What's Still Valuable When Intelligence Is Free

0:00I can see all these incredible ways to use AI and I use a good jillion tokens and I can see how I could use a good jillion more. Everyone uses many tokens as me. We are so short. Like it's going to [music] be this huge thing that's going on forever and everyone and and everyone who's in this build out is going to make a fortune, right? On the other hand, the cost per token is going to collapse and so maybe actually

0:19like we're totally over built, right? And like I can tell I can tell both stories like with [music] equal certainty.

0:39You called me saying it's [music] just intelligence was sort of the theme of our conversation. Describe that like simple idea. There's this idea that if intelligence starts at like an amoeba and it gets better and better and and then it's like a monkey and then it's it's like a person and then it's like Einstein and then it like goes zoom and like every and it becomes God, right? I think that that's something of an intellectual mistake. What people have wrong there is they're missing what are sort of the

0:59real limits of what intelligence can do. And and maybe another way to put it is like if you think thinking of intelligence as like an unbounded scalar like the height of a tree isn't correct. What it you should really think about intelligence as is more like a ratio like the like the roundness of a ball. And what the ratio is calculating is intelligence is is a gadget and we use it to make predictions about the universe. And so we start with this thin scattering of facts about the universe where we have relatively few of them in the world is

1:26in the universe is complex. And we use intelligence to interpolate between them. So as your intelligence goes up, your ability to sort of color in the spaces gets better. And some domains are really amenable to that, right? Mathematics, um you can, you know, solve it like, you know, those sorts of things. And in those areas you

1:44kind of get to perfect. Um but for most of the interesting stuff that is most of the questions in in the world. The mathematical term is they're they're complex dynamic systems. So like um what that basically means is it they're kind of like weather in the sense that you can't just guess your way. You can't make a closed form

2:03answer. You can try to simulate it. Um but if you simulate it, invariably your simulation will have some degree of error in your initial conditions or in each step going forward because you know, at a minimum your steps for example are going to be truncated. And so what happens is your error will build and invariably after the simulation runs for some period of time, you'll be off by some amount. And so the point is that you can't actually like for mixing cream into your coffee or predicting the weather or thinking about how groups of people, you know, who are all very complicated are going to interact with

2:33each other, intelligence just isn't that powerful. And and so my sense is that what what you get is you get this sort of increasing intelligence to the point where you kind of are making all the predictions that something that super smart could make which it turns out is a relatively small set of all the interesting things you'd want to know about the world. To get past that you then need to go out and get more information about the world and that just takes time. You have to like do

2:58experiments. You have to touch reality. But what one interesting corollary to that is that the zoom actually kind of becomes like one of the things you can turn your intelligence to is reducing the cost of intelligence which we're seeing. So what you actually end up getting is like everyone gets unlimited really great like very round ball so to

3:16speak intelligence for essentially free. And and obviously that's a sort of a a limit statement. It won't actually be free but the point is that you know, that you that that what we're seeing is is that like really good enough intelligence is rapidly falling in price and it turns out that really good enough intelligence will solve the vast majority of questions where

3:33people kind of need to use intelligence. Yeah, can you say more about why this notion of like a complex adaptive or dynamic system means that actually even perfect intelligence or whatever intelligence we have in a in a bottle 10 years from now or pick your horizon will still be limited in its applications? Yeah, so like non-linear dynamic system. So the the famous story here and how these were discovered and this this came into the branch of mathematics became chaos theory is um a back in the day there was a um there was a guy who who was building a weather forecasting model, you know, back on big computers that took up rooms

4:07and that sort of thing. And he runs the model and it makes predictions of pressure and temperature over some span of some forecasting period. And he's like, "Cool." And so then, just to make sure it's still working, they reset it and they run it again. And it starts out basically the same results, but then it's the results start to diverge and by sort of the third, fourth, fifth day of the forecast, um the results were entirely different. And the guy thought, you know, some version of like, "Darn, I I must have a bug in my system." Um and begins, you know, the process of trying to figure out why the computer isn't

4:38producing the same results in repeated runs. And it would turned out that the technician who had loaded the data into the computer in the um second run had rounded off some at some decimal, like one decimal less, all of the initial conditions for pressure and temperature and such. And you would have thought that okay, maybe that would have produced a small but consistent error over the whole forecast period. But what had actually happened was that small error through all of the, you know, uh repeated steps, you take your current conditions, you calculate your next set of conditions, they compound and you had an entirely different forecast by some

5:20small distance into the into the forecasting period and the unrecognizable weather. One one one model said it was 70° and sunny and and the other model said it was, you know, 20° and pouring. People called it like the butterfly effect cuz at least in principle, you know, a butterfly could flap its wings in Africa and like that could lead to a hurricane in in North in in North America or whatnot. And it just turns out that like many systems that we're interested in like have this characteristic. You can know everything about pressure and temperature, but you can't model you can't necessarily like write down an equation that you can

5:47solve easily that would tell you that there's going to be a hurricane. You have to like model it. And the model in turn is is is going to be grained at some level unless you actually go to the step of actually having like the whole universe. And so ultimately like in the limit, if you want to simulate the whole universe, what you actually need is the whole universe, right? Now, you can't

6:05actually you don't actually do that. What you actually do is, you know, there are many, many circumstances where simulation is super useful for approximations. Uh and in those circumstances, you know, you can use intelligence to solve them, but there are many more um where it doesn't work. Um and so, intelligence is is limited in that instead of the the sort of fundamental unknowableness, I guess, about the universe. So, back to this notion of it's just intelligence. What does that mean about where you think we're going

6:31from a more practical standpoint? We're all going to have, you know, a team of, you know, 200 IQ people working full-time in our pocket, and that team will have, you know, it itself will have a team of that it can deploy of many more. And they're going to be trying to figure out all the little optimizations that

