
Loading summary
A
Welcome back to the Rundown, one of the top business podcasts in the world. Today we are talking to Doug o', Loughlin, the president of Semi Analysis. Semi Analysis has become the go to research company in the semiconductor and AI space. They post incredible reports breaking down everything from semiconductor supply chains to AI models. Today's conversation with Doug hit on a ton of subjects, including the market's reaction to Nvidia's earnings, corporate Google's launch of Gemini 3, why Google's TPUs are such a big deal, and why the AI bubble could get way crazier. You could tell from this conversation that Doug really knows his stuff and he's very passionate about it. Parts of the conversation got very nerdy but I think you guys are gonna really enjoy it. Alright, let's get into it. Alright guys. Today we are talking to Doug o' Laughlin from Semi Analysis. Semianalysis has some of the best research and reports on the semiconductor and AI industry. I always look forward to their reports. Doug, thanks so much for hopping on.
B
Yeah, thanks for having me. So, yes, let's chat.
A
Big week to come on man. Big week to come on. You came on at the perfect time. How sick are you of talking about Nvidia earnings?
B
Surprisingly this one, can I say, maybe guilty confession. This one was kind of boring. So it wasn't exactly like there's a lot of. Yeah, it was very boring execution. It's a, it's a perfect print. I have no problems with it. It's like exactly what you needed. It's better than buy set expectations. It's better than Street, a little better than what we thought would, they would do too. It's just a really good guide, a really good result. Even with the H20 not even coming back, they're still selling a lot of compute gross margins. You might have a slight complaint but that was kind of expected. I think it's a really good print. But I think, and like look, the markets took, it took that to be. It was a really good print to the morning of. Right.
A
But.
B
And then we faded all day. So I mean I, I thought that this was going to be the all clear. Obviously we even got the all clear and it was still a selling event. Which tells me it's not just about Nvidia in terms of the broader market environment. But I do think when it comes to Nvidia specifically, it was like such a good print. It's boring. You know, just imagine like, you know, they beat all the numbers for the most part and the guide was great but still it's not good enough.
A
So that's what was surprising to me was that how just like again, across the board, they just crushed it. There was zero concerns moving forward. Even the guide was great, but yet the market was just kind of like they brushed it off, which was I think concerning to a lot of people. Like, okay, well if, if this isn't enough to get the AI trade back on, does this mean that the AI trade is fully cooked or what else is going on? What was your take on that?
B
So I think, look, the AI trades had a really good year. Let's, let's be real, right? It's had an amazing year Even, even after a really good year last year. And honestly I would even argue partial part of the year before, right? Like into the end. I think, you know, stocks do go down sometimes and I think there's been a little bit of. Yeah, I know up only isn't, isn't just doesn't work. It is not going to be up only forever. And I think the, the reality is we've had a really, really, really strong run since the tariff tantrum. Right? I think, I think we've been above the 50 day for like I, this is a really surprising thing. I learned. We are above the 50 day moving average for the second longest streak in history. The. Just recently we, we broke through it. The second longest streak in history. I was like, wait, that's, that's kind of. That means we've been in a huge ripping bull market, which we have been. And I guess I was just really shocked. I was shocked by that. And so I think that this is just an example of like, hey, stocks do go down, they don't go up only. And yeah, most of what's actually driving, I think the price and narrative is the Fed, right. The fact that this morning, the reason why things have ripped at all actually is because some of the more dovish comments we went from maybe no cut to definitely no cuts and actually we're cutting again. So, so we went. It's so over, it's so back and. Or no, sorry, the other way. It's so back, it's so over. And now it's so back. Right? And that seems to be, yeah, moving stocks more.
A
So we're recording this on Friday morning, November 21st. And so this morning a Fed official came out and saying that, oh, a rate cuts back on the table when there seemed to be no hope of a rate cut. And so yeah, those comments are moving the markets back in the green. So it seems like the market is more focused on rate cuts right now. But like bigger picture though, if stocks, if this AI trade, let's just say if this AI correction continues to unwind and we still see a bleed, whether it's a fast bleed or a slow bleed in these AI names, do you think that that's going to change the decision making from some of the hyperscalers? Because you know, it's easy announcing $100 billion investment in a, in a data center in Louisiana when stocks are at all time highs. It's a totally different thing when you're doing it when you know you're off 30% from the peak you do you think that would that's going to change the decision from some of these hyperscalers?
