Deep Dive
Shownotes Transcript
Guy, I'm sure you're already up to speed on all things current. Of course I am. Their app makes managing your money, saving, even building credit super easy, Dan. They have a new feature called Paycheck Advance. You can get up to $500 before payday when you switch your paycheck and qualify. Now that I didn't know. So now when you sign up and you set up direct deposit, you unlock so much more.
And it's super easy to get started. Just go to Current.com slash OK. That's O-K-A-Y. That's Current.com slash OK. Current is a financial technology company, not a bank. Banking services provided by Choice Financial Group member FDIC and Cross River Bank member FDIC. Paycheck Advance is offered by Finco Advance LLC, a financial technology company, not a bank.
Eligibility and available amounts may vary and are subject to change at any time. For full terms and conditions, visit Current.com or call 888-851-1172 for more information. Will M&A pick up in 2024? Will this year mark the return of IPOs? Listen to Strategic Alternatives, a podcast from RBC Capital Markets to get
fresh insights on the trends and market forces impacting deal flow across sectors and find out how companies and investors are preparing for potential surge in deal activity and what signals to watch for this year. Listen and subscribe to Strategic Alternatives, the RBC Capital Markets podcast today, wherever you get your podcasts.
All right, welcome to a very special edition of OK Computer. I'm Dan Nathan. I am joined by my good friend Gene Munster. He is the managing partner at Deepwater Asset Management. Gene, welcome back to the pod. Always good to be back. Fun day to be back. All right. Really, we're recording this. It is fun.
It is around 2 p.m. It's one of the more interesting days in the stock market that I can remember in a very long time. You know, the S&P is down about 2%. The NASDAQ is down a little bit more than 3%. Some of the biggest winners in the stock market over the last couple of years are down a lot. And obviously, we're going to focus on NVIDIA, which is down 17% on the day. Why is that? There was some data out over the last week about a company out of China called DeepSense.
Seek, supposedly they only raised $6 million. It was not in existence two months ago, but they announced a LLM or a competitor to OpenAI's ChatGPT. It's a reasoning model. I mean, it does everything else that GPT does, but supposedly better at reasoning. OpenAI launched this 01 set of models back in September, and it was supposed to make some very big inroads on this front. So a lot of the data that was released
And I hope I'm getting this right, Gene. It was better than that of open AI, which is causing a lot of investors to think twice about the hundreds of billions of dollars, which have been spent basically by the fateful eight to build out this AI infrastructure over the last two and a half years or so. I'm going to kick it over to you. I'm going to shut the heck up. What is your thoughts on this? I know that you've been waiting for...
for some sort of headline. I don't mean positioning for some sort of headline, but waiting because all the news has been basically lower left, upper right for this secular trend. You thought that this year, and we talked it late last year, we talked about it a couple weeks ago, that there was going to be periods of digestion where investors were going to kind of take a pause or possibly take profits. So let's think of this in the context of the news that we saw, and was this the sort of event that you would have thought caused this sort of price action?
Well, I didn't anticipate this exact event was going to cause this kind of price action. We recorded, we have a weekly on Deep Tech 315 that we post on Friday. We recorded it on Friday. And so we talked about Deep Seek at that point and we had been playing around with it since Wednesday.
So we had the conversation about all that we talked about, but I'm surprised that the market has had this kind of reaction. And there's so many layers to this story. I have this almost a feeling of where do you start? But it's up to people like you and I to kind of identify the biggest pressure points around this. And I think first and foremost is the question, is the AI trade in jeopardy?
And I think that this debate about where is the AI trade on the hardware side, what does the software piece look like, and the companies that will benefit from that, I think that the broader trajectory is all in place. And this was just the exact thing that a nervous market was going to react to, essentially saying that all of the rules we had around scaling were
all this concept of continuing to invest more to get more out of AI, that that whole curve may be in question.
And I want to pause there because that's why NVIDIA is down so much. Like basically, do we need all of this, all these advanced GPUs to produce what DeepSeq has produced? And the answer is that I believe that what DeepSeq is revealing is partially true. And I think they're not fully revealing everything.
