind Google and many of the hyperscalers to give us a better idea of what is going on here.Guys, we're going to cover three stocks to buy and two to avoid.I love the series we do with you.But we have a lot to cover today on one of the biggest stocks in the market.It's what everyone is talking about, a massive massive drop in Google.Rob, what is happening?Why is Google really just tumbling so much today?Again, a 12% drop in a company of this size is not something the market sees every day.
Yeah, right.I mean, 400, 500 billion dollars of um you know, of of of market capitalization wiped off the table.That's it basically, right?You're wiping out, you know, a United Healthcare and a CVS all at once in terms of what you're wiping out.It's wild.So, I mean, if you really dig in and look at it, and this is what's interesting, because if you wanted to use this as a reason to point for the AI bubble is bursting, you you just clearly didn'tlook at the numbers.Um if you just kind of go through and jump across it.So, one, um the the big two misses that uh caused everybody to panic were one, top line for Google Search, which missed by about a billion and a half dollars, and top line for YouTube ad revenue, which missedby about 200 million dollars.First off, right?
That is not what you when you think about the AI business for Google, that is not the AI business.In fact, if you look, also importantly, the search business still grew by like 15 plus percent, and YouTube revenue still grew by 10%.So, this isn't the idea of this business is being wiped out.basically the glide path is a little bit slower than people thought, but the way more important conversation, if you really want to talk about what happened for Google, was what happened in Google Cloud.Google Cloud beat by over a billion dollars, 1.6.And the number for the actual revenue growth for Google Cloud was 100% year-on-year growth.So, if you really want to point to the story into the narrative of what's going on from the true AI story, clearly what you're seeing is everybody still desperately wants to more use more AI and get more cloud
access and they are ramping their spending with Google.And look, not to put this out of the side, obviously the golden goose for Google that is the greatest business that anybody has ever created is truly search.But, the story here is not that search is actually declining.The The story here is, "Sorry, search grew by 15% this year."Well, the business that is exploding and going to power the next revolution, AI boom, the Google Cloud business is exploding.So, just to talk about that idea of that that's the case is just to put it totally put it aside.That's not the story and in fact there's a lot of really positive things to look for Google.
Bridget, one of the big problems is that the accounting is so difficult.The The accounting punishes innovation.And so, investments like in that Google has and SpaceX were a big reason the company missed an earnings.That's a non-cash item.These mark-to-market rules are the kinds of things that Buffett has talked about and said, "They don't make sense.All they do is create volatility."And that's one of the reasons they missed earnings.Another one reason they missed earnings is because R&D ramps up.And still, a problem in the United States is R&D is treated as as a expense in the period spent, not as an investment.Even in in Europe and other parts of the world, they treat R&D, a portion of it, as capitalized and say, "Oh, this is really a an asset for the future."Now, it could be a bad asset, but don't call cost of goods sold, don't call SG&A, and that's what is happening.
And so, a lot of this miss is because of these items.And lastly, I'll bring up everyone's been talking about this negative free cash flow of all these companies and how scary it is.But, in the case of Google, again, they're reporting negative free cash flow not because their operating cash flow is down, but because their investments are up.Because they know they're making money hand over fist and their incremental returns on the investments they're making.And so, everything is still very, very bullish for this call for Google even though some of the headline risks people have jumped on.
And one last thing, Bridget, to add to is I mean the laughable thing is Alpha, that is a company that trades at a as we look at it a 17 times PE with 15% growth in its biggest business,10% growth in YouTube, and also doubling in the cloud business that's so important.For you to have that kind of panic a 15 a 12% drop when you've got that much growth on a 17 times PE, it just doesn't make any sense.
The weakness is creating a buying opportunity.It really is.
So it's a good buying day for Google.And what I'm hearing from both of you is that the the AI bubble is not cracking, but the real numbers show this isn't a crack, it's more of an emotional reaction.
The speakers explain how accounting rules can make major AI investments look worse than they are, using Anthropic and Google as examples. They emphasize that hyperscaler profitability, not startup optics, is what really determines whether the AI build-out continues.
