Foundation Models are a Commodity | Benedict Evans on a16z

Agentic coding went from being kind of useful to really changing everything. It was going to be magic. And in 20 years' time, we'll just say, 'Well, of course that's how it is. Computers have always done that.'

We are in this extreme scarcity. Like, we can't spend $10 trillion a year on AI infrastructure because there isn't $10 trillion a year there to spend on it.

I don't think foundation models are a product. I don't think a chatbot is a product. I think the value will be further up.

There's like three or four building blocks you can put on the table. One of them is...

Thank you.

So I think we have much more of a sense of diverging product strategy. We have much more of a sense of kind of competitive tension that goes beyond just make a bigger model faster with more compute. We've had several iterations of OpenAI's strategy in particular, from sort of everything all at once yesterday to, oops, no, maybe we should double down on coding. Clearly, agentic coding started working, and so all the focus in tech has kind of narrowed in massively onto that as something that has absolute product-market fit in the sense that, like, the customers are pulling it out of your hands. And of course, that comes with the supply crunch around capacity and price imbalance of supply, demand, capacity, capex, pricing that we see at the moment. So that's kind of the big shift. Like, we had a moment of like, this is kind of sort of working and kind of exciting, but we're not quite sure what we're going to do with it. To like, right, it works for coding. Will it work for anything else? Like, yes, almost certainly, but that's what's working right now. And so that's become, we've got this kind of much narrower focus. Otherwise, you know, the chart numbers keep coming up, the models keep getting bigger, the capex keeps growing, the usage keeps growing, people using this more. But most of the sort of fundamental questions you might have had two or three years ago didn't really have answers. Like, we don't know if there'll be a winner in the models. We don't know if they can capture value up the stack. We don't know how much the models can do. We don't see a way that consumers will use this daily rather than weekly with the technology we have right now. So all of those questions are still open.

Well, deterministically, you could have said, well, look who's messing about with this stuff? Software developers. What are software developers going to try and make work? Software development. So, you know, at a very kind of simplistic, naive level, well, yeah, the stuff that should work first is software development. Just as like, kind of, I often compare this moment to like the internet in like '97, '98, but it's also like the PCs in the early '80s or the late '70s. It's incredibly exciting, but it's not quite clear what it's for, and it doesn't quite work yet. And clearly, the first thing that people did with PCs was make computers. And the first thing that people are doing with LLMs, in a sense, LLMs are computers, is to make more compute. And so that's not terribly surprising. I think the shift has been at the beginning of this year clearly that agentic coding went from being kind of useful to really changing everything. And I'm not sure you could have, clearly there were people who were going to say, well, this is going to be able to do absolutely anything. And so they will say, well, yes, look, I told you. But I don't think anyone kind of could have deterministically predicted exactly when that was going to happen and that it was going to be coding that would work first.

I don't think we've learned anything. I mean, you know, this didn't work six months ago. And everyone is scrambling around trying to work out what it means. And, you know, you can get very, very into the noise and the detail and what did somebody say at a party yesterday. So, oh my god, that's how it's all going to work. You know, it's going to take a couple of years for this all to settle down. You know, if nothing else because of the pricing, you know, we've got this enormous crunch between the demand and the supply and hence the pricing. So we don't know what, you know, what a team is going to look like. I think people are asking new questions around, you know, the sort of the obvious one of, you know, do you hire junior people and if so, what are they doing and why were you hiring junior people in the past and were you actually hiring to do the thing that they did or were you hiring them to do something else? And so if you automate away a class of stuff that used to get done by people, then what will happen? And that sort of becomes much more real now in software development because you actually are automating a bunch of stuff that used to be done by people. So those questions are kind of now rather than theoretical. But I don't think anybody can possibly say they kind of know what the market structure is going to look like or what the career of a software engineer is going to be in three years' time. I think it would be, you'd be insane to think that you could know that yet.

