Ben Thompson谈大科技、中国与资金将尽的AI热潮 - [Invest Like the Best, EP.487]

So, Ben, if you can believe it, look how long it's been since we last did this. The world was very different. There was no AI at the time. We talked mostly about aggregation theory. I thought a fun place to begin, since the world has changed so much, is to hear what you think it would mean for the US to win the AI race.

1. US Dominance Creates Danger

I think it would be very problematic for the US to win. Let's say we take the most fantastical scenario, where if you control AI, basically, your military is better than anyone else's. Somehow, it fixes our manufacturing and all these things that I don't think AI is necessarily going to do because they deal with the real world. But in this world, what is the game-theory-optimal response of China? To blow up TSMC.

Game theory can get very convoluted and complex. To me, this one actually isn't that complicated. There is a fundamental disconnect that I have with a lot of the rhetoric coming out of Silicon Valley, coming out of one of the labs in particular. If we get to a place where we have meaningful superiority from a military and national security perspective, I think that's very dangerous for the world.

But in that state, how much does it extend beyond TSMC being blown up? Because in that state, I would assume we figured out how to build fabs here in the US, to some degree, and are less reliant on that one choke point.

I think there's a little bit of magical thinking, which I just invoked, in terms of manufacturing—whether it be fabs, actuators, or all these precursors. I think the degree to which we are dependent on China is underappreciated, and it's not something that is going to be fixed outside of a conflict. Fixing so many of these things is going to be dramatically dumb.

If your competitor is sourcing from China and you're going to start sourcing or getting things from the US, you're going to be at such a disadvantage, relatively speaking, that you're just not going to do it. You do it when you have literally no choice. That works for very big headline items. You can browbeat Apple to move some of its iPhone manufacturing to India, for example.

But even that is a good example, because Apple is not truly moving out of China. It's diversifying to an extent, but it would just cost so much. It's like paying an insurance policy that, if you don't have to pay it and it's astronomically expensive, you're just not going to pay it.

It's one of those hypotheses that I just have a hard time even grokking, because the only world I see where we truly pull out and have no dependency on China—such that if they want to blow up Taiwan, who cares? There's no impact on us—seems pretty fantastical to me. I think there's a bit of facing reality in this regard that is not present in these conversations.

Put yourself in their shoes. What do you think the motivations are?

Everyone can use a good bogeyman. I think from the AI trade perspective, nothing works better than, “We have to beat China.” And I do think we need to beat China. We need to be competitive.

I despair at the extent to which, over the last few years in particular, so many of our responses, particularly from a political perspective, have been to try to be like China. I think we should be going the other direction: more openness, more innovation, less top-down control, fewer restrictions on speech, and things along those lines. America succeeds by being on the leading edge and by leaning into that.

You said that the US being purely dominant in AI is probably not the right end state for the world. What is your ideal equilibrium for how this goes worldwide?

2. AI Settles Into Equilibrium

There's a bit where AI right now is kind of like the Taiwan situation, in that the current status quo actually doesn't seem so bad. The question is, how sustainable is it? But maybe it's sustainable for longer than we think.

The way I think about it right now is that OpenAI and Anthropic are clearly on the frontier. Who knows what's happening with Google? Then Groq and Meta are chasing them. Meanwhile, the Chinese are very capable and very smart, and are also definitely distilling these models to stay about 6 to 9 months behind. It feels like a pretty good equilibrium that I think is generally favorable to the US.

Now, the question is, how long can it stay this way? There are lots of questions there, such as whether the Chinese can actually pull ahead. I'm still a little skeptical for various reasons, whether it be chips or anything else. Getting to the leading edge when you're 6 to 9 months behind is very difficult.

It's going to be instructive to see how Meta and Groq do in terms of actually catching up, especially as we get to a world of AI improving itself—using AI to make AI better—which I think is definitely a real thing. You see a real acceleration from both OpenAI and Anthropic recently. That was theorized, and it seems to be coming true. To the extent that's true, can you actually catch up?

The other question about this, by the way, is to what extent that applies to cost to serve—to marginal costs. If you can apply AI to optimizing your stack, figuring things out, and analyzing all the data, is your cost to serve structurally lower than anyone else's?

This is the thing about the open-source models. The talk about them being free is bizarre to me because it's marginal costs. You still have to run inference, like with GLM or Kimi. Kimi is very expensive to serve. The cost per answer is significantly higher.

When everyone refers to these as free, it feels like, in the narrative, people have it in their heads that free is free. “Now I can use AI for free.” No, you can't use AI for free. You're not necessarily paying the R&D cost to create the AI, but you're definitely paying for the inference to run it.

Now, I kind of like where we are. The pushback would be, “That's right now. It's not going to stay that way,” which I think is a fair pushback. But I don't know—it may last way longer than we think.

If you could know anything about the future of how this will go, to be more confident in where the equilibrium will end up, what would it be? Is it the length of the S-curve—how far up the S-curve we are? At some point, these things presumably will level out, maybe not. What would be the thing you'd want to know that would give you a better sense of what the future might look like?

I am concerned that, with the scare around people freaking out about Mythos and this Hugging Face incident, the actual implication is not that we reduce these dangers, but that we just stop releasing stuff. We, on the outside, start to lose any sense of where exactly—

What's actually the frontier?

—what is actually the frontier and where it is. There becomes a false sense of security, because right now everyone's basing their understanding of Mythos on Fable. But how good is Fable actually relative to Mythos? That gap is only going to increase over time.

So I think that's a real question that I'm not sure about. This question of the recursiveness of AI making itself better—does that lead to a takeoff? At the end of the day, there are timing questions in lots of different ways. I'm worried about the timing mismatch in terms of the actual return on investment producing enough revenue to fuel investment. We're working our way down the capital curve. We started with free cash flow.

The speed with which the tech companies blew through the debt markets is incredible. It took a year, and now Google’s issuing equity. NVIDIA’s putting together this—

This $500 billion thing.

This $500 billion thing to tap into pension funds, insurance floats, and things like that. What’s after that? Where does the money come from after that? Well, ideally, we actually flip back to free cash flow funding this. But if there’s a gap there, if we don’t get there soon enough, then we could have a big blowup. At the same time, even if we have this blowup, AI is not going away. It’s not going to stop improving. It’s going to keep progressing in a way that makes us look back on the dot-com era, the railroad era, or whatever bubbles throughout history as ultimately immaterial in terms of the broad scope of humanity, even if they were very devastating.

What can the railroads teach us, do you think? It’s now the last big build-out, right, in terms of percentage of GDP or getting there.

I think we might be bigger at this point, or it was the biggest.

It’s in the ballpark.

3. Railroads Expose The Funding Risk

The railroads had a real duration mismatch. To build a railroad and make money off it was a decade- or multiple-decades-long endeavor, whereas you had to issue money to pay for it in the short term, and the world ran out of money.

Got it.

Right? I think that is probably the aspect—that’s why people reach for the railroads. Everyone talks about whether we’re going to have enough compute or enough electricity. Maybe the nearest-term question is, are we going to have enough money? That’s a bizarre thing to think about. That’s what happened in the 1870s. The world just ran out of money. The funny thing is, the railroads kept operating, and they expanded the West. Their contributions to GDP were astronomical. They’re still contributing to GDP. Railroad money is what’s going into Google right now from Berkshire Hathaway. Like, was—

That’s very funny.

No, it—

It’s quite literal.

Berkshire Hathaway has this problem. To me, this NVIDIA deal is very much paired with the Google equity issuance, which I thought was shocking when it happened.

Why was it shocking?