And when we started playing around with the model last Wednesday, we were really impressed. And we've seen some benchmarking that kind of puts it at 70, 80% of open AI, the O1 model. That our sense is we have an affiliate company, Intelligent Alpha, that uses large language models to pick stocks. And so we use it continuously around that as this affiliate project. And that we kind of felt like it's probably 90% of the way there. But the concept of...
doing 90% of what OpenAI has done, just training on these more limited governed-back chips, these H800s that came on late 2023. It's a special chip for China. I think it's just hard to believe. And so I suspect that there's training of this model that was done probably like in
outside of China and Singapore, and there probably was some more advanced hoppers that were used in that, which kind of changes the narrative a little bit. And I think you look at the timing, I don't believe things are coincidence. Look at the timing of the announcement that we saw last week about the increased infrastructure, the tariffs, everything that's going on geopolitically, and then this comes out of Singapore.
seemingly nowhere, but I think it's all kind of orchestrated. And ultimately, I think that this is something to take note of. As investors in AI, we're believers in this. We spent the broader part of late last week and all of today trying to figure out what this means. I don't want to diminish it, but I just want to make the point that we're probably not seeing the full story when it comes to how much this
model cost to build and what the training or what the cost of the tokens are, which is having a big impact on the market today. Let's separate two things. Okay. Let's separate the technology and the ability and what it took to build it. Okay. Let's separate that from the impact of
on public equities, we have no idea because it's not mark to market what it means for open AI, what it means for anthropic. We know that there's been a black market, there's been end arounds to get to high-end GPUs from Chinese companies. We know that that has existed. We knew that there was a big part
of the demand in '23 into '24, and the anticipation of regulation was double, triple ordering by the Chinese. We know that they've been using data centers to train these models in Singapore and other parts of Asia. So we get all that, but no matter what, if you think about how our hyperscalers have been spending, right, since late '22, for all intents and purposes, early '23, there's just nowhere near
You know what I mean? The investment. Now, the CCP in China could have been backing this thing to a great deal. I mean, I kind of liken this a little bit. If we were to have a proper stock market route based on this, and again, I'm not predicting that. I'm just saying, but maybe this was kind of the spark that lit the fire. Then this is like the equivalent to the Wuhan lab leak, if you think about it. This is a win for China, I would say, if they can...
Just in this, the geopolitical, I want to try to stay outside safely outside of that whole conversation. But I think that even if, you know what the dynamics are behind it, I mean, they're, they're, they're causing some havoc on the U S markets today. And that's,
And they're clearly behind this. And so I think that falls in the category of a win for Chinese leadership. Yeah. And what about the idea that a lot of our regulation, a lot of our restrictions have caused the Chinese to innovate in different ways? You know, we throw so many resources or these companies do. Right. So they were forced.
fully, fully, you know, embracing the idea that they were going to ask for forgiveness from their investors rather than permission. And I've also likened this, you know, to this kind of, you know, like war of the next 100 years between us and let's say China, it's a technological war. It's not based off the traditional resources, right? That over the last hundred years that we have involved in no shortage of kind of, you know, I mean-
actual wars, right? So it's kind of like a digital belt and road for them. So putting this out there to the world in an open source way, this is obviously deep seek, you know, it accomplishes a lot of goals for the Chinese. Does that make some sense? It makes a ton of sense. And I think check, check, and check on the things it accomplishes. And it kind of plays back to what is this
the timing on all this and what does it mean bigger picture? And I think that you asked about kind of the innovation and kind of China having some limitations about how they can make advances in AI because they're getting blocked from using some of the best equipment that's out there.
And I think that you can be creative. And of course, you know, that necessity is the mother of innovation. I think that there's definitely some level of innovation that needs to happen with it. But at the most basic level, I want to try to boil it down to literally the most basic level here. Think of infrastructure, AI infrastructure, all that hardware is a brain. Think of it as like neurons and humans.