And I think Joel, you touched on something that we started to hear earlier this week when we got our first glimpse at Anthropic's earnings with their S1, too.We were hearing very similar things about a lot of fears about the amount of spend that's happening.And that spend, do you also see the spend that Anthropic is doing as more of an investment, Joel?
I'm an accountant.Accounts are very conservative people.And so whenever there's some kind of thing you can take a charge for or take an expense for, they'd rather do that and then let it roll into earnings and income down the road.And so when you look at the negative $80 billion net income line, when you do the adjustments to get back to what was really cash, they actually made money.On $16 of revenue, they still had maybe a billion to $2 billion of real profitable income.But these mark-to-market and these other adjustments made it look like Anthropic had a horrible horrible uh S1 when in fact no, it it's producing.I'm not saying it's as strong as Google.In fact, Anthropic lists Google as one of its main suppliers who Anthropic pays a ridiculous amount of money to for operating Anthropic.And so when you think of those things, you say well again, another reason to like Google cuz they're in the right part of the value chain versus Anthropic, which I'm not jumping on Anthropic's IPO.
I'm just saying it's nowhere near as bad as people say and it is I would say very positive for the industry as a whole still.
Quite frankly, the the narrative of Anthropic's profitability or even OpenAI's profitability being the important thing that you need to look at here is irrelevant because those companies aren't not actually where the real profit pools are and the realprofitability in terms of deciding whether or not we actually are going to see the AI boom continue or not.It really matters, to come back to the conversation that we started with Bridget, as to whether or not the hyperscalers are profitable.That's what matters in terms of figuring out if the boom is going to go on or not.
Yeah, and you've already said Google is profitable, Google is growing, and it's a buying opportunity where the stock is today after the big drop that it saw on Thursday.You both have three stocks to share that are big buys right now tied to that hyperscaler growth and profitability you were just talking about.We're going to get to those three names, but also talk about two names in this AI story to avoid because there are plenty of players in the AI build-out that are not good investments for the retail community tobe looking at right now.And the two of you will look at from an accounting perspective why investors want to stay away from these two names and maybe some others like them, too.So, we're going to get into those, which is just as important.I also want to share that Rob and Joel are diving into how investors can get paid during this AI build-out, and that's what their latest report is all about.
It's on Elon's shadow trust, and it really the $4 in dividend payouts that investors have received over just the last year of dividends from the companies that are building out this AI story.It's really a fascinating way to look at investing in the AI build-out story and get paid and really get a piece of all of the money that is going into the sector right now.Um this report is fascinating to look at, and they cover five names in there that are going to help investors get a piece of that shadow trust Elon and other hyperscalers are building out right now.So, scan the QR code, click the link in the description, and you can get a special offer for that report and access all five of those names and the strategy behind getting a piece of that shadow trust rightnow and getting paid during this AI build-out as an investor.
Again, that QR code and the link are in the description.Okay, guys, let's get into this list of stocks today.We have five to cover.We're going to start with the three to buy.What's the first company that you're looking at in the AI build out?
Well, the first one I'm going to mention Microsoft.Microsoft and because of Azure and because of cloud and because Microsoft is also in the same game as Nvidia and AMD meaning they're also making chips.They're they're made by TSMC but they're designed by Microsoft to help build out their space.So I would say for so many reasons Microsoft is occupying the right places of the AI value chain.The market is not pricing in all those things.I think the the headline of co-pilot's not doing as well as not using as well.They don't realize that it doesn't need to be when they've got everything else in the rest of the valley chain that all of their co-pilot competitors have to use like Azure.It's a wonderful situation for Microsoft and that's why it's one of our best long ideas right now.
Yeah, Microsoft has been at such an incredible growth story this year after the big drop it saw early on in the year.It's recovered wonderfully up 35% in just the last 3 months.But I I guess that's the question for investors.Is there enough strength here to warrant premium if you are not in Microsoft and looking at adding it or if you're looking at growing your position, is now a good entry point time based on the revenue this company'sbringing in and the future growth that it could have?Can you still add Microsoft after this recovery and make a profit?