Because it’s Google. They can’t raise money. Why are they issuing equity? Why are they reducing their upside if they believe so strongly in this? But the Berkshire comparison is interesting because, to a rough approximation, they have See’s Candies, famously, right? It’s a tremendously high-margin business. The problem with a lot of high-margin businesses is the percentage profit you can make is very high, but the absolute profit you can make is capped.

No reinvestment runway.

That’s right. You’re just accumulating cash. The brilliance of the BNSF Railway thing was that they took the See’s Candies profits and said, “Here’s another industry whose margins are way worse, but the absolute dollar amounts are so large that those much worse margins result in absolute profits that are much larger.” BNSF in 2025, or something—the amount of free cash it threw off in 1 year was more than See’s Candies had thrown off in its entire lifetime, even though you’re talking about a low-margin business compared to a very high-margin business.

I think there’s an aspect from Berkshire Hathaway where, once your capital gets so large, you start operating in a world of absolute numbers as opposed to percentage numbers. The reason I thought that story was so interesting is that it seems to capture where Google itself might be going. It was very symbolic for them to invest in Google. Google has this unbelievable high-margin business of search, one of the most perfect, beautiful business models of all time, and the purest aggregator of them all. It scales in every direction, doesn’t have to invest any money to do it, and everything’s zero marginal cost. It’s amazing.

Meanwhile, there’s this AI opportunity, which requires just astronomical amounts of money. It’s incinerating cash. But you can imagine that if AI is intelligence, and its TAM is basically all white-collar work, and eventually, with robotics—

Probably more, right? Yeah, yeah.

Everything, potentially. The absolute profits available here, even if the margins are lower, are so much larger. Will we look back and see that Google Search was See’s Candies? It feels like that’s what’s happening. In that world, you use all your free cash flow. They’ve done that. You tap the debt markets to the tune of hundreds of billions of dollars. They’ve done that. You issue equity. What does an equity issuance do? It dilutes your interest in the company as a shareholder, so you have a smaller percentage of the pie. Well, if you have a smaller percentage of an astronomically larger pie, at the end of the day, no one’s going to be complaining.

It is very symbolic that Berkshire is the symbol of that equity issuance. Is Berkshire actually not just an investor in Google, but a model for Google and where they’re going?

I’m curious: setting aside the commercial and competitive components of this, how AI-pilled, on the pure technology, would you say you are relative to other people thinking about this space?

4. AI Struggles Beyond Verifiable Domains

I have a view that is both super bullish and less bullish in some respects.

Okay.

I am not fully convinced about the generalizability argument. AI is clearly incredible at coding. It blows my mind that people were doing this a year ago, actually writing out code. It’s very good at math, obviously, but the obvious riposte is that these are verifiable domains. What is the evidence, or where is the compelling evidence, that being very good at verifiable domains cleanly translates to being very good at unverifiable domains, or domains that have very long verification loops? I think that’s still a little bit to be determined.

It’s interesting because I raised this question, and there were some people at the labs who were on a panel. I was annoyed at the answer because the answer took me for an AI bear. “Oh, well, people thought we couldn’t solve chess or we couldn’t solve Go, and we solved those easily enough.” I’m like, “I thought we could solve chess. I thought we could solve Go because they’re knowable domains.” Scale was the answer to both of those, but both of those were also bounded.

What is the go-to example that’s not chess, not Go, that is genuinely in a new, unknowable space, where it’s doing things that were not possible? That is the “I’m not fully convinced” sense.

However, AI, at a rough approximation, is trained on all the data of the internet. All the data of the internet—that’s distillation. It distilled all of the end state of human thought. The actual—

No traces.

Typing on Reddit. It doesn’t have the traces. It doesn’t actually have the thought, the emotion, or whatever that went into typing that comment or writing that essay. Say Neuralink, whatever. What if the actual payoff from Neuralink is capturing the traces—

Training data.

—of human thought that actually dramatically expand the capabilities of these models?

My sense is that a huge number of jobs, a huge amount of economic activity, does not exist in these domains that I’m not convinced AI is good at. Actually, there are a lot of people in the world who are, to a certain extent, like sentient AIs. They operate very well in verifiable domains. They’re given jobs, they do them, and it’s almost like a somewhat pessimistic view of humanity, to a certain extent.

But I think that market is so large that, if the models did not improve at all from where they are right now, the economic opportunity would still be massive. I wrote an article a while ago. There’s the whole accelerationist movement. I called myself a reluctant accelerationist. I think we need to push forward because we can’t go back, and the worst thing we can do is get stuck where we are.

So I’m very AI-pilled in terms of its impact on the economy and its upside in terms of monetization. I’m not sure about the timing.

What would be the gradient toward it? Imagine law or medicine, where I don’t know whether or not you would consider those verifiable. Law is a code of some sort. Medicine—we have a certain state of understanding of things.

I think medicine is by far one of the biggest opportunities. It’s both one of the biggest opportunities and one of the most challenging ones because of all the regulations and all the access. If you could turn AI and machine learning loose on all the medical records, I think the number of discoveries and improved treatments we could come up with very rapidly would be unbelievable. So that is a very optimistic view. On the flip side—

It’s not happening.

When is that going to happen, right? I think the optimistic frame I put on humans is that our capacity to create needs is sort of unlimited. I think we’ll do a very good job of creating new opportunities and jobs in the fullness of time. The more pessimistic way to put it is that our ability to create red tape and muck is also fairly unlimited. How much of our economy is actually made up of more and more jobs we’ve managed to create that just make us busy and make us slow, to a certain extent?

If I go back to the early 2010s, maybe aggregation theory was stewing in your brain, and then you published it in 2015. I think it's fair to say that theory, that idea, defined the winners and losers of that era of technology. I'm really curious how you're thinking about what theory or principles will define this era of winners, from a financial perspective and a market-cap perspective.

5. Aggregation Theory Meets AI

I go back and forth even just on the question of aggregation theory itself. How much does that apply in this current era?

Just the same thing. Yeah, yeah.

Yeah, because a pushback that people have is that one of the key components of aggregation theory is zero marginal costs. Zero marginal costs shows up in lots of ways. The one that I focused on at the beginning was distribution. People say, “Oh, I don't have distribution. I have to pay Google threads.” Well, no, you have a website. Your problem isn't that you have distribution. Your problem is that you don't have demand, and you're paying for demand when you're paying for ads and things on those because the aggregators control demand.

They control demand because, in a world of abundance, the hard problem is not distribution; it's discovery. How do you actually find what you're interested in? So the companies that solve discovery in their domain come to dominate that market. They get a virtuous feedback loop. That's aggregation theory in a nutshell. The other thing is transaction costs. There are no transaction costs. Google can scale to the whole world, and they can scale to the whole world not just on the user side, but also on the monetization side. The vast, vast, vast majority of advertisers on Google or Meta never talk to someone at Google or Meta. They just go up and buy ads. It's all done by computers.

The perfect business.

And those computers, from a business perspective, cost $0. AI obviously changes that significantly. Inference costs are real. But then again, how real are they?

They're real right now.

I don't know. Are they?

It depends on the company, but they're way more real than those prior examples.

For sure, but you have this incredible spread. You have people—I think the vast majority of people who are using AI today are using it as basically a Google substitute or a recipe maker, or whatever it might be. My suspicion is that the cost to serve those people is extremely low, basically similar to serving them a webpage. I would imagine it's marginally higher, but not that much higher.

Then you have, on the other extreme, people who are actually leveraging test-time scaling. It used to be we just scaled by making the models bigger and bigger. Now you can scale as far as time: How long do you think about the answer? Well, you could think about the answer for days—

Days.