B
I think it just really depends on which hyperscaler look meta. Okay, so I, I specifically Google and specifically Mark Zuckerberg, they're post economic money doesn't matter anymore to them. It's about technology and it's about winning. You don't become a billionaire because you're not a crazy sicko who doesn't want, he wants to keep winning. Right. I think Zuck part of the year of efficiency, right, wasn't like Reality Labs was part of it, but like it kind of just everyone else was doing it. I think that there's a belief in driving willingness to continue to figure out this new technology. And it was clear to me Reality Labs wasn't it and it is clear to me that AI is right. I messed around with an Oculus when, when Zuck was doing Reality Labs and I was like this is pretty cool, but I don't think this is going to change my life. I mess around with ChatGPT when obviously the ChatGPT moment I use Claudi's all in various amounts. And it is life changing. It is a transformative technology now like every other transformative technology in the history of capitalism, in our history, often there is a big capital cycle that builds capacity and supply to, you know, to be ahead of demand. And we're starting to see that right now. In the beginning it's very valuable and it makes a lot of sense to invest. In the end, it's overbuilt. And that's kind of the cycle that every single, every single one of these technology transitions go through. The Internet is the one that is probably the most recent one, so it's the most available. But there was a giant overbuild for even semiconductors at one point, even the automobile industry whenever that in the early, early 1900s that there was an overbuild There was an overbuild for railroads. Railroads are probably the most famous. Right, but yeah, but.
A
So, so I want, I want to stick to that point here. You mentioned, obviously the build out of the railroads. Everyone talks about that. There was a huge boom there and then a bunch of railroad companies went bankrupt. The Internet, you had a bunch of fiber build out, right? Fiber everywhere. But I think the. What people are concerned about with this build out is that like fiber became useful. All these years later, people still use fiber. The railroads are still used today, 100 years later. There's concerns around all this money being spent on data centers and specifically these chips. Right. What is the useful life of an Nvidia chip? If it. Is it just three years, four years. You have Michael Burry tweeting every day talking about how these depreciations are underreported and how the useful life is only like three years compared to what Nvidia is saying that it might be six years or even longer. How do you see that and how does that play into the bubble fears compared to previous bubbles?
B
So I think that's a great question. I definitely think there's a five to six year useful life, if it makes sense from the. But the way that we think about it is less about like accounting useful life and more like economic useful life. There is, I mean it's, it's, it's kind of complicated really. Like we could go into this in a lot of different ways. So it. Look, it's not going, you know, you're building all these assets and they are not, let's say, long term. But the thing is what happens is if you end up spending for another amount of, for a new chip, you know, you end up having meaningful improvement in parity. You're right that this is like, you know, it does create a little bit of a treadmill aspect, but we're also seeing the inverse of it where there's a lot of older chips, for example, H1 hundreds that are still being used today. So that's like the counterpoint, right, is that older assets that like from in this cycle still are being used. And Corey, for example, booked H1 hundreds at 95% of the original price for a contract. That's huge. But, but all of this is like, you know, you are right that there is this like kind of almost this crazy, this crazy like decline curve that's going to be really interesting to see. And I think that that crazy decline curve, you know, you see this in all kinds of things, like shale, for example, has a 12. Are you familiar with the shale well decline curve?
A
I live in Houston, so yeah, I mean I used to work in the oil and gas industry.
B
Okay, yeah, yeah, so, so right, this 12 to 18 month, it's the same kind of aspect here where like you have these crazy cash flow irrs and so you get all the cash up front and you can actually ironically reinvest in it. It's just a, a function of supply. Like, like what's going to happen is like you're going to keep a supply will keep, it's a function of demand rather supply is going to keep reacting to demand and as long as there's demand on the other side to purchase these things, they're going to, you know, people are going to reinvest their cash back into it and so you can actually have even this like more sick perversive thing like same brain mindset of like hey, well if you make so much money so fast on these shale wells you could just take that and do another shale well. Right, right. It can compound infinitely and also crash like crazy.
A
So that's the point. Yes.