6:48we're too lazy to do. And every company, you know, will will have will make similar teams available to their employees in order to, you know, to to do their jobs. There's just a lot of businesses that where that's not good for them. I think there's um you know, variety of ways. So, in a lot of companies, I think that what happens is like consumer satisfies, right? They they Consumers are boundedly rational at present, right? They they only have so

7:14much cognitive space. And so, you sort of go to a consumer and you say, "Look, we'll take your money. We won't steal it. We'll pay you a decent interest rate. Like, you don't have to think about it. Like, and you can go about your life." And the consumer doesn't say, "Look, if I think about this every day, I could save, you know, like, $400 a year. So, therefore, I'm going to have to think about this every day." The consumer says something

7:34like, "Eh, sounds good enough." Right? I'm going to move on. But with an agent, like, you would think the agent will will will continuously optimize for them. And there's also companies where what they produce fundamentally is like some sort of artifact that is just like some sort of crystallized intelligence. So, I'm by far not the first person to realize that AI could be bad for certain software companies, but like, you know, just this idea that like what what an AI model is is is is it's kind of like this general-purpose intelligence tool. And what many software sort of tools are are kind of like crystallized and hardened

8:01like pieces, like, projections of intelligence. And so, you know, as I've been thinking about companies, you know, and and investments, I found it helpful to sort of you know, assume everyone has really great intelligence and ask what's left. And so, the line I've been using when people talk about some version of the pushback is often some version of but you don't understand how hard this is, you don't understand how complicated this is, and you don't understand, you know, how much how many smart people have had to think for so long about this. And I sort of my my my sort of hand-wavy and and sort of approximately

8:30right response is, you know, okay, sure, but that's all just intelligence, right? Um and to follow this logic all the way through to the end is if intelligence becomes cheap and ubiquitous, what you want to own are the complements to intelligence. And so, the complements to intelligence are um the they're they're data, attention, like love, affection, you know, whatever sort of emotional attachment you want. Um it's it's physical proximity, it's scale, you know, it's a little cynical, but it's it's it's regulatory capture, it you know, it's these sorts of things where for some businesses, what's what happens is their finite ability to deploy intelligence, be it economics, you know,

9:08just the cost of it, they're only a certain size company and they they don't have access to 10,000 of Google's best engineers for 10 years, or be it just the organization is behind or whatever. Like the ability to deploy way more intelligence to your business, and your competitors to deploy way more intelligence to their business, and your customers to use lots of intelligence, it can work out to be, you know, better for you, or you know, if intelligence is a barrier to you serving your customers, like the lack of it, or it can be um bad for you if intelligence if your customer's lethargy is kind of a

9:37fundamental element of your of your economic um value. Maybe pick like Meta or something as like the perfect example of a company that I think you believe owns perfect complements to Sure. ubiquitous cheap intelligence, just to like bring the example to life. I'm actually going to set aside for a moment like muse and agents and that

9:55whole thing and focus on business. Uh and and the core business at Meta, right, is is is personalized TV. Personalized TV. Yeah, it used to be social media, but now it's like it's personalized TV. And it turns out that, you know, people really like uh watching Instagram or whatnot. It just discovered that I love like dark British humor and it's great. If you take the a really high like like you people talk about like the 10,000-ft or whatever. Let's talk about like Meta from space, right? Like the

10:20the highest view. They have 250 billion in revenue. And they spend 250 billion dollars a year. Some of it gets tax accounted as for its costs, some of it is accounted for its capex. It it doesn't really matter. Just they spend it all. The question you might ask is like what do they spend it on? And a really reductionist view is that

10:41they spend it all on intelligence. And what do I mean by that? Well, they're buying data centers. What is a data center? Well, it's kind of an intelligent producing tool. What else are they doing? Well, they're they're actually literally buying inference, right? That's just intelligence. What else are they doing? Well, they're paying really smart people to write algorithms. That's intelligence. They're

10:57writing code, also intelligence, right? Like they're surveying videos. That's intelligence-adjacent, sure. It's intelligence. And so in some highly reductionist sense and you shouldn't you shouldn't take these things literally, but I'm just kind of trying to highlight the really big directions here. All they buy is intelligence. And if you think forward with the fact that intelligence the price of intelligence is falling by, you know, like a factor of a hundred or something

11:19over the next five years. First principles, their costs kind of go away. And you know, that's just thinking about the sort of the cost side of things. Um, on the revenue side of things, there are all of the creators. You are a creator. How much more quality um, creative can every creator that makes content for Meta make if they have access to intelligence that's 1/100 the cost and 10,000 times more effective or whatever, right? The point is

11:52probably better. And then Meta in turn, maybe they don't take their costs to zero. Maybe they spend more or more than zero. And they do a far better job of picking the videos to show people next, which should make the content more, um, you know, better targeted and more enjoyable. Um, and then of course on the ad side, um, you know, advertisers could make far better designed, far better targeted, personalized ads that then Meta in turn can target much better. And advertisers can also build kind of much more

12:27slippery funnels or whatever to convert. The whole point of all this is that maybe instead of taking their cost to zero, if with much better intelligence and much lower prices, maybe they can like sell a few more ads. All of this is downstream of of what they possess, which is not, um, destroyed by ubiquitous intelligence, which is data and attention. They they just have this enormous amounts of data about, you know, all all these people and how they view they view videos and whatnot. And they have an enormous amount of attention and these things compound each other. And and so I don't think that the availability of cheap and

12:59ubiquitous intelligence is going to cause people not to be interested in watching funny their 20-second videos. Um, and so if anything, you know, they can win some time on the margin, you know, um, from other things people do with their time. I don't actually own it, but like, you know, a company that benefits in theory from having a lot more intelligence is is Grindr, right? Um, and you know, basically right, Grindr is the gay dating app, but it has been constrained over its life by the fact that it's just not that big a