Oh, yeah, absolutely.And if you just look at it from the perspective of first off that 35% growth in Microsoft was after if you remember earlier this year when they were talking about investing and expanding capacity, everybody panicked.And then what happened?They expanded Azure capacity and lo and behold what happened after they expanded Azure capacity?They blew the doors off in terms of Azure revenue just like Google is doing for Google Cloud revenue right now.And so this is this this cycle that happens constantly.It's you know, we have the nashing of the teeth about should they be spending as much of investing as they are and then they turn around and that and each quarter they say I mean I feel like everysingle time that we've been on here I always use the same reference.Bridget about Satya Nadella.Satya gets on every single earnings call and goes, "If we had more Azure capacity, we would have more Azure revenue."
And they are doing a great job in terms of adding more Azure capacity.And so that means you are going to have continue to have this growth.Again, Google, just to come back to this idea, 100% growth in Google Cloud revenue, and Azure's pointing similar numbers, which is why, you know, Microsoft had such a good number in this quarter, too, in terms of their results.This is all a sign of the fact that we have tons more growth in the cloud business for all of these players, and certainly Microsoft, too, cuz the the market is not pricing in what Microsoft is going toproduce in earnings growth.When you look at the news that came out earlier this week about the 60,000 buyouts.Obviously, they're not expecting to have 60,000 employees accept those buyouts, but all that is a lot of what we talk about regularly when we talk about Microsoft, which is when you look at Microsoft, especially of the four hyperscalers,
Microsoft, this is the most true for.For Microsoft, a huge a huge part of their expense line is employees.It's coders.It's software engineers and everything else.And when they look at more and more AI benefit from that, that doesn't necessarily mean, of course, they're trying to facilitate, you know, basically getting more people in, but it also means more revenue per employee.They don't need to hire people along with getting rid of people, and all that means an ROI lift that the market just is not pricing in at current levels.
The Bridget, the fact is a bunch of your peers out there in the media are out there saying Microsoft is cutting head count because they're burning cash and they're running out of money.The fact is Microsoft is letting employees go because they don't need them, and they can still be super productive with the employees that are left.That's the real reason.
Yeah, we're talking about the the financial impact on the company and whether this is a good investment, not whether it's a good or bad thing that all these people are losing jobs.That's not the debate in this video today, although I know there's plenty of people who have strong feelings about lots of things around that.I think another theme is coming in play here with both Google and Microsoft, and that's headlines.Headlines play such a role in investor sentiment and market sentiment about these companies.And often times, it's not looking at the fundamentals, the real accounting numbers, which is why we love having the two of you on the show to really look at the numbers and what this means for investors, the thingsthat you should be paying attention to, not necessarily every other headline that changes these stocks up or down throughout the entire year.
The conversation shifts to three preferred names in the AI infrastructure trade: Microsoft, Generac, and AMD. Each is presented as benefiting from cloud growth, power demand, or chip mix changes that the market is underestimating.
So, a good reminder to look at those fundamentals.I want to move on the list cuz we've got more stocks to cover here.So, we started with a massive hyperscaler in Microsoft.What's this next company that you're looking at as a buy right now?
Yeah, the next company you would never think of a story about AI, but it's Generac, right?Ticker GNRC.And so, Generac is really interesting because let me historically you think of them as if you have a home somewhere in the north at Bridget, I don't know if you and I, you know, in in in South Dakotado, but I in Maine have a generator and it's Generac cuz Generac has a dominant position in the market for for at-home generators.But, people don't realize they also have a commercial business.And historically, the reason why nobody has thought about that is because the commercial business has been less profitable than the home business.The home business, they're a monopolist.They own 70-80% of the market.There's a lot of competition in terms of, you know, standby power.And so, historically, there's been more competition.And so, it's been a lower margin business.