B
Yeah, yeah. And look, I think that that's going to happen. This cycle will happen. Like I can't be like this, that that's not going to happen. But the way I look at it today is like there's a supply demand curve. We know that like supply demand of tokens. We know that there's a supply of. We, you know, we SM analysis work really hard to estimate and understand the supply of total tokens but we know there's a demand curve for tokens too. Now the biggest dream of all dreams is labor in the global economy. So you know, you know, $100 trillion or whatever number you want to say, like $100 trillion. Right. The reality is always going to fall short. That's just how these new technologies are. But we do think there's a meaningful real demand from most of these customers for most of these use cases and we think that there's a real line of sight for multiple billion people using ChatGPT or some proxy day to day in their life and paying some amount of money for it. Not 100% of them are going to be paid. Some of them will be free. There is a path for OpenAI and anthropic to be making triple digit billion dollar arrs that's valuable. Now the problem is triple digit arrs still can't even support some of the capex and I think that's the real crux and concern that people are having in the market right now. Yeah, it's real.
A
The $1.4 trillion spend that Sam was talking about. And then right now they're sitting at say 12 billion ARR, or whatever. Maybe it's 20 billion, whatever. 20. But that's still not even close to what that doesn't support the $1.4 trillion build out. Now maybe that's just a marketing number, right? It's not really a real number. But I think that when people, when people start seeing these big ginormous numbers and they see like the revenues coming in, it's like well, how do you justify everything? And then that's where the concerns start happening. Whenever the market starts dipping, the people start feeling worse about it. And then there's like this cycle.
B
Yeah, there's definitely a self, self fulfilling cycle of fear. Right, but let me specifically talk a little bit about that. It's a $20 billion ARR, right? So let's say it can support, we're going to do gross margin zero, whatever, that can support $100 billion of investment over five years, right? Because it's a five year useful life. I think one of the reasons why it's 1.4 trillion and some of that is double counting. Some of that is like over announced and some of it is like definitely it's very in this, in the, in the back half of all this. What happens is like they announce a hundred, $500 billion deal, right? There'll be like $100 billion of it will be like already investments in flight. And then they will like, like it's like okay, 100 billion, 100 billion 200, 200. And like so it scales and ramps. These are often very back half weighted deals. And these deals that are back half weighted, you pay with them in future revenue. So that's kind of the calculus I think is that you know, chat, you know, they still believe they can grow triple digit revenues, right? And I don't know, you know, 20 to 20 to 200, that'd be pretty crazy. Like 20 to 100 or something is gonna be a huge, huge, huge lift. But we're, we're talking about. There's never been, or maybe not never, but there's, it's really rare that you get to 800, 800 million users and you don't go straight to like 2 or 3. And so Internet scale will call us 3 billion, right? Like that's Google, that's Facebook, right? And 3 billion users. And they're not. And you know, the vast majority of them aren't even paying there's ways for you to make money from that. And So I think OpenAI has to show that they can monetize quicker and that's kind of going to be some of the push we wrote about how we think they will. We think agentic purchasing. Right. Which essentially is going to be taking a vig on every transaction that happens is going to be one of the ways they'll be able to quickly monetize things.
A
Affiliate fees, right?
B
Yeah, essentially affiliate fees. Right. Like that's, you know, like when you, you know, when you pay someone to pick up your order at doordash, it's a service fee. We definitely. That's going to happen in the near term. We're hearing all these partnerships. I think some of this concern about sustainability is going to accelerate that. Right. But then on the other side, yeah, I mean, it's a lot of money, bro. It's still a lot of money. So. Yeah, yeah.
C
I don't toast the holidays in a new way and raise a glass of Rumchata, a delicious creamy blend of horchata with rum. Enjoy it over ice or in your coffee. Rumchata, your holiday cocktails just got sweeter. Tap or click the banner for more Drink responsibly. Caribbean rum with real dairy cream. Natural and artificial flavors. Alcohol 13.75% by volume 27.5 proof. Copyright 2025 Agave Loco Brands, Pojoaquee, Wisconsin. All rights reserved.
D
So you're about to make a trade based on a friend's text, but which u do you listen to is it we could buy a house in Tulum, get optioning those options. We could lose everything. Or let's do a little research, get your head in the trade and make the investment decision that's right for you. Learn more@finra.org TradeSmart.