—or weeks or months. That is directly marginal cost. Every second longer you're thinking is costing more money. This speaks to how we think about AI and inference as one question. That's why I was pushing back on you. But actually, the marginal-cost question for the different users—the user using free ChatGPT and the user trying to solve a math theorem—they're not even remotely in the same universe.

I think you see this challenge actually in the enterprise in a very interesting way. Microsoft recently shifted its enterprise plan. They came out with an E7 plan, $100 per user per month. That includes some amount of usage, but then they also are charging for usage on top of that. I think this is a kind of fraught position for Microsoft because the positive way to think about Microsoft is that they do everything you need as a business. Every individual component might not be the best, but you get it all for one price, and they all mostly work together. If you're particularly a small or medium-sized business, or even a large enterprise, there's real value in that.

That's right. It makes life easy. The moment you start having to think about how much you're paying, it's not just that that's a new decision; it's also untethered from headcount. Microsoft got the benefit when you were hiring a new employee. You would think about the cost of that employee, and baked into the cost of that employee was $100 a month or $50 a month for their license. It was a thoughtless revenue stream for Microsoft.

Now, if you think about usage, you have to think every single month, “How much do I want to spend?” That introduces 2 problems. Number 1, most companies aren't set up to do this. They make budgets once a year. This idea that we're going to be thinking about our budgetary allotment on a monthly basis doesn't compute.

There's an aspect where they're used to thinking about CapEx decisions or one-time costs, and there's a bit where, when I'm talking about the loaded cost of an employee, it's not CapEx, but it's kind of like CapEx. You make the decision up front, and you don't think about it anymore. The decision's sort of already made. But if you're thinking about usage, you have to do it again.

The final thing is, if you're looking at your Microsoft bill every month and asking, “How much did I use?” you start thinking about, “What am I paying for? How good is each of these products? Should I actually just start thinking about spreading this out?” I think they had to do it because that extreme user who uses a ton of tokens and is actually leveraging AI costs way more to Microsoft than $100 a month. They can't support them, but they want to hold on to the set cost for the vast majority of employees who can fit in that. They need to ask their customers to think a little bit for those extreme employees, but they don't want them to think too much because that breaks the model in very surprising ways.

Are you surprised at all that the recipe builder user, who is very low cost to serve, hasn't had a great business model emerge around them just yet? Google and Facebook sort of perfected the business model in this prior era. They haven't seemed to figure this out at all.

6. Consumer AI Needs Advertising

I am frustrated but not surprised. This is obviously a market that should be supported by advertising. That is why advertising is always the consumer business model. Consumers don't want to pay. There are 2 things to understand about consumers that Silicon Valley has to relearn about every 10 years. Number 1, consumers do not want to pay for software, and number 2, consumers do not care about being productive.

We went through this in early SaaS. The canonical company for this, in my mind, is Dropbox. Dropbox was an unbelievable product, especially when it first came out. In business school, I was one of the first people to use Dropbox, and that went off like crazy. I have so much storage still in my free Dropbox because I gave out my code to so many people.

Drew Houston made this amazing product that was so easy to use and absolutely seamless. He was very clear about this. He wanted to build a consumer company, and there's that famous story of him meeting with Steve Jobs. Apple was interested in acquiring Dropbox, and they were like, “Oh, we wanna build a company,” and Steve was, you know, “You're a feature, not a company.”

That plain-Jane, just-file-sync product—Apple did make it a feature, as iCloud Drive. With Dropbox, they grew very fast, and then they had a 2-year lull. In that 2-year lull, what they had to do was basically completely rebuild the app from the bottom up because not enough consumers were going to pay for it.

Enterprises could see the value. But if you want enterprise, you need permissions, you need control, and you need someone else to be able to set all these sorts of things. Their app wasn't created to do that at all. So they had to rebuild the whole thing and realize, “The only way we're going to make money is by selling to companies.”

Why do companies pay? Because companies are paying employees. To the extent they can make their employees more productive, they're getting a greater return on their investment. It's the complete inverse of a consumer. A consumer's like, “I spent all day working. Why do I want to come home and be more productive? I want to sit on the couch and watch Reels.” But you see that with AI, and you also have this overarching skepticism of advertising.

I've gotten so much traction on Stratechery by being an advertising appreciator, and I go back and read my early articles about advertising that were directionally correct but also not very good at all. But I got so much traction doing it because I was the only person writing about advertising.

In a world where everyone wanted to have a blog and Twitter, no one wanted to talk about advertising. But even now, in Silicon Valley, there's this embarrassment about the fact that the Valley is, in many respects, monetized by advertising. Particularly during the last 8 years, there was this sense that Facebook was icky.

The best engineers don't want to go work on this problem.

And so you literally had OpenAI replaying the Dropbox story, but at 100× the size, being like, "No, we're going to sell subscriptions to consumers." They did. They sold a lot, but they didn't sell enough. If you're going to be in the consumer market, you have to be doing advertising.

They're doing advertising now. It's a little weird that they finally pivoted to doing advertising. At the same time, they're like, "Oh, crap, we need to go after the enterprise because Anthropic is kicking our rear end." So I'm not quite sure what they're doing there.

They have been rolling out ad features very rapidly, things like copying the connections with retailers, so you know if a purchase went through and can do all the tracking and things like that. I'm very interested to see how that goes. There's a bit where, had they leaned into advertising immediately, as soon as ChatGPT was a hit, I think they would have a killer ad product right now.

I think Google would be in much bigger trouble. I think Meta would be in much bigger trouble because if you have this flywheel—the thing about advertising with consumers is your ability to monetize the consumer—

Goes up as the volume goes up, yeah.

—is infinite because the advertiser is bearing the price increase, so there are zero elasticity issues. If you're charging consumers a price, if you want to raise the price, like Netflix, this is their problem with the subscription plan. How much can they raise prices before consumers rebel and drop a tier or give up the service entirely?

Charging people money is hard. Giving people things for free is easy.

Is easy.

And it's very frustrating that OpenAI did not pursue this sooner.

I know you've been spending time with some of the big money firms and sources of capital. What is your sense of their appetite right now, and how are they thinking about the future? Because I think this year it's going to be 800 billion or something that we're going to spend in CapEx. Next year is supposed to be 1.3 trillion, I think, is the current estimate. It's going to keep going up from there.

We're burning through all the compute that gets installed basically immediately. It's such a strange circumstance that we can use the capacity right away, as soon as it's online.

7. Compute Scarcity Masks Commodity Risk

Well, that's the thing, though. There are a few timing mismatches happening right now. We can't use it right away. All the bulls on Twitter are always like, "We don't have enough compute, we don't have enough compute."

We don't have enough compute because there was insufficient investment made in 2023 and 2024, which, yes, absolutely, is true. And by the way, if you think there's not enough compute, TSMC decreased its rate of growth in 2023, in 2024, and in 2025. Our shortage of compute is going to get worse in the next few years because the lead time for a fab is even greater than for a data center.

Today, when we say there's not enough compute, it's not like all the money that the companies are putting in today manifests in compute tomorrow.

That equals a shortage of compute, yeah.

No. It all manifests in compute in 2028 and 2029. On the calls, both Andy Jassy and Satya Nadella are out there saying, "Look, we're just building data centers. These are the shells. We might not use them now; maybe we'll use them in the future, and we only buy GPUs when we know there's demand for them."

That is a great story to tell. I'm not sure that I think it's a lot of BS because the reality is, if you've built the shell, that money is sitting there. You're not going to let it just sit there. If you've invested a fixed cost—and this is the whole logic of commodity markets—I think tech in general doesn't understand commodity markets.

Tech is, by and large, focused on: If I produce a highly differentiated product, and that differentiation could be software, it could be a network in terms of developers, it could be a social network sort of thing where we're peer-to-peer—where I'm highly differentiated—then my ability to charge higher prices provides my profit margin.