13:26company. And it turns out they've got lots of attention and they've got lots of data, um, but their ability to build all the things that you could imagine building on what ultimately amounts to sort of a finite size, you know, team and budget and ability to like create things. Um, you know, it has has meant in some ways that Grindr can't be all the things you could imagine it could be. And again, I don't I don't actually own any Grindr or whatever, but we study I study lots of companies and let's just say it occurred to me that like this is another example of something where, um, you know, here

13:56the here the sort of mental model is like what what if you had 10,000 of of Google's best engineers to go work on your problem for 5 years, like what could you do? And Grindr is one where you, you know, maybe they could do more. Ramp is the only platform built to make your finance team leaner, faster, and better, saving businesses 5% annually on average so So can stay focused on growth. Ramp just opened for business in the United Kingdom. So if you're running a business in the UK, you can now use Ramp's AI-powered finance platform to manage cards, expenses, bills, approvals, and accounting all in one

14:26place. I run my business on Ramp and so should you. Learn more at ramp.com/invest. OpenAI, Cursor, Anthropic, Perplexity, and Vercel all have something in common. They all use WorkOS. And here's why. To achieve enterprise adoption at scale, you have to deliver on core capabilities like SSO, SCIM, RBAC, and audit logs. That's where WorkOS comes in. Instead of spending months building these mission-critical capabilities yourself, you can just use WorkOS APIs to gain all of them on day zero. That's why so many of the top AI teams you hear about

15:02already run on WorkOS. WorkOS is the fastest way [music] to become enterprise-ready and stay focused on what matters most, your product. Visit workos.com [music] to get started. Rogo is the AI platform purpose-built for financial institutions serving tens of thousands of bankers and investors at hundreds of leading firms worldwide. Rogo's AI agents autonomously execute large chunk of your firm's [music] workflows. They can screen deals, draft SIMs, run buyer outreach, and due diligence on data rooms. All with full

15:31security and compliance handled. Every action feeds Rogo intelligence, the firm's context layer, turning years of deal experience into a secure, governed institutional memory [music] that makes each new transaction stronger than the last. To learn more, visit rogo.com/invest. I want to talk about the labs. So back to the original point, I think lots of great investors would say um intelligence, specifically frontier

15:56intelligence, is not a commodity. Historically, the conversation has been about the gap in intelligence between closed-source and open-source. Increasingly, the conversation is about the gap between what's publicly available from the labs and what's privately used by them in inside their own businesses. They seem to have figured out how to stop or slow distillation. Lots of people would say that there's this escape velocity thing happening and that the frontier token is going to be way more valuable and useful and that the labs themselves can then distill their own models and provide the best version of the cheap intelligence, too. I'm I'm curious for you to riff on the labs as

16:30companies through this sort of mind of an investor. If they have models that are better and the ability to produce sort of adequate intelligence at way lower cost than everybody else and nobody can copy it, then it's going to be a great It's going to be a great business. I guess I have a couple of thoughts. You know, one is as I've watched models get deployed, it feels like one way to think about it is like you start out with these very inadequate models that are basically

16:53able to do a very small set of jobs. Then what happens is the models get better and they work their way kind of up the stack. And it with each successive step up, you know, the frontier model has unlocked a larger, you know, call it like addressable market. But behind it, what's happening is that once the models get saturate once the sort of whatever get saturated, you get this like you know, you would never use a frontier model for like a chatbot in a in a customer service application. You'd be an idiot. And so what's happening is um those like areas, they they clearly commoditize. And so there's like this

17:26wave of hard tasks where you really need a frontier model and then behind that you you don't. And so, you know, the friends I have who who are big, you know, frontier labs companies bulls or whatever and obviously these companies' financials aren't public or whatever and they might have an information advantage over me, but would say something like, "Don't worry, we can also along the whole efficient frontier produce cheaper tokens." Maybe they can right now, maybe they can't right now, maybe they will in a few years, maybe they won't. It's very I I find this this to be very hard to to think about. Certainly the trend has

17:56been rapid commoditization of of these things. If you would look at the rise of Chinese models and the sort of the the step every step. Uh and at a minimum, the cost has been collapsing. So, um even if they do sort of have a position where the tokens necessary to do tasks that are sort of well behind the frontier, it might not be a very big market um because it just doesn't cost that much in the scheme of things to deliver and it is competitive even if it's um even if they are a cost leader in a commodity in a shrinking commodity

18:27market. And then the other thing is that I think that you know programming was this huge use of models. As you push forward the question becomes like for the next next next like layers of intelligence do you still have bigger and bigger places to go after and the bull story would be some version of like yes like people will pay anything for a model that can like figure out how to cure cancer. Maybe that's true but I think figuring out to go to our prior conversation I think figuring out how to cure cancer requires touching reality and I'm not sure that more IQ gets you that. If you don't believe that there's

18:58some super incredible use cases where you can just charge an enormous amount for the next for like more intelligence then you start getting to this point where like the next step up in terms of like like you got coding and then maybe it's like I don't know something else and then but it gets into smaller niches. It's like this diamond shaped demand frontier where like initially you've got wider and wider uses and then actually it starts to get more and more rarified where you actually like how often do you really need a 300 IQ one might start to ask and you know look these things are

19:25all going to like we're going to find out and I don't want to be too dogmatic but it it strikes me as like totally reasonable that what we end up with is some version of like there are very few use cases for like super elite intelligence. There are lots of use cases for like very good intelligence. That very good intelligence is highly competitive and commoditized and if these frontier labs can somehow become like cost leaders in doing that there's a margin to make in being like a cost leader but then there's also just rampant deflation and then this question becomes of like like you know how big is the market if