A
Sorry, what I was going to say was like, I think the. I think from my perspective, I feel like the shine has come off a little bit from OpenAI though. I think the interviews that Sam is doing and then like what the CFO Sarah Fryer said about potentially getting government backstops, I think the shine has kind of come off. That hasn't helped the narrative. And I think Google has done a really good job kind of coming in. Sundar has just kind of gotten like he's locked in and now Google just launched Gemini 3 this Nano Banana 3 Pro, which is just viral on social media. And I feel like they've kind of come in and taken the all the hype. Now when it comes to AI, do you think that Like Google coming in with their existing cash flows with an existing business kind of helps. I don't know, it helps like, remove some of the fear about an AI bubble because there's a company that's actually doing it with cash flows and not.
B
Necessarily, I don't know, big promises and dreams.
A
Yeah.
B
Yes, yes.
A
What was your take on this?
B
So, yeah, I think it's a great question. Okay, so, okay, let's put this in this way. It would be a lot healthier, for sure. I think, in fact, it'd be extremely healthy if Google did all of it right. We would know, hey, it's probably be money good, right? They have the full stack. It definitely works out. Probably bad for stocks, though, if you think about it. All of these things, I mean, you could argue that some of that demand would be be fungible and whoever ends up winning the 3 billion users will be using all the chips. And so, hey, OpenAI, all of a sudden, those large clusters that they signed up for, they're not going to be OpenAI clusters anymore. They're going to be Gemini clusters, right? So that would, I think, be kind of some of the thought process. But, yeah, I think it would be bad. And that's kind of one of the reasons why you're starting to see some of this fear too. But what I would really, what I really truly, utterly believe, at the end of the day, what it's about is about rates, right? All of this is a function of, like, I think of it as a ginormous party. Pretty much every single one of these technology builds. You have these ginormous capital builds. That's just history. Like, that's just how it works. Capitalism is extremely obsessed with margins. And so every 20 years completely throws that away and says, fuck it, we're building the biggest. You know, this is the future. And all of the, you know, all of the entire world is going to point at the same direction to get this done. And so I think that that's, that's kind of what, what we're seeing and where we're at. We're starting to see all of the world kind of point in the same direction because they want this technology. You know, whether if it's, you know, shareholders, to a certain extent, shareholders are rewarding Gemini for winning, right? You can, you can see credit markets are very happy and there's a lot of demand for debt. The government is very, very, very yolo and all in on this. And so you see all of these things kind of lined up and pointed in the same direction. And that's, it's kind of a big deal, man. And so the thing that, that kind of is like okay, two sides to the coin. We're probably already spending it at a rate that is unsustainable unless of a lot of revenue comes soon. But the other side of it is that like if people are really all in the upside of what that looks like, especially if you have uneconomic spending from the, from the big players is very, very, very, very large. You know I learned recently that Microsoft Lens at essentially let's say 15 bips above the United States government. That's crazy. That's crazy. Like yeah, there's right. Yeah, they have no debt too. Like these guys all have an ability to really invest if they really want to. I don't think that's Microsoft right now, but I do think, well, I mean actually after the satya interview with Dylan Patel, business partner at Semianalysis. Right, yeah, shout out to semi analysis. But there's a lot of. I think that Microsoft's back. But you have to remember who is the most aggressive of the players. That's going to be Meta and Google because they're still run by founders. Founders are really important in this because at the end of the day it is a person based game and these people do not want to lose. And so that's kind of what we think is that the capability and how the market is rewarding people for winning has kind of led to this incentive to win at AI at all costs. And so I think that that's where we're at. It could get a lot crazier. But then also this whole thing is like, let's just say that's the party, okay. That's the energy, the vibe. How I think about how crazy the party can get is all function downstream of capital and in this case capital and how cheap it is from let's say the Fed and long term rates is the alcohol, okay? Now if the beginning of the party, it's really expensive. These drinks are 40 bucks a pop. You know, you're like ah, I have to be at this party, I'll pay, I'll, I'll buy one drink and leave. Now if you cut rates and you cut rates and you cut rates and you cut rates and all of a sudden they're like yeah, drinks are free. You're going to get pretty drunk, dude. So that's kind of where I think that that's what's happened is we good analogy. We've seen, we've seen the rate cuts and so one of the reasons why there's a freak out is I think, and I thought, honestly, the market appreciated this, is that we, it became very consensus we were not going to cut rates in December. And now after this morning, people are like, wait, are we cutting rates? And so there's kind of this like, there's a little bit of a tape, a little bit of a tantrum around rates as well. And so you have all of these things kind of at this like weird inflection point. That's just how it works. I can't tell you definitively what's happening. And that's kind of a weird thing because usually I'm like, oh, I know, I know, I'm familiar all my clients, I talk to everyone. But at this moment in time it's kind of weird. It's just a vibe shift. And so yeah, we'll see, see where it lands.