The classic example is Apple. They have their ecosystem, their software, third-party products, and all those things, so they can charge 50% margins on their iPhone. Everyone looks at Apple as the ideal business model. That's how you run a business.

But in a commodity market, the price is set by the marginal supplier.

Cost to serve is all that matters.

That's right. I had a good friend in Taiwan who is in shipping. It's a fascinating industry. It's kind of like the airlines, too, another industry that I love to look at.

You buy a ship, and the cost of that ship is depreciation. Your marginal cost is actually quite low. It's the fuel to run the ship, the cost of the crew, and your port fees. Not that much. What that means is, you're going to run that ship—

As full as you possibly can.

No, you're going to run it no matter what.

Yeah.

And you're going to bring down the price of a container as low as it needs to be to cover your marginal costs. Now, your paper losses in this situation might be very large because your accounting loss includes depreciation, but the depreciation is an accounting figment. You already paid the money.

You're going to run that ship at whatever the market will bear. And the container—the beauty of the container is that it is a pure commodity—the price in the market is going to be the marginal cost.

Now, if it gets low enough, at some point people will exit because their marginal costs mean they're actually losing money on a shipment. Not just paper money, but actual, real money. They will exit, but then the supply is diminished, so the price will go back up, and you get this interplay of coming in and out.

But then, if the market's very high, like it was during COVID, it's like, "Wow, we're making so much money right now because there's not enough supply." There wasn't enough supply of ships, so containers went from usually being 3,000 or 4,000 to 17,000 or 18,000. The amount of money that these shipping companies made in a very short amount of time was insane.

What happens, though? Well—

Build more ships.

Imagine if we had more ships, right? The problem is it takes 2 years to build a ship. If everyone makes this decision simultaneously, you suddenly have a lot of ships, the price plummets, and so forth.

Where we see this is in components, in memory in particular. Memory is famous for boom-and-bust cycles, with people entering the market late. But to what extent are data centers going to be memory makers?

Right now, everyone can see we don't have enough compute. So everyone's like, "We absolutely have to be investing because there's so much money to be made. And look at our payback period." The problem is, you're measuring your payback period in a time of scarcity. Is that payback period going to hold in a time of abundance?

The bulls say there's never going to be a time of abundance. AI—

We're going to be short forever.

That's test-time scaling.

Yeah.

We're going to be short forever, which maybe we will be. My concern is, even if that's right, we could still have an air gap in that there's so much money going into it right now and not enough has come online to actually make sufficient revenues to handle the situation where we run out of capital.

I believe in AI. I think it's a real thing. I think the economic impact is going to be astronomical. I think all the concerns about societal impact are very real and are going to come to bear in a major way.

You can believe all that and still be worried about whether we're going to make the bridge to this actually generating the level of returns necessary to continue fueling this going forward.

Can you zoom in on TSMC and the component makers where fabs are involved? So far, at least my understanding is that they've been quite conservative in their willingness to expand capacity and build new fabs, and they've not met the market's demand with similar growth.

Is that just rate-limiting this whole thing and preventing us from getting one of these giant overbuilds?

8. Chipmakers Shift The Risk

We can talk about a few different ones. We'll start with memory. Memory used to have tons and tons of memory makers. Every time there'd be a boom, memory makers would reenter the market. New countries would come in—Taiwan used to have a memory market, for example—but you would get these exact dynamics.

If there's a shortage of memory, there's so much money to be made, but you can't bring capacity online immediately. It's the same as shipping and the same as what we're seeing right now. That would spur people to come into the market, you'd get too much capacity, prices would plunge, and people would just get blown out.

The issue is that the upfront cost for these is so large, just like buying a ship. Building a fab is even more expensive. And memory now—the leading edges of memory are using things like EUV machines—so the costs are getting into the billions of dollars for these lines.

What happens every time with these boom-and-bust cycles is that some people would enter, and more people would get washed out.

You go through these famous historical moments for these memory cycles. Companies just get blown out. One of the most interesting memory stories is how Samsung sort of took over memory. They saw it as an opportunity, studied history, and realized that the way to take over the market was to invest into downturns so that they were ready when the next cycle came around, which requires a ton of guts, discipline, and money.

But they did that and basically wiped out the Japanese. That’s when the South Koreans generally took over the market in a major way. But it got down to 3, and the problem with 3 is that it’s not a monopoly, but it’s kind of an oligopoly. They all became a lot more disciplined about not making the mistakes of the past. They’re not colluding, but they’re all on the same page about not doing that.

I think that dynamic ran head-on into the current moment, where it just took a while for them to realize that there was a secular shift in memory demand that didn’t exist for a very long time. I think the memory solution will be solved eventually. The other risk they run is Apple’s lobbying to get Chinese memory. What is the number-one focus of algorithmic changes? How can we use less memory?

I think the memory makers probably screw themselves in the long run by creating such a massive target on their backs. I’ve analogized memory makers to Iran. The issue with the Strait of Hormuz is that it’s very effective. It’s more effective if you don’t use it, because then it’s always hanging out there as something you could do.

Now they did it, and it turns out it worked, but the UAE and Saudi Arabia are going to build pipelines and new ports. They’re not going to let this happen again. It’s very painful right now, but say Iran wants to close the Strait of Hormuz in 2035—it’s not going to have any effect because it will have been built around.

My concern for the memory makers is that they might have done the same thing. No one’s going to let themselves get into this situation again as far as memory goes. TSMC is arguably worse because there’s only 1. There is 1 company on the leading edge. Obviously, Intel and Samsung are trying to get there.

It’s the same thing. All markets carry risk, and a lot of the question is who ends up holding the risk? What I think a lot of the tech companies didn’t fully appreciate is the extent to which TSMC has offloaded risk onto the big tech companies.

The way they’ve done that is that the risk TSMC is worried about is overcapacity. If we build too much, it’s not just that we built too much and have all these fixed costs that aren’t being fully utilized. If we build a fab, we expect that fab to run for 30 years. We baked in too much capacity into the system for years and years and years. So they are very biased toward being much more conservative.

There’s a little bit of a culture component to this, too. One of the most interesting TSMC stories, analogous to that Samsung story, was that Morris Chang retired in the late 2000s. New leadership took over. There was the Great Recession, and so they pulled back their planned spending.

He comes in, fires everyone, and says, “The iPhone just launched. This is the biggest opportunity we’ve ever seen. We need to be investing, not cutting.” They invested through the Great Recession and through that downturn. That’s what laid the foundation for them taking over leading-edge semiconductors during that time.

Morris Chang is a one-of-one on my Mount Rushmore of the greatest and most impactful tech executives of all time. The entire fabless model is so critical to what tech is and what it does. There was also just the guts to do that at that time, particularly for someone who lived there, in a culture that doesn’t necessarily tend to make those sorts of bets.

TSMC was pretty conservative, to be totally honest. So what happens, though? Where did the risk go? TSMC is saying, “We don’t want to take the risk.” Risk doesn’t disappear. It just moves.

The risk is right now that every single big tech company realizes that if we had more compute, we could be making more money. So there’s lots of foregone revenue and foregone profits that are the manifestation of the risk TSMC handed off to them. Risk doesn’t disappear. It just gets handed off.

Sometimes that risk doesn’t manifest in losing money; it manifests in not making money. There is money not being made right now because they were very excited about 5G. They did a big wave of investment and expanded their fabs around 2020, 2021, and 2022. They said, “Oh, yeah, we’re good.”

Like I said, ChatGPT came out in 2022 and was a big thing in tech in 2023. In 2024, their growth rate went down. In 2025, their growth rate went down. In 2026, it’s up now.