19:59you know good enough intelligence is very very cheap then how much good enough intelligence are people going to buy? I'll just keep playing the devil's advocate. Yeah. So the two things people might say in this case are Jevons paradox right that like as it gets cheaper we're just going to want to use more of it so I'm curious what you think about that. The second would be you know historically the new uses of intelligence have typically been discovered by highly intelligent people. My my good friend says intelligence creates its own demand. What about that? Like what about the fact that maybe it's diamond-shaped for humans if humans are driving the new

20:32questions that the intelligence is set after to answer. But what if the thing itself is creating the new questions and we train it to do that really effectively and just sort of like explore in an open-ended way in the way that like all of humanity would be doing led by the people with, you know, 150

20:49plus IQ or something historically. Jevons paradox, right? It's just a sort of narrow example of of price elasticity of demand. And in in general, all markets are U-shape, right? Like so initially when the product is terrible, um there's no demand. And then the product gets better and people will pay a lot for but but but it's uh but they won't buy very

21:07much. And then as the price goes down, the the size of the market will will will rise because, you know, you know, it becomes um uh you you Jevons paradox essentially. But then you get to a point where further price decreases don't actually drive enough demand and the market shrinks and eventually, of course, if you take the price to zero, um then demand goes to the the size of the market goes to zero. You know, one example that I've I've thought about is um the cost of fiber optic communications, right? So or just communications. Like the cost of moving information used to be super high

21:44and people used moved very little of it. And and then, you know, through a whole series of innovations, the cost of um uh of moving information uh went down a lot and the amount of information moved uh went up a ton and the market grew a lot for a long time and and and then uh the market stopped growing. And the cost of moving information kept falling and the market kept shrinking and nowadays, you know, um Cogent Communications I think routes 30 or 40% of all IP traffic across the internet and it's through one of its divisions and Cogent, you know, and Cogent which which is like not even its

22:20biggest division Cogent Communications is like a few billion dollar market cap company. This is like just not a very big industry. Now, I'm not saying that we're going to get quite that small in terms of like intelligence, but there you know, Jevons paradox is um is a statement about a particular spot on a curve, which doesn't necessarily have to be true. And you know, you know, it remains as your other question about like can and can more intelligence create more sort of desire for intelligence? I mean,

22:44I have a couple of thoughts on that. Like what may like certainly maybe and I um but I started out by saying like you need to touch reality to like answer questions, right? And so, if you wanted to like if you had a model and it was going to think more about physics, it could certainly think lots of interesting thoughts about physics, but eventually it's going to need to do

23:02experiments. And that's not to say we couldn't also do experiments, but now we're bound by like our ability to real world stuff. physically do experiments. And so, I think there's going to be this incredible time where we sort of there's going to be so many cool discoveries, which you're already starting to read about where like this paper was written and this paper was written and this paper was written and if somebody had read all three papers and thought about it, they would have realized this, but nobody had. But it turns out the models read all the papers and they'll they'll they'll like figure that out and there's

23:28going to be we're going to there's going to be all kinds of great stuff that falls out of that. But at some point, we're kind of kind of finished that and and then it's going to be about like, "Hey, you know, there's some way to think about things that are outside of what we know and we have a hypothesis, but a model, even a super smart model, won't be

23:45better at guessing that necessarily. We'll we'll be better at guessing that than the best smartest person, but the smartest person's odds are super low anyway. So, it'll be able to come up with great ideas for experiments to do and they'll be worth doing, but that'll limit our ability to expand. The other thing I just think about is

24:01like my own use of like AI. Um a lot on a lot of levels. Like one, just to go back to the commoditization thing. I've totally commoditized the my use of models. For what it's worth, I just it's cuz I'm lazy about how I do a lot of my like building for stuff. I'll I'll tell like a GrokBot to build things and cuz I don't want to use up all my GrokBot wheat tokens, I then have it run all of them I subscribe to all the things and I have it use all this and and I'll have it build using those

24:25subscriptions on like a server. And so I actually don't know what model I'm using anymore. Like I used to know all the which this model is better, this one is worse. I no longer know. All of this is being optimized on the back end for me and all I know is that stuff is getting built and like it's not using up my subscriptions. And so like we're good. There's this batch of companies that I'm curious what you think about which is effectively real world experimentation companies that create a loop that's automatic. There's this happening in robotics, there's this happening in several places in bio in in

24:56bio labs and you know automated wet labs, automated testing labs. Um any any commentary on like that? Super great idea, right? Like brand of company. Yeah, like super great idea. Closing the loop in the physical world is going to be great and it's going to accelerate the process of learning and and and this is a great idea. But again, like there is there are speed limits. I I guess I'm my my sort of the the thing I keep coming back to is some version of like this is amazing and I'm super in a sense like AI pill in the sense of like what I think these are things are going

25:23to do and they're useful for. But I'm but I'm also sort of cognizant of the idea that like things bump up against The way I like to tell this to people is I say look, you're really smart so you must understand. Being smart has limits. Being smart has limits and and so the ability to just think your way to solutions is is limited. And so if you're going to build an an automatic biology testing laboratory and you're going to connect it to a computer and the computer is going to come up with an idea and and then it's going to run an experiment and it's going to iterate. That's awesome. But if

25:50that experiment requires breeding mice and mice need a year to grow up and die, that experiment going to take 10. And it might be a great experiment, you might learn a lot and that's great. Still going to take time. And you know, the lab's only so big and so you might have 5,000 great ideas,

26:05you're going to pick 20. And you you the model may be better. You you your odds your odds of that experiment working might even be 1%. Turns out that might be 10 times better than the average researcher and that's amazing. It's still 1%. None of this is to be pessimistic. This is like super amazing. Like discovering all the things that we should know that cuz it's already in our corpus that we forgot is awesome. Discovering all the things um like and like better directing our research is awesome. Automating that