A
I, I like. So the, the, the connection that I'm seeing is that, okay, it's crazy right now, but it could get crazier because if the Fed cuts rates, it makes borrowing money even cheaper. And we haven't even entered the debt phase of this bubble where like companies like Microsoft, Meta started borrowing money. But Amazon, Google, they haven't even really started tapping the banks yet to borrow money and really pour money into the, the build out. Oracle started to do that, but nobody else has really done that. And once that starts, that's when things get fun slash crazy and we haven't gotten there yet.
B
Yeah, that's a bubble. Okay. There's a big difference. Like I think it's really hard for you for an entire thing. Like yeah, it'd be bad for example, like, and like look, I think free cash flow hasn't gone to zero. That's really what it comes down to. The hyperscalers can move free cash flow to zero. If they move free cash flow to zero, then like all of a sudden.
A
That means borrowing more money and spending everything they make.
B
Every single dollar they make. Yep. They spend every single dollar. Right. Everyone is not spending every single dollar. And so spending every single dollar is one thing we think. I mean, I think at this point in time we'll probably go there just because like Google for example, their business model is the most under attack they've ever, it's ever been for sure. Right? Yes. Even Microsoft, man, all these companies have, these have some of the greatest, most profitable businesses in the history of capitalism. And you can argue they've also are at the biggest threat in their entire lives too. And that's all because AI So you, you kind of have this concept where you have to continue to like, maybe, maybe you can pull off the gas. But look, Microsoft pulled off the gas and it really hurt them last year. And so now, you know, it's kind of this weird race to. You have to stay in. You can't, like, you know, if it's, if it's going to be a lot longer race, you can't go, you can't go out too fast. You could argue Oracle is going out too fast. They're running out of capacity. Right.
A
They're betting the company on the, on the.
B
Zack.
A
They're betting the entire company.
B
They're betting the entire company, like, you know, pretty much. I mean, if things go really poorly, there is an actual, there's an actual case that Oracle could go like, bankrupt. Crazy. This thing's been, it's a very old company. Larry's been doing this for a long time. That'd be absurd. Right? But that's just how it is. Like there's, you know, and so in this whole race, yeah, it's kind of a crazy thing. I still think. It's not like we're not going to call it early, but the potential for a true bubble is much, much, much crazier than today. And I think, as you know, if it all stops, probably bad for stocks, but I still don't think it's, I think it would be a drawdown, but it wouldn't be like this. We wouldn't go down like 40. It wouldn't be a bubble. Right. Like, it would be this painful adjustment period. But free cash flow can pay for a lot of things. And so that's okay. OpenAI takes the mark. It's very painful, whatever. But yeah, it's definitely not too late to stop and have like, no damage. But I don't know where we're at. It's just this weird thing you feel everyone pushing forward And I think Gemini 3, ironically, is going to be a huge shot across the bow. Like, you know, they haven't been working on pre training in OpenAI and that's, you know, they can, if they really take it seriously, that this might be a really good competitive threat to get OpenAI back on track.
A
So sticking with Gemini 3 and just Google in general and please stop me if you have to run. I just want to, you know, stop me whenever you have to. But I wanted to ask about Google's TPUs, which have gotten a lot of attention, right. TPUs are, they're personal and I'm not very technical. The TPUs are Google's specialized in house built chips. And some people are saying that this might be a legit threat to Nvidia's business as well. Because if Google ends up winning the whole thing, everything's going to run on TPUs and then no one's going to need GPUs and you know, and black belt chips. Can you just kind of explain that to someone who's, to someone who's not technical on like how big of a deal these TPUs are?
B
Yeah, these CPUs are number two. There's a, and like some, for some people there is a belief that they're the best chip that's available to purchase. Give me one second.
A
You gotta run.