It was very funny because I was writing about this a while ago, and then I think it was 1 or 2 earnings calls ago. Suddenly, C.C. Wei, the CEO and chairman, was talking about use cases for AI throughout the whole earnings call in a way he never had before.

This is why the memory makers are scared. Usually, there’s a bullwhip, and they’re worried about being at the end of the bullwhip, where demand happens and works its way down the chain. They’re at the end, and then they double down. It’s already too late. They’re wasting all their money.

I think the thing with AI is that if it’s a bullwhip, it’s the longest bullwhip of all time. There’s still so much to be built, and it just took a while for Asia to get the message to these companies. I think they’ve by and large gotten it, but once they get the message, it then takes several years for that to actually materialize.

Do you have a sense for how long you think it will take, given the extreme shortage of compute?

The interesting thing is what this means for Intel and Samsung’s logic businesses. I’ve been writing about the problem of this dependency on TSMC for years. One of my first articles, in 2013, was exhorting Intel. I said, “You have to build a fab business. You’re not going to be a designer anymore. There’s a huge business in manufacturing chips.”

I thought I was late writing it then. Their stock went to the moon throughout the 2010s as they rode the cloud wave. It wasn’t until 2020 that they finally realized, “We fell behind. By the way, there’s this huge opportunity. We’re totally unprepared for it. We don’t have a customer-service mindset or culture, organization, or all the IP building blocks—all these things that TSMC has.”

And they need a customer. They need customers to help them actually build a real foundry business. So I would write about this as a problem, and I would write about the China issue: You’re dependent on a company that is 60 miles offshore from our greatest geopolitical opponent, who thinks it’s theirs. These are big problems.

That’s where I came to appreciate this insurance issue. If a big tech company went to Intel and said, “Intel, you make our chip,” the biggest benefactor of this is going to be Intel, because it’s going to learn how to work with a partner. The biggest pain is going to be the big tech company, because it’s going to have to figure out how to work with Intel.

We could just go to TSMC. They are awesome. They are so great to work with. We know they’re going to do a good job. It just never made rational sense for anyone to go work with Intel. That was their fundamental problem.

In an unchanging world, TSMC would just win forever. But this is where TSMC, in some respects, made the same mistake as the memory makers and made the same mistakes as Iran, if I can continue the analogy.

Because they didn’t invest in the last few years, the shortages are going to be so acute. Big 10 companies that we're foregoing so much revenue and so many profits because they don’t have enough compute will go through the pain of getting Intel up to speed and getting Samsung’s logic business up to speed. The scarcity is what ultimately saved Intel.

I expect that at some point they’re going to announce some major partner for the first time, and it’s going to be a big deal. But ultimately, TSMC brought it on themselves.

It’s the “cure for high prices is high prices” thing.

For sure.

We’re going to route around them.

Yep. There are all these things that, as an analyst sitting on the side, you can write about, and it’s one of those things I learned: No one’s going to pay for insurance that they don’t need when that insurance’s expected value is negative.

The way to solve the geopolitical problem of dependence on TSMC is to come up with a compute use case that is so massive that everyone is economically incentivized to bring other people up to speed. Then we get the geopolitical insurance for free.

If you think about the top 10 or 15 technology companies, which ones do you think have the most interesting setups for their businesses today?

9. Tech Giants Choose Their AI Strategies

The answer’s always Amazon. The reason Amazon is so compelling is the extent to which they build for themselves. They are their first-best customer. They provide the scale to get basically anything off the ground, which they then sell to other people.

AWS is the most obvious example. AWS, contrary to popular thought, was not spare Amazon capacity. It took a long time to get Amazon.com onto AWS. What it drove was the understanding that we can’t be having so many meetings. We need to have compute that you can plug in—a pure API surface. You don’t need to talk to anyone; it’s just there.

And, oh, by the way, if we do that for our internal retail teams, we could do that for anyone. It turns out the retail business is so big, we have to start with everyone else.

AWS actually started serving external customers before it served internal ones, but now it serves them all. You've got other products like, say, logistics, where it was the opposite. Right now, we're using external providers for our logistics—UPS, FedEx, and USPS. We need to build this up ourselves. Now they've built it up themselves, and they're offering it to third parties. Other people can use their delivery services.

You see this in market after market. They're talking about some of their AI products or their chip products. What's the beauty of the Graviton or the Trainium, particularly the early versions? The early versions were terrible, but if you're on Amazon and you're using some of their managed services, like, say, their Redshift database service, they don't tell you what the processor is underneath that. You're just buying a managed service. So they can put all their crappy processors underneath the services they're selling, and that gives them the volume and capacity to—

It's better, yeah.

—iterate them and get better, and they get to the point where they can actually sell them externally. Because they were the first best customer for Graviton, Graviton got better. Because they were the first best customer for Trainium, Trainium got better, and now Trainium is obviously running Anthropic's AI products. We'll see if any of them take off. They have call-center software. Their call center, or their customer experience, is going through AI. By the way, it's pretty good.

I haven't tried it.

Moving back to America, I've been buying lots of stuff. Every summer, I'd buy lots of stuff in a very brief amount of time. Sometime in the last year or so, you can go on and you're clearly talking to a chatbot, but the chatbot does a great job, and it actually does take care of the problem, so you could see that starting to work in that regard.

They're building up these AI services for their own business that they're going to make broadly available, and some of them will work, some of them won't. It's such an elegant approach, given they have so many investments in the real world. Their core business feels so impervious to the model version of AI. It will benefit from AI, but their moat feels deeper than anyone's as far as their core business, and their ability to generate new business lines organically is very compelling.

What about Apple? They sat this whole thing out, it seems.

It feels like it might be a situation of better be lucky than good, to a certain extent. Apple has their whole ecosystem. At the end of the day, they do own access to customers, so they can get suppliers. This is the classic aggregator play: If you own access to customers, suppliers come to you, not the other way around. So they can get suppliers for their AI as needed.

To the extent it's true that people don't want to be productive and just want a chatbot, not only can Apple serve them a chatbot and finally get a Siri that works, but you can see a future where this absolutely can work on-device, and they don't even need to pay for inference costs because they're using the customer's electricity. I don't think we're quite there. There's a reason they're using Google Cloud and NVIDIA chips, but you can certainly imagine a future where that's the case.

They're in physical goods. Actually making phones is hard. Having retail and distribution for physical goods means they're more insulated. The smartphone is so perfect. It's small enough to fit in your pocket, but it's big enough to watch basically anything on. You can run your whole life on it. All your entertainment is there.

When we talk about customers who just want to be entertained, the TV is now an accessory. It's all on your phone. I don't see anyone taking over the phone. The question is: Is the phone always going to be the center? Or is there a bit where, particularly in the home, this is where OpenAI's efforts are very interesting, where you want an ambient AI that you just talk to and it tells you what you need?

Apple is the best positioned to provide that, but can they provide that without having leading-edge models? Can they provide that if they're so phone-centric, or is it like a Microsoft situation? Microsoft didn't miss mobile. They were very early to mobile. The problem is their mobile was a small PC. They assumed the PC would always be the center, and their phones were going to be something that was based on that.

Apple realized, "No, we need to reset." The iPod helped them realize that. Microsoft went with Windows and all that, but will Apple fall into a Microsoft-like trap, assuming the phone's so good it's always going to be the center, and then figuring out what goes around it? Or is this finally the time when ambient AI, the cloud—just AI in general being everywhere—can manifest through your phone, through a device, or on your computer, and is actually better and disruptive to them?

I think it's possible. I also think it's totally valid for Apple to double down on what they do. The other thing about the AI stuff is: On what basis should we expect Apple to be good at this? At the most crude level, AI is this probabilistic endeavor. Apple is the king of deterministic products.