What's really interesting that's happened is what's happened for them recently is very similar to what's happened for GE Vernova.So, GE Vernova, which we've talked about here before, part of the big story that's made them so profitable recently is they used to basically make all of their money by these service contracts that they get for 20 to30 years after you install a turbine in a power plant.But, they basically gave away the turbine itself practically for for zero cost and not even a loss leader for that.Meaning, they had no profit in it or it was a slight loss leader.Well, for Generac, they had to compete with all these companies for the same way.So, they were looking at the same thing.But, now you're in a situation where Caterpillar is 3 to 4 years backlogged.Generac is also several is starting to grow their their backlog.
Though, Generac has capacity in the commercial space.And so, but Generac is actually getting pricing power now.And they just in last year completed a massive expansion of capacity for their commercial business to benefit from all this.So, you've got this setup where the business is structurally more profitable in terms of the commercial business and they're adding capacity to it while the residential business is being as amazing as it was and all that points tous the reason be really excited about Generac and it's especially telling because this is a company that the as reported metrics hide how profitable it is.Right, uniform ROI is actually was 16% last year not the 3% that it is on an as reported basis.It's forecast to rise to like 20 plus percent this year and so we're seeing a lot of stuff because of what's happening for the AI build out still being lasting and durable and all these all these hyperscalers need
that power Generac is going to be a big winner for that.
Yeah, I want to talk a little bit about what we've seen in the price action for Generac because it is very interesting.There's enough stocks out there that have seen that peak in that June July time frame and have had a downtrend since then and this is another one of those similar charts and so I want to know if itis it just general market sentiment and trend that had the stock moving lower the last few months.Is is the price action really closely tied to the AI story and that's why we saw it it drop or is there something else going on in the market that has the market reacting the way it has inthe last few months to the stock?
Yeah, I mean you hit on it Bridget.The primary issue has been the narrative out there that the AI infrastructure boom is over.Um and a lot of that right I mean if for everybody as as you watch this if you wait about five days um I don't know what happens next Tuesday but I think there's some important that happens nextTuesday that's been driving a lot of the air the narrative in terms of the AI boom.Things are probably going to get resolved cuz the fact of the matter is a lot of this has been people worried going into the election.Oh, we see a lot of nashing of the teeth about you know the AI build out and will politicians slow this down.And when we get to the other side of that you're going to see the other story which is Azure revenue that's going through the moon, Google Cloud revenue that's doubling year over year, Amazon Web Services revenue that's exploding,
Oracle's insane backlog that all means all this is actually going to get built because the private companies that have the demand are going to find ways to build it but the solution for how you get it built isto lower political temperatures.You build it behind the meter power solutions like what Generac can offer so that you don't have all those political that political friction.And that's what we're going to see play out.So yeah, this is a lot this is entirely narrative driven.It's not about anything fundamentally slowing down for the business.Google wouldn't be beating revenues for Google Cloud.Azure wouldn't be beating revenue expectations for Azure for Microsoft.That's the true story going on here.
Yeah, so interesting to see that and we'll talk a little bit more about that election impact down the road with the with some of the next ones we're going to talk about.But let's get on to the third stock on the list to keep this rolling.What's the third buy on your list today?
Well, Bridget, we like AMD and we like it for some of the same reasons that we liked it back in Barron's in January of 2015.Uh it was the first time that Rob and I were quoted in Barron's magazine and Barron's did a piece on AMD and they did a bullet bear case and we were very much bulls.Stock was only $2 at the time, very cheap and what they were missing was the same thing as today.They didn't understand how truly profitable AMD was.So people thought, "Oh, they don't have the money, they're going to go bankrupt."We said, "No, they're not going to go bankrupt, they have more than enough cash flow."We love what the CEO Lisa was doing.We said she was on top of things, that she was fixing things.We could see it happening in the uniform accounting, you could not see it in the GAAP accounting and we were all in buy AMD.