B
Feel free. So let's just do this last TP thing and I'll bounce. Okay. So TPUs are tensor processing units and they're by Google. They are also manufactured with their design partner. Sorry, they're designed and they're co designed with Broadcom, manufactured at tsmc. So the thing that's different though is Google makes it and designs part of it so they don't have to pay as much. You know, you talk about 75% gross margin. Imagine if instead of paying a 75% gross margin, you're paying a 50% gross margin, you know, 50% less. That's a lot of money that you get to. And so because you pay 50% less, I think I'm using that math right. I might be wrong because you're paying so much less. What happens is that you're able to get to have better cost of ownership when you purchase a tpu. You're spending less on what you're effectively getting. And you own your data center and you own your own model so you can margin stack in a way that no one else can. That's the beauty of, that's the, I mean that's the beauty. That's really what it comes down to, right? When you own every layer of the stack, you only need to make profit once. At the very end you can, you can eat the rest of it as a cost of goods sold. And so that's, that's the beauty and that's what, that's what Google strategy was, is focused on. To be clear, Google strategy was sucking until very recently, let's not forget.
A
Right Bard? Oh my God.
B
Yeah, Bard. Also you have to bet on Google being a really good, a really good like, like essentially like have you noticed that they kill products all the time and they kind of are, are like a, kind of a crappy company when it comes to google.com killed by google.com but they're also crappy company when it comes to execution. On the back, on the back of this new release, what I do, and I do this for every single new model is I force myself and what I do is my action button effectively opens a model. So my action button right now is to Gemini. But it's like, really? I'm like, wait, am I on the right thing? It's kind of a shitty experience. Straight up. I'm going to be honest with you. It sucks compared to OpenAI, compared to Claude. You can just tell there isn't that level of finesse. There's a desktop app for Claude, there's a desktop app for, for ChatGPT, there isn't one for Gemini. And there's just like all these little details where they just aren't super ultra commercial and because if they don't, if they don't figure this out, Google doesn't die versus the other two. They do. And so I think that that's the, the difference is two startups versus a big incumbent. But the big incumbent does have a lot of advantages at this moment. So yeah, that's, that's, that's kind of the game. So. Yeah, yeah.
A
30 second question. Does OpenAI IPO before or after January.
B
2028 check before or after 2020? January 2020. So that's like. Yeah, I mean, so look, I think it's been pretty widely rumored that they will do before because they're, they're like, look, these checks are getting so big, right? There really is only one check left that you can do like this. And it's public markets. Public markets don't quite scale to infinity, but they're pretty close. They're the most scale that we've ever had. Right. And so, you know, a trillion dollar company, you know, let's say, let's say he wants to raise $100 billion for trillion dollar valuation. That's, that's only capable. There's only one place in the entire world you could do that. So yeah, I think that that's the, that's the dream. We'll see. Okay. Right. Yeah, I, I think, I think that it demands, right. There's this pace that, this, this pace that Sam demands by, by essentially always raising, always investing, always like you, it kind of forces all the things to move forward. Even if you're like, and, and pretty much is if you're not keeping up with that pace, they die. So that's, that's the plan. I think that really is a superpower.
A
Fundraising is his superpower, man.
B
He's 100% at it. Yeah, 100%.
A
I really, really appreciate it, man. Everyone go check out semi analysis. Every time I get a report in my email, man, I feel like I wish there was like a chat GPT explained like on 5 like buttons because it's such a detailed, well thought out report and I have to spend like 45 minutes reading through it multiple times. So highly recommend everyone check out semi analysis. And Doug, I really appreciate the time, man.
B
Yeah, thanks for having me. We should do this again. Take care.
A
Oh, yeah, man. Thank you. Well, all right, guys. Hope you enjoyed that great conversation with Doug o'. Laughlin. Like I said, wide ranging topics. There was some more topics I wanted to ask him, but we ran out of time. I'll have to save those questions for the next time Doug comes on. If you guys enjoyed today's show and have like five extra seconds, consider giving us a five star rating on YouTube, Spotify, wherever you listen to your podcast. And if you are listening on Spotify, don't forget to vote in today's Spotify poll. Leave us a comment on Spotify. All that engagement really does help us out and it helps other people find the show. Thank you guys so much for listening, watching and commenting. Shout out to Mike and Connor for all the work behind the scenes and we'll see you guys back here on Monday.
Host: Zaid Admani | Guest: Doug O'Laughlin, President of SemiAnalysis
In this episode, Zaid Admani hosts Doug O’Laughlin, president of the influential semiconductor and AI research firm SemiAnalysis. The pair dive deep into the nuances of Nvidia’s latest “perfect” earnings, the broader market’s reaction, Google’s Gemini 3 and TPUs, and the current state and future trajectory of the sprawling AI investment cycle. The conversation is candid, technical, and insightful, addressing both the hype and the skepticism surrounding the AI trade.