A physical product—you ship that iPhone, you ship it once, and it's got to be good. If it's bad, it costs you billions and billions and billions of dollars. Apple's never had an iPhone recall. It's amazing. That care and decision-making and diligence and fierceness in terms of your supply chain, and making hard decisions, is very, very different from everything that goes into making great AI. I generally prefer companies to do what they're good at. So from my perspective, I'm fine with Apple not doing AI. I want them to keep making great devices.

Of the five potential frontier AI winners—OpenAI, Anthropic, Gemini, SpaceX AI, Groq, and Meta—which of those firms do you think has the most interesting setup?

OpenAI and Anthropic obviously are the riskiest, but also have the biggest upside. Never discount the power of belief. They think they're creating God. The most impactful things in history have usually been fueled by religion.

The 2 religious organizations in Silicon Valley are OpenAI, which is kind of like mainline, and Anthropic. They go to church every Sunday. They're sort of like evangelicals. It is core to their belief. That goes a long way.

The fact that you need to make a business work for you to survive goes a very long way. Google just needs search to not die too quickly. Meta has the huge advertising business. In a world where Meta was run by anyone other than Mark Zuckerberg, they would not be on the leading edge. That is one of the purest manifestations of founder energy, for better or for worse.

Their business is so amazing. You see them easily doubling down on that. Google has Google Cloud and TPUs, and they've been doing research in this. It makes sense why they're pursuing this. Meta being like, "Actually, we're going to hire a completely new team and we're going to start from scratch"—this, again, is pretty insane. Credit to Mark Zuckerberg in that regard. Again, you can decide whether that's a good idea or not.

And then SpaceX AI, data centers in space, the theory is there. Do they have to own their own model, though, to do that? They'd get better margins if they do. Then again, if we actually run out of data centers on Earth, whether through political opposition or power or whatever it might be, they can run whatever model they want, as we're seeing with them selling their capacity to Anthropic right now.

They're all pretty interesting. The case for SpaceX AI is probably the weakest because the data center-in-space play is so highly differentiated. If that plays out, I'm not sure to what extent they need to even have their own model, so why are you wasting billions and billions of dollars in the meantime? That's a fair question.

From a tactical perspective, I love the Cursor acquisition. That makes so much sense for both companies, and so I've been intrigued to see what they do. Meta is probably the most interesting.

You've written a lot about this recently.

I think there's a very good case to make that it is more reckless not to be on the frontier if you're a digital company. The counter to Meta is actually Microsoft. Microsoft is not on the frontier. Microsoft had $40 billion of free cash flow last quarter. Microsoft paid a $10 billion dividend last quarter. There's some money, but their play is, "Oh, we're going to play all these off each other. We're going to provide middleware. We're going to provide the platform that enterprises will build on us, and we're going to disintermediate the models."

I think it's a rational play. It's the IBM play of the '90s. History echoes. Everyone talks about Google following in Microsoft's footsteps, but Microsoft follows IBM, and you can see that to an extent.

What did IBM do? What's the analogy?

Well, IBM had this dominant position. We talked about it in the '70s, and then you fast-forward to the '90s, and IBM is this very distressed asset. The thought was that IBM needed to break up into all these different pieces.

Lou Gerstner comes in and takes it over. Gerstner's real key insight at IBM was, "We're pretty mediocre at everything." It's kind of like what I talked about with Microsoft before, and that's the price of monopoly. Once you've been a monopoly, you kind of lose your capacity to be good because you didn't need to compete anymore.

I think a lot of incumbent tech companies have this problem. It didn't matter what they did; they were going to rake in money. If you don't have the pressure, if you don't have the incentive, if you don't have the fear of death or the fear of God, as we talk about these monopoly companies—

Then you don’t do your best work. The problem is that once you lose that muscle, it’s gone. You’re just sort of fat and flabby. So what Gerstner realized is that the worst thing IBM could do would be to break it up into component pieces, because all those component pieces are actually not very good.

Our biggest asset is that we’re big. It’s like, what? No. What does it mean that we’re big? It’s the ’90s, this Internet thing’s coming along. There are all these companies that know they have to figure out the Internet, and they don’t know what to do.

They need someone who can come in, understand their business, and help them get online. That’s basically what IBM did. So they built out—and this is an echo of what’s happening now—a huge consulting force, and they put all their time into building basically middleware. They would go in and put this layer between a company’s old-school mainframe, which all these companies had, and modern web services on the other end, so they could have websites and e-commerce sites and all these sorts of things.

That gave IBM a 30-year lease on life. Yes, in theory, you could go get point solutions from all these hot Silicon Valley startups, but you don’t understand that. You don’t know how to do that. You know us. We’ll come in, we’ll create all this middleware, build this big consulting force to help you implement it, and you’ll get online.

IBM basically brought all of corporate America online. That’s Microsoft’s playbook. Microsoft will help you figure out AI. It will help you figure it out in a way where you’re not giving away the crown jewels to these companies. We’re going to build this platform, this harness, this sort of middle layer. We’re dependable, we’re stable. You know us. We have backward compatibility to the ’80s. You can build on us, and then we’ll manage all the changing models and what’s updating and do all those sorts of things.

Does that mean you’ll get the absolute best experience? No. Middleware saws off the sharp edges. You sort of get a lowest-common-denominator capacity. This is the oldest enterprise sales motion. How did Oracle go to market? Oracle went to market in the 1980s with Larry Ellison and another technology taken from IBM—or IBM just didn’t want it: relational databases. They said, “You don’t want to be locked into IBM. Relational databases you could run anywhere. Come with us.” The reason this is a joke is because Oracle—

Everyone’s locked in to Oracle.

—they lock you in more than anyone, right? But all enterprise sales are companies whose long-term goal is to lock you in, getting you on board by trying to make you scared of being locked into somebody else. All the cloud companies are like, “Oh, portability, whatever. You can do whatever.” They’re like, “Oh, just use our service that only runs on our cloud, and now you’re locked in.” That’s Microsoft’s playbook. It’s a very rational playbook, and I think it makes sense.

That’s why they have extra money, because they’re not on the frontier. They are building massive data centers, but they’re building data centers for inference. They’re not building them for training, and their story that they’re investing in response to customer demand is more believable in that regard. They’re not having to tell a fungibility story where they’re building big data centers for training that will be used for inference down the road, maybe.

Go back to this notion that it’s reckless not to be in the front.

The reason why that’s concerning, though, is that at the end of the day, why are we using Microsoft products again?

Because we did before.

To what extent does it actually make sense to have all these artifacts, all these documents, all these email inboxes? Can’t AI just do that? There’s a real threat here to Microsoft’s software business. The whole systems-of-record thing is funny, because one reason why systems of record are so powerful is that it’s so hard to move them somewhere else, because it’s a very tedious, repetitive job.

AI is actually surprisingly good at that. Microsoft isn’t so much a systems-of-record company. They do have some of the Dynamics business, but it’s the user interface. It’s where you actually interact with the computer.

That’s the part—when you see Codex, Claude, Cowork, or whatever—it is aimed like an arrow at the heart of what Microsoft has. In the long run, all digital companies are threatened, but Microsoft is very much so. Their strategy is sound. It’s also desperate in an existential way, and also in a they-might-pull-it-off-because-they’re-desperate sort of way.

Meta is not threatened immediately, but this is where my bullish view of AI comes in. I think all digital companies are threatened, and Meta is a digital company. They have software. One worry is that AI takes up more and more time. Time ultimately is Meta’s currency.