Now, today we're looking at AMD again we're saying the same thing is happening.You look at any major database, the return on assets of AMD looks like 3 or 4%.When in reality, it's probably quadruple that, maybe quintuple that.The databases are saying that and I mean all the major financial databases are saying it's still in 3-4% return when in fact it's much higher.So we like AMD for those reasons.AMD is chips are getting used so heavily into all of these hyperscalar centers because they're realizing they don't need Nvidia's top top chip for everything.Most of the stuff I do doesn't need the highest level compute, right?I'm using AI all day long.But I'm sure a good couple of AMD chips will satisfy whatever compute needs I have and the data centers realize that and so AMD's filling that that space in of of chips that they need.
So, for all those reasons, AMD is not getting the credit it's due.It looks inexpensive relative to its opportunity.It's another great long for us.
I love that hearing that you're the second person in the last month I've talked to who got in on AMD at around $2.What a good win story to you just for a track record of looking at companies and getting in on them very, very early.I think every investor would have loved to get in on AMD at $2, but looking at getting in on AMD today, they're wondering, can they get that same, you know, massive return that you got getting in on AMDat $2 if they enter AMD today?Can they still get that huge return another 10 years down the road?Do you believe that's still possible even at the entry point that we're looking at today for AMD?
I don't know that you're going to get a 21,000% return on AMD again.So, I I will be Joel, I'll be the first to tell you probably not that kind of return will you get on AMD.But, when you look at AMD, and this is the important thing, they have two benefits here.One is people are understanding that they don't need just Nvidia chips.There's another side of what's actually going on that benefits AMD, too.And it's that AMD is the only one of the chip makers who truly sits on both sides of GPUs and and CPUs.Training AI models, you need GPUs for them.Right?For training the the latest cloud model, training the latest open AI model, training the latest Gemini model, you need GPUs.For execution on actually then using the AI models, a lot of that starts to move to CPUs.
And so, there's this really big question of we've been to a point for the last few years where in data centers we moved to four, five, six GPUs per per CPU inside the data center, which has been greatfor for Nvidia.What we're starting to see happen is people are saying, "Well, actually, we can introduce more CPUs back into the mix."And if you do that, AMD wins in spades because AMD also has the integration between a GPU and a CPU that it can do because it makes both of them.And also, just in general, it doesn't mean that we're going to need less GPUs.It means we're going to use same many GPUs as we as now, but we're going to have far more density of CPUs next to them for actually handling now the compute for for actually using it.And AMD benefits from both of those things that if you look at it, the market is not pricing that second leg of the story.
The market's kind of embracing the first leg of the story, which is yes, they are effectively taking market share from Nvidia, but not that second leg, which is and also we're going to use a lot more of theother side of their business, too, which which really interesting.
Yeah, one quick follow-up question.We've got two others to get to and these are just as important.We're going to talk about the two to not buy, but I had another analyst who's very bullish on AMD, had a little hint towards a concern that even though AMD's new product launch has everyone very excited and andit helps it compete with Nvidia better, that there may be so many contracts already inked with Nvidia that are multi-year deals, that will there be room for AMD to fit in even though Nvidia has already secured some ofthese contracts.Any concerns for you there?
I mean, this comes back to the idea of if you believe that we are going to have the massive ramp in demand for AI that we're going to have, there's a lot of room at the table for everybody.This is not a four-person table that we have five people fighting for seats for.This is a 12-person table of which maybe six of those seats are full.There's a lot more growth to be had here.This is not a structural issue in terms of whether or not AMD can find a place to feed at the trough cuz Nvidia is getting fat and happy.
Rob and Joel explain why CoreWeave and Cipher Digital sit in the wrong part of the AI value chain and lack durable pricing power. They close by noting that some niche players may survive, but many neo-cloud names are unlikely to generate attractive long-term returns.