And think about how humanity has gotten smarter over time. And part of the reason I think that humanity has gotten smarter is actually more people out there who have the ability to be smarter. And if you're going to look at this in a very stale way is that effectively the number of neurons that are in the world today, there are more neurons in the world today than there were 100 years ago because there's more people in the world. And so you need these types of advancements. And we're talking artificial intelligence, you
You need a brain that keeps getting bigger and bigger. I mean, that's what nature would show us that you need that. And I think that's been the same thing that we've seen over the past, not just the past three years with GPT, but even leading up to this, we were involved in a lot of AI when we started DeepWand in 2017. And so we're coming up on eight years and these scaling laws have been holding together. And so my point is, is that just the concept of AI,
trying to have just this incredible breakthrough on a brain that is like significantly smaller by orders of magnitude, a hundred times smaller than what we is, uh, than what's out there.
That is just a hard concept for me to wrap my head around right now. I mean, there'd be some pretty big precedents that would be broken if that was the case. And so I just want to, again, start there. We're not going to get a press release that says this is actually what happened, but I do believe that the scaling laws, this is bringing it back to the substance here, I think the scaling laws are largely intact because of another piece, if I can just quickly jump forward, Dan, is that
The question from our perspective is not, are there more efficient ways to build infrastructure or to train the model or cheaper tokens?
It's never been for these hyperscalers or enterprises or sovereign entities. It's never been about trying to get to a set level of AI, AGI, for example, just at that. It's been about getting to AGI faster than anybody else. And that speed piece is really important when you think about what's going on with NVIDIA. Because if these models are more efficient, and it turns out to be substance behind all this,
then you still will have incremental benefit from having the best chips and the best GPUs. And so I think that demand for the fastest and greatest, you
you can build a case that this could even be positive for it. And so, again, truth lies several areas below the surface, but those are the types of things that we're wrestling with. You think that we're going to look back in a few, I think I know the answer to this, in a few years, and let's just say this causes a sustained sell-off, something that takes a lot of froth.
out of what has been a very narrow trade, if you will, right? It's been the fateful eight and a handful of other names where, you know, Nvidia obviously garnered a great deal of the market cap appreciation in and around this trade for obvious reasons. But do you think we're going to look back and say, oh, that was just a blip and these companies are going to be at much higher levels in market cap terms and sales terms and margin terms, that sort of thing, because they're getting the leverage out of all these investments, which you kind of just hinted to because,
I think it's interesting that Marc Andreessen, the A in A16Z, he's calling this a Sputnik moment for AI. And then I've seen a post on Twitter from Gavin Baker, a treaties management, who's a pretty brilliant guy, very technical guy, investor in both public and private markets. And we'll put all this in the show notes where a lot of these folks that you think have a vested interest in
and kind of downplaying some of this data and this model, are actually looking at it and saying that this is a game changer. There is some advancement. There was some progress that was surprising. Even if you find out what really was going on, there was some, and I think that that is noteworthy. I think that the Sputnik moment, I think that's a good, there's probably two ways to look at that. One is that that kickstarted
really, I think the US's ambition there too. And I think that there is going to be this, I mean, just getting really high level stuff here. But what we're seeing here is the intelligence is becoming more readily available, whether it's subsidized by the Chinese leadership or not, but it's becoming more widely available. And so I think that this will prove to be a catalyst as far as
What does this mean for the broader AI? My goal here is not just to be positive on AI for the sake of being positive on AI. My goal is to get it right.
And ultimately, I believe that we're still scratching the surface. We can go through even with today's news today. And so I think that this is going to be, I wouldn't say it's a blip. I think, again, historically, it's 54 days from when you see these types of corrections before the market kind of gains itself back. We should probably talk about what the read through on Wednesday when we get Meta and Microsoft and Tesla back.
But I think that if we take that period, that's a two month period here. I think two months from now, we're going to be higher than where we are today. Okay. I mean, that makes perfect sense. But from your viewpoint, and again, you know, I go back to a couple of things, the concentration of these names, the fact that like a stock like Nvidia up nearly 200% last year represented 25% of the S&P 500's gain rate.