Nvidia’s flawless execution:
Doug calls Nvidia’s recent results a “perfect print,” fit for high expectations, beating nearly every forecast even without certain product lines making a comeback.
Quote:
Market’s muted response:
Despite Nvidia’s strong performance, the stock market responded with a selloff. Both discuss how this shows the market’s attention has moved beyond single company performance to macroeconomic factors—especially interest rates and Fed policy.
Quote:
Reflection on the AI trade’s bull run:
The hosts note that AI-related stocks have performed phenomenally over the past two years, but “stocks do go down,” and the streak above the 50-day moving average was unusually long.
Quote:
Focus shifts from earnings to monetary policy:
The Fed’s shifting stance on rate cuts is now the primary market mover, overshadowing even stellar earnings.
Quote:
Party analogy for capital cycles:
Doug compares rate cuts to cheaper drinks at a party, fueling more exuberant spending and potentially riskier behavior in AI and tech.
Quote:
“How I think about how crazy the party can get is all function downstream of capital...capital and how cheap it is from...the Fed and long term rates is the alcohol.” — Doug (18:36)
“If you cut rates...and all of a sudden they’re like yeah, drinks are free. You’re going to get pretty drunk, dude.” — Doug (19:39)
Continued investment “at all costs”:
Doug emphasizes large tech companies like Meta and Google are “post-economic”—driven by a will to win at all costs, with investment incentives disconnected from near-term profit.
Historical bubble/cycle analogies:
The current AI boom is likened to historical overbuilds—railroads, fiber, semiconductors—where excess capacity was built ahead of actual demand.
Data center lifetime economics:
There’s concern over the useful life of high-end AI chips; Doug suggests a 5–6 year “economic useful life,” but admits fast innovation causes rapid hardware obsolescence, creating a treadmill similar to shale oil wells.
Gap between revenues and investment:
Current AI company revenues do not support the monumental capital investments being made and announced (e.g., Sam Altman's cited $1.4 trillion spend vs. $20 billion ARR for OpenAI).
Self-fulfilling cycles of hype and fear:
Narratives of both exuberance and fear create feedback loops in how capital is allocated and how companies react.
Monetization challenges and affiliate/agentic models:
Doug suggests future monetization may rely on taking a cut from actions and transactions conducted through AI assistants (affiliate-like fees), but stresses even then, reaching escape velocity for these investments remains a tall order.
Google’s comeback:
With Gemini 3’s launch, Google’s AI division is regaining excitement and “hype,” benefiting from deep cash flows and an existing ecosystem.
TPUs as a threat to Nvidia’s dominance:
Google’s in-house TPUs (Tensor Processing Units), co-developed with Broadcom and fabricated by TSMC, offer cost and performance advantages since Google controls the stack and can operate on lower gross margins.
Execution and risk with incumbents:
Despite their technology, Google struggles with consumer-facing execution (“killed by Google” syndrome) and lack of refinement in newer releases like Gemini.
We haven’t hit the “debt phase” yet:
The real mania may start if hyperscalers begin funding massive AI investments with borrowed capital rather than just free cash flow.
Potential for a true bubble:
Doug warns the bubble can get much “crazier,” but a pullback now would merely be a painful correction, not a catastrophic collapse.
“There really is only one check left that you can do like this. And it's public markets...Let's say [Altman] wants to raise $100 billion for trillion dollar valuation...there's only one place...you could do that.” — Doug (27:10)
“Fundraising is his superpower, man.” — Zaid (28:15)
This fast-moving episode is both a sober and energetic examination of the AI investment ecosystem as of late 2025. Nvidia’s stellar earnings highlight that even best-in-class results can’t overcome shifting macro winds, especially as interest rates come to dominate the narrative. The discussion contextualizes current AI capital spending in the long history of tech overbuilds and warns that things could get much wilder if we reach a full “debt-fueled” bubble stage—something yet to materialize.
Meanwhile, Google’s Gemini 3 and its TPUs represent a return of serious competition for OpenAI, though both hosts acknowledge persistent execution risks for incumbents. Ultimately, the fate of the AI bubble will hinge on whether real economic returns catch up to hyped investment and whether the global capital party continues, or money becomes dearer once more.
For more in-depth semiconductor and AI industry insights, check out Doug O’Laughlin’s work at SemiAnalysis.