26:32research to move faster is awesome. We will learn. It will push the frontier. But the But like this is not um This is not like suddenly the model like looks looks around and says, "You guys missed it. Like here you know, here's like you know, some new world that like you guys can't comprehend or whatnot." Like it I don't I don't think that's where we're we're headed. It seems like the market has been especially in private markets and sort of the leading edge of investors thinking about AI very focused on investing in supply related stuff. Like how are we going to build and serve all the intelligence versus like what I'll

27:06call demand related stuff. Someone watching their personalized TV on their on their meta app or whatever. What do you think about that first batch of companies as investments? Super hard. Right? Like I mean, I've sort of been I've joked I've been willfully ignorant about them and probably I'm sort of worse off for it. But like

27:23you know, there's this big buildout. And then you're trying to forecast like inference demand and I've I can't forecast supply or demand within several orders of magnitude, right? Like this idea of like how much more inference will people use as intelligence gets better and cheaper is like a really really hard thing. And equally hard is how much better and cheaper inference is going to get. And then even if you could somehow guess at those things, then there's the whole value chain and you sort of need to be at the bottleneck in order to capture the rents. And then of course finally you get to like your actual competition

27:56with your competitors in that particular piece of value chain. And like it's just a super hard thing. I think about my use of AI as a consumer, you know, increasingly I'm doing stuff like instead of um like having the AI chatbot like talk to my daughter like you know, or you know, text with my daughter about how to you know, do math. Now I have it make like a three blue one brown style video explainer so that you know, she can see like a nice five-minute or 10-minute video on like you know, powers or area or you know, circumference or something. It's super cool. I don't think I would do that if

28:31it was going to be $5, right? And so there's this thing where like um there's real price elasticity. The point I'm trying to make is like if a lot of demand comes from the idea that like you know, it's one thing to like I'll film myself working out and with my Muse app and it'll like show me pointers on my form and I'll do that for very little money, but I won't do that if the price point is real and I'm probably much less sensitive to the price than like most people. So I think there's like real like like AI demand is going to be really

28:59sensitive to the to the price. Which intrinsically I think does limit in some sense like the the amount in total consumers are going to be willing to spend on this. Maybe this is a time to talk about capital cycles or something. It sure does seem like the smartest people in the world and many or most of the best entrepreneurs in the world are heavily personally invested in this supply build-out thing. You know, if you watch SpaceX's uh investor day yesterday, like this is

29:25basically the theme of it. Arguably maybe the greatest entrepreneur of all time is dedicated to effectively doing this, building out the data center supply, etc. This is just like a historical question like um if that's all true and there's so much uncertainty about supply and demand and there are other ways to go build businesses, products, make money why do you think so many of like the best entrepreneurs, the smartest people are so completely dedicating their lives

29:51to the build-out of the supply? This is a huge infrastructure build-out of like a huge and important technology. And it would be impossible for there to be a huge infrastructure build-out of a huge and important technology if there weren't a strong price signal and a ton of talent going at it. So you may have your causality backwards. You know, obviously these thing things work both ways, but like could you imagine a telling of the story where nobody interesting who'd made no

30:21money was building out the AI, right? That's probably probably the best answer. Um And and to be clear like this is a huge and important project. It's only happening because super smart and talented people are focusing their lives on it. It's great that they are. They will have and deserve to make lots of money. It's just very, very hard and from where I sit and there are probably people who are smarter and figured it out, but like for where I from where I sit, it is very, very hard to figure out where one would invest along this and like how like whether or not you'll you'll make

30:54any money doing it. It's not to say that there isn't like plenty of opportunity potential there. I've just sort of decided to focus my attention more on the longer-term equilibrium results that come from what the world goes to. You know, instead of betting on like which railroad tycoon is going to make money and how many railroads we need, I'm trying to think about, okay, once you have railroads, like what, you know, that like I I I'd love to have the foresight to figure out like that like CPG brands come next, right? Like that's kind of the um the way I've been trying

31:23to think about it. Can you riff a little bit about what you've learned about capital cycles? Like it it and innovation cycles like this. I'd love your perspective on having studied so much market history. If we're in the middle of one of these, maybe the biggest one ever, what does history teach us about these cycles and how they work

31:40that's interesting to you? You could think about American society right now as like a household that's like saving money for a house, right? Like we're building a house and we're all kind of feeling pinched, right? Because all these resources are dedicated to like building out like this big AI infrastructure. All these electricians and, you know, plumbers and, you know, everything else are going to do these tasks and as consequence there's higher costs elsewhere in the

32:00economy. Um and undoubtedly we're going to get a lot of value for it. And we're going to keep pushing it because no one really knows. We're going to keep pushing it until eventually like we overdo it. And along the way you're going to make people have to make tons of money because otherwise people

32:18won't keep pushing it. And because things have lead times and because it's sort of a winner's curse, you know, invariably the last person to stop will regret it. And it's just so hard, right? It's just so hard. Um and then you You that with the fact that there's just so much dynamism in the technology here, which I think was less true for sort of other historical like infrastructure build-outs, right? But like um you know, there's just so much change in

32:46in terms of how you do these things. You know, I mean other people have made the point which is I don't think is is necessarily wrong that like even going, you know, we talked about the models, but even at like the chips level or whatever that, you know, the potential for um AI to commoditize even things like Nvidia or sort of recursive self-improvement everywhere like at the Yeah, like everywhere. And a lot of it

33:07ends up being like just intelligence. Like Nvidia literally quite literally like makes flat files that they send to Taiwan, right? So like Now, their flat files are very hard to make. Um but like what they have is a lot of people who think really hard about it, right? And so in theory you could make them. And also then the question becomes like at what point are you Like what if you step back a little from the frontier? Like then the cost of