Yeah, very good.Yeah, there's a lot more room at the table, but not every company is qualified to sit at that table and those are the next two names that we're going to get to.And there's also a way for you to get some of the the food off of that table, so to say, and that's by earning an income from some of these companies that are all a part of building outthis AI story right now.Again, I want to remind you about that report that Rob and Joel at Altimetry put together.Again, using their accounting knowledge, they found the five companies tied to the AI build-out that really are part of the XAI build-out too, of building out the infrastructure for the entire AI story and there ways that you canget paid.In just the last year, these five companies have given out over 4 billion dollars in dividends to investors and you can get a piece of that right now.
They're calling the report Elon's Shadow Trust and there's a way for investors to make up to $12,000 a month using these kinds of income stocks that are tied to the AI build-out story.It's a fascinating concept on how to invest and get paid and really get a piece of this AI build-out as investors.So, scan the QR code, click the link in the description, and you can get that special report with the five different names that they are looking at.They also spell out their strategy of how to get paid each month through this form of investing in the AI build-out story.It's a great way to look at it.But, that's not what we're looking at today in this video, guys.Now, we are going to look at the two stocks to avoid in this AI story.What's the first company that you are staying away from right now?
So, Bridget, when you look at the value chain, there are good places to be.We've talked about that.And there's bad places to be, and the worst place to be is where CoreWeave is.They're getting squeezed because they have to pay top dollar for all the equipment to all the chip makers like AMD and Nvidia, but they also don't have the supplier power where they can say, "You have to use ourdata center, and you can't use somebody else's."Because there's plenty of other places to go.They don't have their own AI frontier models, so it's not like someone has to use CoreWeave because they're at the cutting edge.And so, they're getting squeezed by all the bigger players, and they're the ones that, you know, when we see a reported single low single-digit return on assets of 2 3%.That number is actually correct.
From a uniform standpoint, this company is not earning its cost of capital.Its economic profit is negative.Its forecasted economic profit will still be negative.And so, why would you want to invest in a company like that?There's a lot of hype around it, but if you want to invest in AI, there are certain parts of the of the value chain to get into.CoreWeave is in the exact wrong one, and that's why you get out of that name if it's in your portfolio.
And right now, there's a lot of negative hype around CoreWeave.This is the stock that really kind of was the poster child of the AI tumble the last few weeks.CoreWeave was down, you know, 12% day after day after day.It really saw a huge downturn the last few weeks.You have been pointing out why there's a risk in this company for a while.This isn't a new a new call for you to say avoid this stock, is it?
We've been saying it for quite a while, but you know, we do have people say, "Hey, has it gotten cheap enough?"No, it has not.It has a lot further to go down than it has to go back up in in any stretch of the imagination.So, no, even at these much lower levels, still stay out.
my other question is I've heard some people predicting that companies like CoreWeave or some of these neo cloud companies that are investing all of this capital right now to build these data centers for rent could go to zerosomeday.There's pre- predictions that these companies will just go to zero and will will will be gone and and not companies that we even are able to talk about a couple of years from now.Do you think that is a possible future for CoreWeave?
We found a way that large enough companies can remain survivors and survive a long time just eking out a return that's just enough to cover their debt costs.It's possible they survive.That just doesn't mean because they survive you'd want to own the stock, right?We want successors and winners, not not not just survivors.Um and so, you know, there is a a need for what they're doing.They just don't have any pricing power on the buy side or on the on the supply side to to manage it.So, could there be other companies that we hear about today and then are no more?Yeah, of course.Of course, there are.
All right, let's move on to that final name on your do not buy list, a stock to avoid in this AI story right now.Rob, what's that stock?
Yeah, and that's Cipher Digital.Um so, Cipher Digital, much like CoreWeave, right?To Joel's point, I mean, the whole entire idea of these companies is they're they're running a commodity.But, the funny thing about Cipher Digital is it used to be called Cipher Mining.And the reason why it used to be called Cipher Mining is because it used to be a Bitcoin crypto miner.And when you look at it, it's been so funny watching all of these crypto miners that their only genuine asset at the end of the day was they owned GPUs and they had some power contracts.And they go they go, "Well, wait a second.We don't really know where Bitcoin is going."It dropped from the high hundreds to 16, now it's around 80, who knows.But, I remember somebody else who needs a lot of data center compute.Oh, that's right.