The fact that names like Marvel and Broadcom joined a party late last year. You and I have been talking about it. This was something that you were very focused on. The idea of custom silicon that all of NVIDIA's largest customers over the last few years, buying all these hopper chips, committed to buying Blackwell. I'm sure they were all on board for Rubin. It's coming out that they were working with these two companies or they had signed up
or contracted with them to build specific chips for specific duties, right? Like, or whatever your tasks, that sort of thing. And ultimately, that had to cause a meaningful slowdown in the sales of these general, you know, more generally used GPUs. And so the fact that- I agree with that. I think that-
NVIDIA's numbers would have been better if not for custom silicon. I would agree with that. And just the quick math on that, NVIDIA has doubled their business for three consecutive years. I mean, this is like unprecedented, but it would have been even better if not for that. Yeah.
Yeah. And you and I talked about it, though, in December when Broadcom CEO Hock Tan gave this guidance. Remember the guidance about custom chips? It was like going to be between I can't remember, 50 billion and 100, 90 billion or something, 60 and 90. It was between 60 and 90 by the end of 2020. And it just sounded like a load of shit as far as like the guidance and why the stock rallied up like that.
That $90 billion number might be the case. But when the stock gained as much as it did, I think it was like $300 billion, you're putting a multiple on the potential capture of that TAM at a number that made no sense two years before they estimated it. So the fact that they joined the party, Marvel, it just seemed very late stage to me, Gene. We talked about that. And when I say late stage...
I don't mean we're calling a top here. That's not what I'm trying to do by any means. I'm trying to anticipate what the next quarter might bring for some of these companies. What is the sort of guide? And the reason I do this, because the end of '21, okay, in November, Mark Zuckerberg went into some freaking computer with this avatar of him and he told the whole street and everybody who was willing to listen, okay, in his goofy sort of like little metaverse thing that they were changing the name of the company,
They were changing the ticker, okay, of the company and they were all in on metaverse, that this is the way forward for computing. And at the end of the day, Facebook or meta, whatever the hell you want to call it, it lost 70 plus percent of its value from its highs in 21 to its lows in 22s. They were also huge buyers of GPUs from Nvidia. You know who else lost 70% of their value from the highs in 21 to the lows in 22?
It was NVIDIA. You know, Netflix did too. And, you know, listen, and Tesla did too. You and I have talked about this before. The fact that these stocks now are down 20% from their recent all-time highs, you know, if people want to be defensive about it, you...
better have a lot of good reasons why in the next couple quarters, you know what I mean? That this, that the news changes, the news flow, because without a change in fundamentals, these stocks don't work from here. They just don't. And this is in the near term. This is, you know, for the middle part of 2025. Let me ask this is the, by the way, the stock's been down 10, 5, 10, 15, 20, 25%. We're seeing that all across the board here. And this is, I agree that that has,
That doesn't mean that it's a buying opportunity. When things pull back, it doesn't mean it's a buying opportunity. This could be the start of something much bigger. And I think, by the way, let me just talk about the much bigger piece. I'm less concerned about what's going on in AI and the build out and the need for speed and how NVIDIA's business is going to play through here. What concerns me more is just that part of, I believe that we're still in this now a two-year bull market that's going to end in a bubble burst.
Part of to have that, you have to have optimism, just underline unbridled optimism in terms of what these technologies can do. And now we've kind of thrown some cold water on that. And so there's a market psychology piece. And I think some of the, you know, one way to kind of track what's going on with this market psychology piece is look at Palantir reports. That's
probably bigger than in terms of this AI trade, bigger than what Microsoft and Meta are going to say about kind of AI hardware. I mean, I think they're, you know, they have to show on the software side, there's just blowing things out. And so ultimately, I think that, you know, there's still a psychological piece to the market. And it's definitely been hit hard here. And
I think it's worth kind of anticipating what some of the companies are going to say later this week and then when NVIDIA reports. But that's the risk. That's where I think I could be wrong. I still believe that AI is going to be more impactful than we realize. And it's a pretty broad statement there. But I think your optimism is probably not unfounded. You'd have to be a really cynical stock market person.
person, I don't know any of those, who would actually be willing to take a long-term bet about this technology changing so many different industries. That's where the outside returns are, Dan. I mean, it's not a consensus and right. That's where you make your money. But I'm also agreeing with you because you and I have the
benefit of remembering the late 90s and the skepticism and the powering of you know returns and then knowing for sure that it was a bubble based on valuations but the idea of trying to pick a stop a top and that could have been you know it could have been lights out for any investor trying to do that and then when the bubble popped
No one really knew it was popping. It could have been just like the sort of pullback that we're seeing right here. So it's an impossible task. I think you make a great point though on the Palantir. Here's a company that is a use case from a software standpoint, right? They've been making these investments. The stock has appreciated because of their ability to tell a story, how they're using this technology.