Well, you know, it's always been such a tranquil, drama-free environment. So, you know, it's... and you know, obviously they've had the issue with Fidji Simo having to take a medical leave, which kind of shuffled things up a bit. Look, clearly the second half, last quarter of last year, their question was, right, well, the models are the models, but what else? And how do we get people to do other stuff with this? You know, ask ChatGPT for 15 ideas for what we could do to build value on top of infrastructure, and then we'll do all of them. It's almost literally what it looked like. And then Anthropic, with having less capital raised, said, 'No, we're going to focus on coding.' And they got coding working. Whether that was like a deliberate strategy or kind of they stumbled into it is, you know, for other people to say, but like clearly that worked. But the question kind of still remains. It's like the stuff that's working right now is software development and some things in some other fields, and then there's a lot of people who are kind of excited about using this around the edges and using this for some things. And there's clearly this kind of very widespread gap between people in the Valley who bought, you know, a cluster of Mac Studios and are running OpenClaw all day versus, you know, those other 40% of people who say, yeah, it's kind of useful, I used it last week for something. And I'm like, how do you bridge that? And I don't think that question... software is a place where that's really, really bridged over that bridge. And I don't think... and then there's a lot of other places where people are kind of scratching their heads and using it up to a point. And then there's a lot of places where corporations are using it to automate some like specific back office process where you're not asking the user to work out what they do with the new tool. Instead, you're saying, okay, here's a problem that we can solve. And you know, I go and talk to, you know, companies outside America and outside of tech and talk to consultants and, you know, investors. They're looking at those one-at-a-time point solutions. So like I spoke a couple of days ago to a commodities company and they want to use LLMs to get better predictions on their cash flow because they deal with all sorts of small producers and they don't necessarily know when their invoices are going to get paid and it's a very low-margin business. So that's a big deal and so they want to use LLMs to get better cash flow forecasting. That's a very different thing from kind of going to ChatGPT or Claude and saying, hey, you know, give me a summary of my meetings this week.

So I think there's a bunch of different ways to answer this. One of them is like we're always standing on the shoulders of giants and the growth is always compounding. So mobile didn't need to wait for the internet or cellular networks, like mobile data. Mobile internet didn't need to wait for it, kind of needed to wait for cellular data, but it didn't need to wait for like the internet to happen. And the internet didn't need to wait for PCs, and PCs didn't need to wait for consumer electronics and semiconductors and so on. So you've always got this accelerating adoption. And you know, when your boss, my old boss Marc Andreessen, was working on Netscape, there were like double-digit millions of PCs on the entire planet. So like, no, you couldn't have 900 million weekly active users because there weren't 900 million PCs. So there's always that acceleration. So that's one point. I think the second point is like at the early stage of any of these shifts, it's not really clear how it's going to work and nothing works. So you know, like I'm just about old enough to remember this. I'm not sure how old you are, but like, you know, anyone in their 30s doesn't really remember a time when it was completely normal that you'd be working and then everything on the screen would just freeze and you just have to crawl under your desk and unplug the computer and then pray that like some of what you'd done in the last hour might still be there. That just doesn't happen anymore. And go back to, you know, the '80s and like you bought a sound card. Well, that's $300. You won't have sound on your computer. Okay, that's $300 and it's like a, that's the weekend to make that work. I mean, I remember this trying to get this stuff to work. And the same thing with the internet, like, you know, you've got to get a floppy disk that has TCP/IP on and, you know, it's slow and none of the stuff that you need to do existed. And the same with mobile. And we're kind of at that stage. And of course, it's not clear which of these things are going to work. And that's the same thing now, like a browser's going to work, is it going to be this? Is it going to be that? How's this all going to fit together? And there's a gap between what's incredibly exciting and the small number of people who are willing to put the work in to get something to work and just turning that into a thing where you can just press a button in real hands. I think the third point here is, and you know, there's a much more tangible observation, is that the pricing crunch that we've already mentioned looks to me a lot like what happened with mobile data instead of 2009, 2010, where suddenly people got bills for like $5,000, $10,000 of data on the one side. And on the other hand, if you had flat-rate data, which is kind of what happened in the US with the iPhone, the AT&T launch, AT&T Cingular launched the iPhone with flat-rate data. And then unfortunately, everybody buys iPhones and starts, and then they get 3G and people start watching YouTube and the whole network goes down because they just don't have capacity to do that. It's funny, there are still people in tech who don't understand that cellular networks have marginal cost and like they have to add more capacity and that costs more money. And so the networks kind of had to scramble to get like the cost curve aligned with the infrastructure, the pricing system aligned with the underlying cost and aligned with perceived value, which they kind of did with, with, with cap bundles and fair use and throttling and so on. But the other side of that, and that's exactly, you know, where you see now, it's like on the one hand you're paying $20 a month and you get 10 grand worth of tokens, and on the other hand, like, you know, you messed about for a couple of days and you get a bill for 10 grand and you're like, what the hell is this? That's exactly what, you know, you see, literally see these stories now, which is exactly what happened in kind of 2009, 2008, 2009, 2010, also what happened in like 2001 and 2002 and 2003 with GPS. But I think the other interesting part of that analogy or that comparison is that since then mobile data traffic has risen by something like 1,500 to 2,000 times, and the mobile networks collectively have revenue of about a trillion dollars, and they spend about $200 billion a year on capex, and the stocks have been flat for 20 years, and all the cool stuff got built by somebody else, and they kind of all thought that they were going to build all the cool stuff. Like I worked for a phone company that had a banking license because they thought like they would, they would do, they would do mobile banking, which now seems absolutely insane. But that's kind of the point is they built this amazing piece of global, incredibly sophisticated, very expensive global infrastructure with enormous growth in use all the time, and it changed all of our lives, and we all pay for it, and they didn't make any money from it because all the value moved up the stack. And this is of course absolutely, as I said earlier, this is the absolutely central question for, for LLMs is can the model do the whole thing or do you have to have 300 apps built on top of it? Can you just go to the model and say, do my taxes for me, or do you need to have a tax thing that uses, that might use some AI in 10, 10 different ways inside it? And if not, then what is it to be a foundation model provider? Is this just commodity infrastructure that gets sold at marginal cost, which is somehow seems to be a very difficult concept for people to grasp right now because you can sell all the tokens you can make, so you can price it at ROI. But over the next couple of years, we've got like a trillion, two trillion dollars of capex coming down the pipe, and the models get 100x, 200x more efficient every year, and then there's new models, and will the models use more tokens or less tokens? But wherever, like, we'll get to a different equilibrium, and why would that equilibrium be one where the model companies have pricing power when the models are all kind of the same, doing kind of the same thing with the same chips? Why would they have pricing power? And I think that's so, it's a long answer to your question, but you know, you go back and look over time, like chip companies didn't capture the value, ISPs didn't capture the value, mobile network operators didn't capture the value. Windows and iOS did, but they were doing something else. They were, they had all these levers to go up the stack back. And of course, they have network effects which models don't have. So that's sort of the, the question is do they end up like the infrastructure layers or do they end up like the operating system layers and capture value and actually get to decide what gets built or do they end up... I mean, the irony of there, of course, is Netscape where, you know, Marc Andreessen famously said that he was going to turn Windows into a set of badly debugged device drivers, and then Microsoft kind of crowbarred their way into the market, but it turned out that web browsers weren't the point because all the value was somewhere else. And so I think that's kind of the more, the kind of swirling mass of questions about how this settles out, which comes back right through to all, all my kind of answers to your question. It's like, you know, some of this stuff, but you don't know how it's going to work.