Well, let's just rebrand ourselves as a neo cloud and then we'll be able to sell all of our capacity to somebody else to be able to have them use it.And that's exactly what's happening here.I mean, but the issue and what's go what's going on for this business is the exact same problem as CoreWeave.You've got a company that effectively is selling a commodity.The hyperscalers, they have a whole bunch of genuine assets.They have software that they're selling on top of this.They have technology in terms of how they're managing all this.The neo clouds are selling a dumb box.Um right?This is what the um the Digital Realty's of the world, the Equinix's of the world have been doing for decades.And those companies have been 3 to 4% return on asset businesses every single year.And so to Joel's point, I mean, for CoreWeave and for Digital Cypher, do they need to go to zero?
Not necessarily as long as they're thoughtful about their capital structure, but these are not companies that you were ever going to see generate 8, 9, 10, 12% return on assets.Cypher Digital in 2026 is forecast to generate a negative 20% return on asset.That means that literally for every dollar in the business, they're losing 20 cents.And that just points out the issue of like I mean, just cuz you slap on the term AI neo cloud or AI exposed does not mean that you're a buyer a buying recommendation.Even if you're exposed to the infrastructure buildout, you really have to have that discretion about understanding where they are in the ecosystem when that when that means they are in terms of exposed to profit pools or not.
Yeah, I know the first question many of our viewers have, uh does this apply to all of those companies?Because like you said, there are dozens.There are tons of companies out there that switched from that Bitcoin mining company and made the pivot into being a data center company and providing that compute power to hyperscalers and and other people who are renting that computepower.Does what you say about Cypher Mining apply or Cypher Digital now apply to all of these companies or are there some things to look for to say some of these companies might be successful while others might struggle?
If your business is purely on owning uh effectively GPUs and owning a data center and then just purely renting it out as a barebones um data center, for the most part, over the long period of time, that's nota really high competitive moat business.But there is an exception.And that exception is, if you're solving the key issue that everybody has right now with being able to get capacity, which is being scaled with access to power.Not being small with access to power cuz then you don't solve Entropic or Gemini or OpenAI's problems cuz they have zero interest in going and signing contracts with a thousand of these tiny miners.But if you have a huge huge huge capacity of GPUs and you already have the power in place, then in the short term, the next few years, you probably have the ability to generate But there's only one of
these {quote} {unquote} neo clouds that I can think of that does that, and that's SpaceX.SpaceX is the only one with xAI that's pivoted to really being a neo cloud with Groq on the side that has that cuz Coreweave doesn't have that.Coreweave has the chips, but they don't they haven't solved the power issue.All of these these former Bitcoin miners, they all have the power and the chips, but they're not big enough to be worthwhile for anybody to do a tangible deal with.So they're busy picking ones and twos.So it's just really hard to see any of these guys earning a persistent and high return on assets.That means any of them really are buying story.You have to have scale and then also have the already the infrastructure in place to mean that in the short term you could earn a superior return.
That's our two cents anyways, not that we're right.Coreweave is the WeWork of this space.A lot of hype, lot of excitement.It's been coming down.It's going to continue to go down.
Such an interesting conversation on the AI story.Overall very bullish, especially on the hyperscaler sector, but not every AI stock is something to be bullish on right now, and I think that's important for investors to pay attention to along with the actual numbers, the actual accounting,that uniform accounting.And again, reminder about that link in the description.If you want to continue learning from Joel and Rob, learning their accounting outlook on investing in the market, that special offer will also get you access to sign up for their monthly newsletter where you can get the insiderknowledge on some of these stocks and the information that they share on this channel, too.Thank you both for your time today.If you want to hear a little bit more of Rob's theory on that AI build-out, make sure to watch this video if you haven't yet.We talked about blue-collar millionaires and some of the different companies that are helping with the the nuts and bolts, the the blue collar jobs behind building out this AI story here.
These are really interesting companies to look at, too.You can watch that full interview here.