One thing's really interesting, though, when you talk about optimism, not just on the technology, but in the stock market. So, for instance, NVIDIA, okay, before coming into today, 90% of the analysts, that's 70, okay, 70 of them had a buy rating on the stock. The average price target was about 175, okay? So the stock's
You know, 118 right now, it's down 17, but there was 25% embedded upside in those price targets. Look on the flip side of that. Palantir is down 6.5%, also one of the biggest gainers last year in the S&P 500, obviously a much smaller market cap, but it was over $100 billion. It snuck up there pretty good, right?
Four analysts rate the stock a buy, 13 a hold, seven a sell. So think about the skepticism that exists for a Palantir. Isn't that almost like positive for Palantir? I think it is positive unless the analyst community has it right. But if they have it right, they have it right on valuation. Does that make sense in a way? They don't have to be right on fundamentals. All of those holds...
might have very reasonable upside as it relates to the EPS or sales growth and margins, but they might not like where the stock is trading right now based on those estimates. You're a former research analyst on the sell side. Does that make some sense to you? It makes perfect sense. That's exactly how the business is done, the game is played. That's exactly right.
Hi, everyone. It's Guy. You've probably heard me mention Current before on the pod, but we can't stop talking about how easy managing all your money is with their app. It helps you grow your savings, you can build your credit, and it works for the entire family. Plus, with their new Paycheck Advance feature, you can get up to $500 before payday when you switch your payroll and qualify. No credit checks or mandatory fees required.
It's super easy to sign up. Just go to Current.com slash OK. O-K-A-Y. That's Current.com OK to get the app.
For more information. In
In today's hyper-fast markets, it's never been more important to consider every option to raise capital, drive growth, and create value. Stay one step ahead with Strategic Alternatives, a podcast from RBC Capital Markets. In this season, RBC's experts will examine how corporates and investors are evaluating their strategic plans, reassessing their portfolios, and reallocating capital to help them lead today and define tomorrow. Tune in to Strategic Alternatives, the RBC Capital Markets podcast,
podcast today.
I want to go back to one thing that you mentioned about the Andreessen Sputnik comment, because I think it's really interesting. I think if you just walked away and perceived that as a negative, like if you were just to kind of not take it one step further and just say, oh, the Russians are going to run away with us. The Chinese, you know, the Chinese are going to run away with it. Well, the Russians didn't run away with it. It was an arms race, you know what I mean? For a while, but we started... That was the crazy part about this, just go full circle on this.
Sputnik is what actually started Silicon Valley in its current form because that started the kind of the initial venture arms race was around space. But keep going. Yeah, no, I think that's a great point. You're a student of history. I think that's really important to understand these different cycles. I'm a stock market guy. You're both a market guy and a technology guy. I just kind of pretend to be a technology guy when I have a mic in front of my face.
But I want to go back to that whole Sputnik thing because what your point is is that large investment went into how to innovate and how to beat the Russians. It had happened in Silicon Valley, and it was really the dawn of what we know to be one of the most innovative, you know what I mean, like little areas that exist on any part of the planet. So my point is do you think –
this causes a lot of these companies in the US to kind of hunker down. Does it cause an increase in CapEx or right? Because if it was an increase in CapEx, if the numbers that Zuck and the numbers that Microsoft just kind of alluded to over the last two weeks, 65 billion minimum, right? 80 billion for Microsoft, the prior being Meta. Do you think there's a chance that accelerates CapEx?
The simple answer is yes. And part of the reason why let's just zero in on deep seek. And so it's China, great performance, and it is probably not on the table for most companies to build on top of.