We saw OpenAI try the Sora thing, but it didn’t really take off. Social networks are actually pretty hard; they also cost a lot of money. It’s really interesting. So this came up with the creator-payment stuff.

YouTube very famously has paid creators from the beginning, and that’s a much bigger drag on the business than people appreciate, because YouTube has marginal costs for its content. Now, unlike Netflix, they don’t have to pay that cost upfront. They’ll pay it after the fact, so their revenue sharing is a better model than Netflix’s model. Netflix has to pay upfront for content and then ideally make more money. YouTube pays along the way, but Facebook, or Meta—

Pays nothing.

They pay nothing. Instagram is this unbelievable product that generates all this money for which Facebook pays zero dollars for content. It’s unbelievable.

It’s funny because you could see a world where, for YouTube, AI-generated content could theoretically be a positive, because the inference cost of generating content could be less than what they’re sharing with creators. For Meta, AI-generated content, to the extent they’re the ones generating it, is actually a worse margin profile than what they have today, because what they have today is free.

So they have attention. There’s a bullish world where Meta is actually very well placed, because in a world where we’re interacting with AI all the time, the desire for a human connection becomes greater, and it’s Meta going back to its roots.

One of Meta’s biggest mistakes, actually, was that Meta was always a social network company. They killed Snapchat, or stopped Snapchat’s growth, by realizing Snapchat was a great product: “Let’s layer it onto our network.” They brought their network to bear to kill Snapchat.

The reason why TikTok was just a blind spot for them is that TikTok is classified as a social network, and it’s not a social network at all. TikTok is an entertainment product. It doesn’t matter who you follow on TikTok. What you see on TikTok is a function of what you watched, and you’re going to get more of the same.

It’s a user-generated content network, and the insight from TikTok was that limiting the best content to your social network is an artificial constraint. We’re going to give you the best content from across the whole network, and the vast majority of content is going to be crap, but this is an absolute-numbers question. You don’t think about margins; you think about absolute numbers.

Even if the margin for great content is infinitesimal, if we have a ton of content, the absolute amount of great content is going to be very large. Then Meta is like, “We’re a social network.” Meta is serving you content from your network of people you know, and TikTok is serving you the best content from around the world. That’s why they took a huge chunk out of them.

Meta had to shift. That’s what’s happened with Instagram and with Reels: it’s not really a social network. It is an entertainment product that pulls from the entire network, and social networking is like the group chat.

It’s possible in AI, actually, that social networking is important again, because we actually want humans. We want to have some sort of connection to them. That will be interesting to see how that plays out.

But the other thing with the models is they’re so impactful on advertising. The biggest impact of the models—the biggest monetization right now—is probably not Anthropic or OpenAI. It’s the incremental gain that is happening for Google and Meta.

Most of the stuff is pre-LLM, but we’re getting to LLMs, whether it be generating advertising content. What do we want? We want verifiable domains. How do you verify if a generated image is good for an ad? Does the ad sell or not?

They actually can validate their image creation and their text creation in a way no one else can, and their validation is the ad marketplace—running a gazillion A/B tests on all these different things to see what works and what doesn’t. Most ads are a throwaway. It’s fine. The vast majority of ads don’t convert.

They have this massive advantage, this huge liquid market that is a verification machine, where the verifiers are humans deciding whether to click on that ad and make a purchase or not, but they’re doing it at global scale. That can actually create a feedback loop to make their products better.

You’re also going to get a world where ad matching is actually still fairly crude. Here are the qualities of the person; here are the qualities of the ad. It’s like you create an embedding, a vector calculation, and see what numbers match, and then you sort of match an ad to the person.

What do LLMs do? LLMs predict. We’re going to move to this world where Meta’s going to look at people and say, “This person probably wants to see this next,” and they’re going to go find that thing and show it to them. The potential upside in terms of showing people better ads that are more relevant to them is enormous. They only need to increase a few percentage points for the returns to be billions and billions of dollars.

This alone is worth Meta investing in being on the leading edge and having these amazing models. I think a big problem Meta has is that they don't tell this story. It's weird, but Mark Zuckerberg has the same problem Sam Altman does: he doesn't love ads.

They have the best ad business in the world. They have an ad business that I think is a societal positive. You and I have set up these little content businesses that make great money. Content is—you get a ride on social media. I grew up on Twitter, with people sharing my links. It was amazing.

If you're selling some product, the beauty of the internet is that there is a niche out there that wants that product. The question is: How do you find the niche? Facebook advertising. That's what it does. It helps products find the people who didn't even know they wanted that product, but when they get it, they're so happy they got it. That's a huge societal positive.

You have a new business from a new entrepreneur making a new product. You have customers who are happy they got something that they didn't know they would get otherwise. Those customers, by the way, got lots of free entertainment, and they didn't have to pay for it along the way. Meta made a bunch of money for itself and its shareholders, which is basically everyone in the world.

This is why advertising is great, and Meta's advertising in particular is awesome. I get frustrated that Meta doesn't talk about that. Mark Zuckerberg has never really talked about the societal benefits of advertising, except in passing, in 20 years. He's handed it off to other people to take care of, and maybe there's a bit where him not paying attention is why there is a certain grit and grind that goes into building an advertising business.

People get frustrated or have questions about data and all those sorts of things, and maybe there was a bit where he didn't want to be involved in it and washed his hands of it. But you saw this when Apple passed ATT, App Tracking Transparency. It was one of the worst antitrust violations in the history of technology: Apple unilaterally obliterating all these business models while they're simultaneously building their own advertising business and doing all this tracking. Why? “Trust us.”

Meanwhile, they're running these advertisements. Remember that advertisement of people on the bus, overhearing everyone around them and what they're saying? That was such a dishonest representation of how advertising works on the internet. You had Tim Cook in Congress talking about companies selling data. Facebook's not selling your data. That's valuable to them. Why would they sell the data?

Meta was not prepared to respond. I think you saw this with Sheryl Sandberg back in the day. On every call, she would talk about advertising and how great it is, and have a bunch of case studies of people who were benefiting from advertising and these new entrepreneurs. Then she left, and it's kind of like that hole never got filled.

It feels like it's a company that's embarrassed. “We make a lot of money from ads, but we got glasses and we're doing AI.” It's like, you have ads, and ads are awesome. I think if they had communicated that more consistently, they would be in a better place generally from a PR perspective. They would have been in a better place relative to Apple, and I think they would have an easier time right now convincing Wall Street: Let us invest.

The other problem is they've spent a cumulative hundred-some billion dollars on Oculus, which I hated all along, and so there's a bit of: Why should we let you spend money again?

The one major player and company that we haven't talked about much is Jensen and NVIDIA, and I'm curious how you would tie this back to the notion of not understanding commodity markets in Silicon Valley, whether or not you think compute ultimately is a commodity. I'm curious whether or not you think intelligence will ultimately be a commodity. It's interesting that intelligence and compute, which seem to be by far the most interesting and important topics in tech, both might be commodities and less differentiated than tech's prior products.

10. NVIDIA Faces Commodity Pressure

Well, that's always the case, though. The most interesting thing about the internet is free distribution. Bandwidth is a commodity. The fact that I can pull out my phone right now and connect to any information source in the world for free, free on a marginal-cost basis, is because it's a commodity. It changed the world. Commodities change the world.

There is an aspect of differentiated products: By definition, they have lower TAMs because there's an elasticity aspect to it. Not everyone can afford to pay for it. People's willingness to pay is going to differ, and your market is going to be constrained. Apple is never going to serve the whole world by selling a device, whereas Google can because it's free. That matters.

You're paying for a commodity, but to the extent it is available to everyone is the extent it is impactful. The internet is a commodity. It changed the world, so I don't think it'd be weird that intelligence ends up a commodity and changes the world.