33:29making something comes way down. In that world, would your model be that physical physical and scale are sort of the things that you'd want in a business that you were to back on the supply side of this equation? TSMC? I don't want to carry myself around as some sort of TSMC expert. I mean certainly what they do has better

33:48attributes and I think in that regard. They've They've also got this sort of interesting problem where they did they did this super hard thing that requires like all this process power built over many years and is super hard to replicate. Um and much less amenable to like an AI because just because it's it is so physical and because it it it's all these little things that are many which aren't like even legible to an AI, right? It's like it's like some guy in who kind of is in the system kind of

34:11knows that you have to do this, right? Because there's so much demand and because they haven't been able to meet it, they've basically caused everyone in the world to seek to copy their moat. It's like a not a great fact. Um but I you know, I think more generally sort of stepping outside the AI build-out like if you think about um you know, businesses We talked about the meta example. Um just the idea of businesses that have like physical proximity or physical, you know, um scale might might benefit if they can use intelligence to serve their customers better or have like lower costs as as a consequence. I want to go

34:43back through the natural complements to intelligence and just have you explain why why it's a complement to intelligence so that people can go away and think about okay, what companies have this feature that might be interesting. Good example to go talk about physical proximity. How would you think about DoorDash or Uber or these businesses that have local density

35:00effects that matter a lot. Does intelligence being ubiquitous sort of naturally magnify all of the scale winners of this type where physical proximity is some is some sort of key advantage? Well, so if you're the low-cost producer of something, I'll give you a a sort of stylized example. So like at one If everyone has an AI agent, then what happens is many decisions get offloaded to the AI agent cuz the AI agent helps you like optimize, helps you pick up all the like pennies and dollars and nickels and stuff that you like were leaving on

35:30the ground. And so let's take the example of So So I think about it consumer choices in like a spectrum, right? Like on the one hand is like maximizing my bank rate, being deposits. Like This is something where I think I can almost entirely like defer to the AI's judgment. Like as long as it's FDIC insured,

35:47I just want the highest rate. On the other hand, spectrum end of the spectrum might be like picking my favorite sweater. This is something where it's just all of my judgment. The AI can source me things to look at, but in the end, you know, I need to decide. It's going to I can't really delegate that judgment to it. And there's a spectrum. And so let's just pretend for a moment that garbage collection is something where I can defer the judgment. You might say, "No, I I really care about the guy and I trust the service and they're driving to my house or whatever." But let's just set all

36:12that aside for a second. So right now, if I'm in the waste collection business, um I might be great at it, but it's still really hard to get people to bother to switch. They sort of have like I doubt you know what you pay for garbage collection like off the top of your head. Um many people do, but like I doubt you do. And so what happens is you have to do all this work to get people like pay attention to bother to like switch garbage collection companies even if you're offering a better service at a

36:38lower price. You could imagine that like my agent, their agent could talk to my agent and be like, "Hey, I'll save you a 50 cents, you know, a month." And the agent will say like, "Will the garbage go away?" And I'm like, "Yes, the garbage will go away." And like, no problem, and I'll change. So, the first-order effect is

36:52that you move to the low-cost provider. But the general market effects are actually sort of interesting, too, because now let's say that there's a street. And before, like, it would be more efficient for society if all of the people on my street had the same waste garbage collector, cuz the guy could drive up the street and pick up the trash, and he could offer all of us a lower price make and make and make more more money. But it's really hard for us to coordinate that. Because what happens is like all of us are just lazy about switching garbage collectors, and so we all kind of have one, and um you know,

37:19the guy with with two guys on the street can't call everyone else up and say, "Hey, I've got two guys on the street. Everyone else has just one. I'm I'm the low-cost producer. Let me roll this up." So, a second-order effect might be that, you know, the guy get who has two two people on my street can go to everyone else on the street and say, "Hey, I'll offer you 50 cents, you know, a a lower price on the garbage." And everyone will switch. And so now you'd get this like like local, you know, like everyone moves to the low-cost producer, who in turn gets to be like a lower-cost

37:44producer. So, um the that actually creates in some sense like a better like business. Um where the where the low-cost guy wins a bunch more volume at like much lower margins. Um so like, you know, a lot of companies have like, you know, have like mixed As you study them, like a lot of companies have like mixed effects that point in different directions. Um uh you know, because because of these

38:08things. What do you think the impact of agents is going to be on your world? What will be your role? How will it be different in 5 years? Like, will the best agent be able to build a portfolio of Carvana and Hilton Grand Vacations and Meta and in the same way that you would, cuz you

38:22trained it how you think? better. In terms of what I do, right? Like, I think that in the long term we'll probably get competed out. Pricing pricing risk as the thing you do? Yeah. When computer-aided chess came along, there was this long span of time where freestyle chess was better, where where the human working with the computer program was able to outcompete either the human or the computer by

38:44itself. Um and this is just because humans thought about things differently and computers thought about things differently and they were synergistic. And I think in investing for for a while We're in that era now. We're in that era. And that's because, you know, there are things that I understand in the world in ways that the

38:59model just doesn't. Um, by virtue of being a person and having lived a life. And there's information I have about the world that just isn't in um, models cuz as much as as much information is on the internet, not everything is actually on the internet. While that's true, which is going to I think be a while, um, you know, the ability of me to use models to um, assist me and then make judgments will

39:24make better judgments working together. Um, you could you could imagine a a circumstance though um, in the future where that gap diminishes to the point where it doesn't matter anymore. And then, you know, models, you know, I I don't think models will be able to predict the future as we talked about like the future is in some ways very

39:42intrinsically unpredictable. Um, but the model's distribution of outcomes may be um, the best that you can do and that adding a human to the forecast would be kind of like adding my judgment to whether or not it's going to rain in 7 days. You know, Cliff, I don't really care what you think. I'm I'm using the forecast. Like this is