This is a great consumer prosumer, large language model. But just the fact that it is owned by China, I think, creates some limitations. They've been pretty well documented in the last few days about how they have some pretty tight rails on it. Very similar to Baidu. And I covered Baidu for five or six years out of China. And so I know that this dynamic is limiting to Western cultures to try to work with these companies. And so
The question is, does the model architecture that DeepSeek has created, does it somehow spring itself and find its way into kind of the rest of the world? And the answer is most likely it will.
And we in our venture business, like this concept of more efficiently trained models, we've seen probably 30 companies over the past couple of years that have been trying to do what DeepSeek is doing. And so this is something that, you know, enough shots on goal, something's going to break through. And so I think that as the back to the moment, what does it mean for most is that there is still this basic question that companies are having to
enterprises are having about ai which is how do i even begin to approach this like that's the level that's where the conversations are for most of them and starting to get some implementation in but the actual like use cases and and uh some of that uh the cost of the tokens of those we're still kind of in 2026 2027 and so i think if you can bring some of those costs down uh then i think that that will have a liftoff and again i come back to
that speed is of the essence. If we bring in DeepSeek into the club, there's six now foundational models in the world that are kind of making all the noise. But if you are just a half a step behind, if one mile is a half a step behind another model, you can be totally irrelevant. And so I think that, again, the race continues on. I think this is going to be something where it's going to continue to accelerate the demand for hardware. And I also believe that we're
we are going to see more adoption if, in fact, the cost of training these models does ultimately come down, not from DeepSeek, but from some of their architecture that they're talking about. Yeah, I think you'd have to have your head examined if you're downloading DeepSeek. You live here in the U.S. And I know you were using it for research purposes and the like, but when you think about what's going on with TikTok, you think about... TikTok, yeah, like this is... Well, this is, Gene, really quickly...
I think it's low-hanging fruit now for the U.S., for Trump to kind of do an about-face here. And this is an issue of kind of sovereign competitiveness in the technological world. And I think these were two of the highest downloaded apps as of like a few weeks ago in the U.S. I think they both have to be banned. You know why? Because Facebook and Google and Amazon and Snapchat –
and X, they're not in China. You know what I mean? So it makes no sense that we would allow these companies to be here, especially when we're in this digital arms race. Last thing, Gene, before we get out of here, there's two stocks that you and I know and love. You love them a bit more than I do. Meta reports Wednesday after the close. Apple reports Thursday after the close. Both of these stocks are up on the day. It's a bloodbath for most names here in the U.S. that are associated with this generative AI trade.
But Apple's up 4%. I think it opened down a little bit and then rallied. And that's a big move to the upside. Let's talk a little bit about why you think it is. We kind of alluded to it in the beginning of the show, but I'm just curious.
It's like, do you believe in this move? Do you like the move? Does it make you feel a bit more bullish? It kind of happened at a big technical level at 220. It bounced off. It was also a big 200-day moving average. I know you don't give a crap about charts and technicals, but you put some of that stuff together, it looked like a good spot for it to bounce, especially if you think that they're not particularly tied to the bear case in and around a lot of these infrastructure names and the spend that they've had.
If I would have gotten the script before, I'd have expected Apple stock would have been basically flat on the day with the Nasdaq being down a little bit. I'm surprised to see it's up 4%. And I think it speaks to just the amount of confusion that's in the market. The specific reason why I think it's up is that Apple is basically a consumer of AI compute. And so the prospect of the cost of these tokens declining benefits Apple's margins. I mean, that's kind of the
I think the essence of this is they're not building the infrastructure, they're buying it and the cost just gets a little bit cheaper. The reason why I think that the 4% isn't really justified right here is because the cost, even with those elevated kind of the old token pricing, it's still so small in terms of the grand scheme of what their overall OPEX is, what they would be spending on tokens.