Commodities often are not thought of as as good of businesses as these differentiated, high-margin products. I'm curious for your thoughts on Jensen and NVIDIA specifically.

NVIDIA's position is definitely unnatural. You look at NVIDIA—they've maintained all their margins. Isn't that amazing? It's 2026, and everyone's coming for them, and they're still charging however much money for a chip.

But they're actually not maintaining their margins, because of this whole question of circular financing. People talk about Lucent and things like that, and NVIDIA is providing a 25% backstop. If you actually ascribe a value to NVIDIA taking equity in the neoclouds or whatever, they guarantee they're going to buy all their compute through 2030. Why do they do that? So that the entity in question can get a lower cost of capital, so they can buy more GPUs—

Buy the stock chip.

—et cetera. But implicit in that is: Why do they get a lower cost of capital? They get a lower cost of capital because NVIDIA assumed risk. This is my point before: Risk never disappears. It just appears somewhere else. Taking on risk has a price.

There is a world where AI takes off, it never stops, and everything is fine, and NVIDIA captured all the upside of their risk. But there's also a world where, say, it's this new cloud they backed. A ton of compute comes to market. The hyperscalers have plenty of compute. They don't have enough compute. NVIDIA is paying for compute that no one wants. They just lost a bunch of money.

If you think about it, there's an expected value of that investment. That expected value is not 0. It's not 100%. It's somewhere in the middle, but that is a diminution of NVIDIA's profitability if you actually look at their business holistically. What that is is a price cut.

The price cut didn't show up in margins. It didn't show up in what they're offering, but a lot of what NVIDIA is doing is asking, “How can we maintain our margins even if, in the wide-view, sort of discounted-cash-flow, expected-value, holistic view of our company—” People do discounted cash flows, but are you actually considering all these pieces?

The reality is that moving stuff off the balance sheet, by and large, works, but they're doing all these deals to maintain what feels somewhat unnatural. We have seen price cuts. They're just manifesting in these very bizarre sorts of ways.

In the long run, I think the challenge is that their ultimate competitors are the hyperscalers, particularly Google and Amazon. Google and Amazon aren't just building their own chips; they're also looking to sell those chips externally. Google already made a deal to sell, like, 20% of its TPUs to Anthropic.

On the last earnings call, Andy Jassy practically confirmed that they'll be selling Trainium 3s, or maybe Trainium 4, or those Trainium chips, eventually externally. Which makes sense. That gives them a long-term buy into these companies. There's a huge amount of R&D that goes into developing chips. They get more leverage on their spend. It all makes sense.

By the way, they're not selling their chips on differentiation. They're selling their chips as commodities. NVIDIA's the one selling the differentiation. People aren't going to Amazon to use Trainium, so they're not cannibalizing the attractiveness of their cloud by selling Trainium outside. So they're NVIDIA's biggest problem, because what's the number-one advantage that the hyperscalers have?

Scale.

Lower cost of capital. It's a capital fight. They have a lower cost of capital than the neoclouds do. The neoclouds will buy NVIDIA left, right, and center.

By the way, it also makes total sense why SpaceX Si, like Elon's out there saying, “We will always buy NVIDIA because they're the best.” No, you'll buy NVIDIA because they're the most fungible. NVIDIA is truly the most fungible.

CUDA's moat is dramatically diminished because the models don't care what they run on, and that's what actually matters—what's built on top of the models. But it still matters. It's still something of a moat.

If you want to play the game SpaceX is playing, where we're going to build a lot and rent it out, but reserve the right to pull it back, of course you're going to be on NVIDIA, because the easiest way to rent it out is to be on NVIDIA.

You saw this very early, by the way. You go back to 2024, 2023, and NVIDIA starts talking about all these sovereign clouds. They start talking about—they tried to come up with these Neotron models. They had this thing in 2024, I remember. It was the first one where it was the rock-star GTC in San Jose, in this huge coliseum, and Jensen Huang comes out.

It was a very boring GTC. The old ones used to be NVIDIA demonstrating 50 gazillion things as they threw stuff at the wall.

They knew they had something with GPUs, and they were trying to find the use case.

Get people excited about them, yeah.

Once LM showed up, it was like, “Oh, we have the use case.” But they were coming up with all these enterprise offerings. I can’t remember what they were called, but they were these modules, basically, that, of course, were free, but only ran on NVIDIA. You could see what they were doing: They were trying to lock people in.

Intel is a good example here. AMD cleaned them out in hyperscaler sales because the hyperscalers would put in the effort to get things working on AMD versus Intel. There are still small differences even though they’re x86. Because they’re buying at such scale, the investment to do it is worth it to get a better chip, a lower price, or whatever it might be.

The part of Intel’s business that never floundered was selling to governments and selling to enterprises. They don’t have the resources of a hyperscaler. They’re not buying at that scale. They’re just going to keep buying what they had before. That’s why NVIDIA talks about selling to sovereign clouds. That’s why they talk about selling to enterprises, because they want to get into these markets where they’re not going to be balancing this chip versus that chip.

The hyperscalers have always been the threat to NVIDIA for that reason. They’re actually bigger. So you have this issue where the hyperscalers are the threat. The hyperscalers have a better cost of capital than the other companies Nvidia wants to buy them. That’s how you get this deal this week. I see this deal as a response.

That’s why it goes with the Google deal. Google can just issue equity. The shareholders don’t love it, but their monetization capacity is much higher than NVIDIA’s, or than NVIDIA’s customers’. I think what NVIDIA is hoping for—maybe they wouldn’t say this in so many words—is that if we get to a world where we actually run out of power, that’s probably good for NVIDIA. Because in a world where we’re totally constrained on power—

Everyone will want the best stuff.

—we have to get the best efficiency—

Yeah, yeah.

—the best token efficiency. And I think NVIDIA is still the most token-efficient, so that is a good world for them.

Probably the biggest problem for NVIDIA over the last couple of years is that I think the US has actually brought a lot more power online than expected. It surprised me. Whether it be what Elon did behind the meter, which has been replicated, or West Texas and natural gas, or even restarting nuclear plants—

You love how the US responds to these things.

It’s awesome. It’s actually one of the biggest encouraging signals about the US. I was writing early on, “Assume this is a bubble. You want there to be a long-term payout.” With the dot-com bubble, we got fiber in the ground. And, by the way, Google’s played this game before. Google built its business by buying up dark fiber. They had the killer search engine, but so much of the power behind what they do is because they bought up all this dark fiber that was basically free after the dot-com era.

Our core internet still runs on WorldCom fiber. That was a lasting benefit. The railroads—BNSF is throwing off money that’s going to Google from Northern Pacific and Jay Cooke selling bonds to retail investors. You want a bubble that produces something that lasts.

Very early on, it was like, “What’s going to last from AI?” The GPUs don’t last that long. Data centers, okay, fine, but what is it going to be? Power. It has to be power. If we’re in a world where this all blows up and we have way too much power, that is an amazing world to be in.

We’ve always been energy-constrained. Energy undergirds everything. What would it be like to live in a world of energy abundance? It’s hard to even imagine because our minds are so constrained by the fact that we’ve actually always been in energy scarcity.

I think we’ve done an unbelievable job. Power, for sure, is a constraint. It’s going to be a constraint, but I think it has taken longer to become a constraint than anyone expected, and I wouldn’t be surprised if that includes Jensen Huang. I think he thought insufficient power was going to be NVIDIA’s moat sooner than it happened.

It turns out that the longer we have enough power, the more time Amazon has to make Trainium better, and the more time Google has to make TPUs competitive from an efficiency standpoint. And if we get into a world where that happens, this is a world where those margins seem very hard to sustain.