39:59this is this is going to be better. Whatever However you fudge it, it's just going to make it worse. That day may come. It's a very interesting like moment there. Like there's a lot to investing, right? You think about like judging management or something, right? But like maybe maybe the models will be able to judge management. They'll be able to interview

40:15them. They'll be able to figure it out. And then and then it gets really interesting like how do you raise capital? You'll be like, you know, I I'd like to run an amusement park and the model will say like, "Sure. Here are like six locations that have a 51% chance of an amusement park work park working." And, you know, as long as you're like really committed to it. And I've assessed you and you have like a 51.2% chance of like succeeding cuz you're actually slightly below average, right? And you'll be like, "Okay, well, what can I do better than average?" And I'll say, "Well, you know what? Have you

40:38thought about like running a you know, a go-karting park?" But um, you know, unfortunately it I don't think it's going to affect uh, my career that much cuz I think it'll take a while. What about replacing price as the organizing principle and signal that coordinates economic activity? Price is a miracle, right? Cuz it's a signal

40:56wrapped in an incentive. What does that mean? So, price is an is an incentive. It causes you to want to buy more or less of something because it goes up or down. Um and it's also a signal cuz it says something about the relative value of something and the scarcity of something and the cost to produce something. It's like a miracle of a mechanism to coordinate things. Um and uh I think you'd be really hard-pressed to [music] find a better system even if even with like unlimited uh unlimited intelligence. Um it'd be

41:21pretty hard. Vanta automates compliance so your team can spend less [music] time on security reviews and more time getting customers. Vanta cuts audit prep by 82% and gives you instant up-to-date proof of trust without all the manual work so you can close deals faster and with less friction. Customers report a 526% [music] return on investment and more than 16,000 companies use Vanta including Ramp, Harvey, [music] and Snowflake. Get started with $1,000 off

41:49at vanta.com/invest. [music] Ridgeline is the only AI-native investment management platform that brings the front, middle, and back office together in one system. [music] It gives firms one place for their data and uses AI agents to automate repetitive work like onboarding, reconciliation, post-trade compliance, and reporting [music] so teams can spend more time on investing, serving clients, and growing the business. Ridgeline is built to give you time back that [music] can change the trajectory of your firm and be the inflection point that takes your growth from incremental to accelerated. Rideline also wants to help asset and wealth managers understand and improve their firm's growth [music] trajectory so they built

42:23an interactive dashboard that lets you see how your firm compares to 16,000 peers. If you file a Form ADV, your firm is already in it. Visit ridgeline.ai/trajectory to see where your firm stands. [music] Then talk to Rideline about how you can do more with the team you already have. Maybe another way of asking the question [music] is kind of like investing if ubiquitous intelligence gets so much better and so much cheaper and in some way um it can replicate what you do and you're really good at pricing risk. A friend of mine, Dan, was saying there's sort of three levels to this. There's I put in a

42:56prompt and I get something back. That's level one. Level two is I set up a watchtower. So, I set up a you know, my instinct every night runs like a set of tasks and it happens in the background on a schedule. And level three, which produces the most use of intelligence, is it's just always on proactive, figuring out its own way to use its own intelligence, and that just like goes on forever and it's sort of unlimited in

43:18how much it can do. If I think about that level three applying to like everything, like what things should be priced, efficiencies, optimization, all this kind of stuff. Does that add something interesting to the price? Sharpens markets, right? There's a lot of like stickiness to prices, right? It's a well-known macroeconomic phenomenon. It's like price stickiness. There's a lot of like lazily priced stuff. Think about deposits and such. I mean, it's a really competitive world out there, but it's actually not as competitive as in theory like it could be. So, I think you know, if everyone has agents who are working on their behalf and constantly sort of

43:52optimizing, I I think you know, the price of garbage collection will move around a lot more. As all the input costs of garbage collection evolve, and every you know, a garbage collecting company is constantly like re-running its financials and trying to figure out its true marginal cost and bid every customer where it could like serve for like you know, more than marginal cost or you know, and every agent is

44:13switching to the lowest cost provider. That's just a much like the price of garbage you won't just pay $60 a month for 12 months and then have the guy come back to you and say it's a 5% increase, it's going to be 63 next year, and you don't even notice and you just like you know, like it's they're just picking up your garbage. Your your garbage collection will vary by day and and you know, and one week it'll be whatever time it there'll be all this interesting much more accurate, much more rapidly moving

44:38prices, I would imagine. Have you seen anything interesting in capital markets innovation? There there was this awesome economist story you're talking about the behemoth that is American capital markets and how good it is at adapting to new stuff and creating new new instruments, new ways of financing things. And that that's been [clears throat] as much a part of the American story as the technology innovations that typically happen to cause, you know, like that there's this amazing interrelationship between Wall Street and Silicon Valley would be one way of thinking about it. Anything interesting percolating for you in in markets, in capital markets, in ways

45:13of financing things, things like this? Well, I mean, I just it's fun to take a moment to pause and maybe congratulate ourselves as as a as capital markets for what we're doing with AI. Like this I mean, how you know, people were worried about this, the decline in American dynamism. I think we just like redeployed what is it? 3% of GDP, right? Like right into this like highly speculative venture basically on the on the turn of a dime. Um it's just this remarkable thing. Of course, who knows? Maybe it should be two, maybe it should be six, right? But it is amazing that we can redeploy

45:43resources at this scale um this quickly. And there's no one else in the world that's been able to do it. And um you know, certainly the the inability to move investment and productive resources um can be a binding constraint on productivity growth even with especially with rapid technological progress. And and so um you know, at least with regards to capital and obviously you know, we talk about capital, but the capital investments are tied to people doing different things. Like Americans are doing different things today. Like every day, like people are waking up and going to some place in I don't know, Pennsylvania and like building a data center, right? And like