And so I think that this is just a sign, I think, of the market trying to get their arms around, you know, is the software trade on now? Is this the lower AI hardware trade?
or, or, uh, inference trade, the lower AI inference trade. So I think that's what's going on. What matters is what's the iPhone going to do over the next year. And I think that what gets me optimistic is that the numbers are pretty low. The street expectations, there's been a lot of concern about China in the last few weeks in terms of Apple and trying to be down 18%. And the simple takeaways, I think they still will meet or exceed, uh, their, uh,
where the street is at for December. And I think they will guide essentially in line for the street for March in terms of where the iPhone's at. If they guide in line with where March estimates are, I don't believe it. I think they'll miss that. And I'll tell you why. If you were a Chinese consumer and there are 300 million iPhones that are set to be up
or supposedly because of the two, three years, whatever it is. And all of a sudden now, you have a model that you can download over there, okay, and use it on any phone that you want. And you understand that ChatGPT and Claude and Llama and I'm missing, oh, and Gemini are not going to be behind the firewall. You put it on there with your WeChat, which is your super app. Maybe you buy a cheaper Android phone. So we're in this digital war
with China, and they've made some very big nationalistic consumer decisions over the last couple of years. Just look at Tesla, right? And the drop-off that we've seen there. So when I think about Apple intelligence, it was never going to go to China. It wasn't going to be a reason to upgrade your iPhone there. Now you have a reason to kind of move away. You have a couple reasons to move away from Apple. So again, I get it. I think expectations are really low. I think the
upgrade cycle that a lot of folks were kind of hoping for with Apple intelligence is not playing out. I think if they don't upgrade the capability significantly by the time the next one comes out, then I think it's pushed out again until the 2026 thing. So again, you know, the 15% decline, it's taken a couple turns off that valuation. Um, you know, they better kind of point to, let's say better than expected back half growth. Um,
And then maybe investors stick around. I'm not particularly bearish on it. I did not like the fact that it broke out in November and just kept on going higher for no apparent fundamental reason.
One piece on the China is, you know, what's that impact? If you take the China business and the rumors are that it's down 18%, usually those rumors kind of overstate how bad it is. So let's say it's down 15%. That essentially takes two percentage of the overall iPhone growth for the quarter. So last quarter, in the September quarter, iPhone grew at 5.5%.
So if you assume that status quo, that the Apple intelligence is not kicking off, the street's at a percent and a half for the December core. It's at three and a half for fiscal 25.
And so I just think those numbers are, are, can basically are low enough. This is going to be a lot of fun to see how this one plays out. I think the numbers are essentially low enough to kind of weather the storm. All that said, I do believe there's another, I'd love to join you for another, uh, another pod where we talk about kind of where, where the, where the, where this,
the kind of the outperformance is going to come. I think that the Mag 7 is still going to do good. I think we're going to see other pockets. Let's hold it for a different day, but other pockets of the market that I think the tech market, they're going to start to do better. Yeah, I want to hear all about that. So hopefully we can do that really soon. Last one, Meta, it's up on the day here. It's an open source Lama, right? That is their large language model, DeepSeek's.
open source here. I would have thought that maybe this is one where developers around the world might kind of lean towards a deep seek because of a whole host of technical reasons like you laid out. Why is this stock up on the day, you think? Deep seek is China open source. And I think that Lama is kind of...
getting to the heads of developers that this is kind of the Western leaning, Western facing open source model. And I think that there's the stock is up today on a, on a belief again, the market's factoring a lot of stuff in really quickly here.
is that this idea that they're probably, I mentioned that all these small companies that are trying to break through and make improvements in the model architecture, there's probably something around Lama being well-positioned. I think that's what the market is saying is that Meta can have a deep seek moment.
Yeah. Listen, man, I really appreciate it. I know you're going to be really busy this week. You got Tesla on Wednesday. You got Meta. You got Microsoft. This is what we come to work for. I love this stuff. Yeah, and Apple. I love we didn't get to Microsoft and the relationship with OpenAI. I know that you've kind of highlighted the fact that there's been some kind of kinks in the armor over there as far as that relationship. So maybe we'll get a little more detail on that.
on that on Wednesday's call. But maybe hopefully if you had some time next week, we can check back in and do a little review of some of these thoughts. And I think it'll be fascinating to see week over week where market sentiment is around this. All right, my man. I really appreciate it. Good luck this week. I know you're going to be busy. All right. Talk to you soon, Gene Munster, managing partner, Deepwater Asset Management. Thanks so much.