Well, so two answers to this. I mean, there's all these sorts of quotes about how history works, and you know, my favorite one is history teaches us nothing except that something will happen. And you know, you can always ex post facto say, well, of course it worked out like that, but it generally wasn't obvious at the time. And you know, in particular, I remember, you know, like sort of 15 years ago, a lot of really, really clever people in tech looked at the iPhone and Android and said, you know, this is open versus closed again, and Android is going to crush the iPhone, which of course isn't what happened. And then I can go and explain why, but you know, all of these, all of these comparisons are useful, none of them are predictive. And you know, it's always obvious in hindsight. I, you know, it's funny, I've done a couple of podcasts recently and I've published this presentation, and there's like a, there's like a class of comment on this stuff which is to say, you know, Benedict, you're not doing your job. You're supposed to tell us what's going to happen. You're supposed to make predictions, and all you seem to do is say, well, we don't know. And there's kind of two problems with that. One of them is there's a bunch of class of places where I actually do say, like, I don't think this is going to work. I think it's going to work like that. Like, I don't think foundation models are a product. I don't think a chatbot is a product. I think the value will be further up. But the other side of this is like when you're at this stage in the cycle, you, there's, there's, there's many paths and you don't know which of those paths it's going to be. And to try and say, well, I think it's going to be that one is, you know, you might be right, but you do have to kind of be conscious of, of like how uncertain this is and how many different paths it could take. That's the nature of this part of the cycle is all bets are open. You know, we get to the point where the S-curve kind of curves up and it narrows in. And, you know, there was a moment when, you know, Windows Phone might have worked. In hindsight, no, it probably wasn't going to work. But, you know, there was a moment when it wasn't clear how mobile was going to work. And there's a moment where it was clear, right, this is what's happening. Now we move on to the next question. One of the characteristics of tech is that the moment that you understand something and you know how it works and what's going to happen is the moment you should move on to something else. You should always be looking for the places where we don't know what the answers are because, you know, I haven't updated my Apple spreadsheet in like five years because we know what happened. I don't care what this year's iPhone looks like. I don't pay attention to their market share in China. Like, it happened. Next question.