46:31that wouldn't be happening if it wasn't for this whole infrastructure we have that's somehow been able to move this enormous amount of capital. It's incredible. As like deep innovation, I don't I think the fundamental idea that like people are taking risk on speculative ventures and sort of slicing up the pie in different ways is sort of irreducible. I think people will come up with lots of different names for what are essentially the same things over time. I don't think that's like deeply different. Um but it is incredible. I should also note that like I'm sure we'll discover all kinds of like fraud and bad behavior and stuff

46:59eventually like when when like the water goes out, but I think given the magnitude of this, I suspect like it's so clean. Like it's amazing. Like if you read about like the railroad build-out or something, like the amount of you know, like waste and and fraud and like whatever. Invariably there is some, but but like the like this is like there's a lot of people trying to basically do the right thing here and like the governance parts that seem to kind of we just take for granted. Like it's incredible. Like America's capital markets and the capital markets in general I think are

47:29just like incredibly under appreciated. Like the idea that you know, you know, the any person can sort of buy $100,000 of stock and get, you know, a portfolio of well-run, you know, well-motivated teams, you know, sort of building the companies that are they're sort of basically doing mostly the right things and do it basically instantly and at basically no cost and then sell it as soon as they're done. Like you know, is amazing and then the fact that that that all those companies and and other ways people invest like can then allocate society's resources like pretty well like in the scheme of things is just I mean,

48:02you know, just incredible. Like the whole system is just very underrated. Why has America so dominated this, do you think? My pet theory and everyone has their own theories, um, it is some sort of interplay between genetics, cultural evolution, and, um, institutional evolution. So basically like the people who came to the United States were the subset of people who were most interested in taking risk because taking going to the United

48:32States was very risky. And I think that um, risk taking has a meaningful like genetic component. And and so I think that that tendency is just more common amongst, you know, Americans. When you're in a place where um, risk taking is more the norm, there's also a whole bunch of like other things that come from that. Like um, it it's it's more the norm and then that leads to build-ups of sort of cultural and institutional things that reinforce that. It's understood that people can take risk and fail. That leads to bankruptcy laws that make sense. That

49:07leads to labor laws that support that. That leads to you know, venture capital that supports that. And all of those things reinforce that. Um and then that leads to success stories which we So you get kind of this whole soup. Um but that's my pet theory. So if there's a moon base at some point you're going to be bullish on that community's ability to

49:26do really well economically. Yeah, actually that's a great that's a great way to put it. The United States has has been persistently like innovative in in like a very amazing way. And so if there's a moon base I wouldn't be shocked if it's ours. Even though the Chinese are brilliant and and numerous um and the Europeans exist. [laughter]

49:50When you see rates go up like the the you see that great long-term chart of rates let's say the 10-year rate in the US. And it just looks like this for forever basically our whole lives and I was going uh comment on that chart. Well, I mean there's so much and I'm certainly no macroeconomics expert nor do I spend much too much time thinking about it but I'll make a couple of observations. Like you know, one obviously is just you know, in general when society has like you're you're society and you're you're building it you're making a giant investment in the

50:18future. Means you're going to have to consume less today. Like fundamentally all of the resources going into this AI buildout are about the future. Are about the future and they're and fundamentally like the the plumber building that today is not making consumption today. Like he could be making today's present consumption. Instead he's making a data center which is future

50:36consumption. So one piece of a very complex puzzle that I have thought a bit about is this idea that rates are higher cuz we we we're inducing people to save for the future. Um and the way you do that is you pay them. Simple. [laughter] People would kill me if I didn't ask about your portfolio a little bit which is uh so interesting to me. It's like 80% Carvana, vacations, [clears throat] 5% Meta. It's basically your portfolio. But like rough justice it's like five things. Sure. And one thing is 80%. And so if you think about your portfolio um and those companies which obviously you know extremely well

51:11and you were advising them on how to think about using Yeah. AI to transform their businesses, what's that speech look like? Well, it starts by making sure that everyone is vibe coding. Starting at the board and the CEO and like everybody needs to start building some, you know, they need to have um hackathons. And of course I I brought this to to Carvana and they said, "Yeah, we call them pizza parties." And I don't mean like the developers, I mean like

51:36everybody needs to have like hackathons. Um and it's not so much that you're going to build hugely useful software necessarily that comes out of these hackathons, but it's that you're going to teach everybody what can be done. And that's going to change their expectations, I think, for what can be done. Um and then the other thing, the second thing which I I I tell them is the CE- CEO needs to basically turn to the head of like IT and say something to the effect of "So, I want um my Grok bot, MetaMuse, agent, whatever to be able to access every system in the company on a read-only basis. Everything

52:09I could look up like if I like asked somebody I want to be able to get that in a prompt." And the guy's going to say, "That's really hard. We have like umpteen trillion problems with that." And he's going to say, "That That's my point. Like I want that." And you know, that's a big process and it's system by system and all the rest, but if you can get to that um then that gets you

52:29like read-only access. starting line. Yeah, to the starting line. That That gets you read-only access. Um and then the next logical thing is the CEO or whatever needs to like use this everywhere and everyone once is going to get jealous and you're going to say, "Oh, you can have it, too, but we need

52:40to work out your um permissions." [laughter] And and then that in theory percolates down and now you've got permission to read-only access and then invariably someone is going to say something like, "I'd really like to put data into the system." And you're going to say, "That's fantastic. We got to work out write permissions. Um and then someone's going to say, 'I'd love to change the system cuz we just this vibe coding thing and I'm confident we could do it.'" And you're going to say, "That's also fantastic. We need to work out, you know, a process by which we're going to

53:06manage changes to the system. And, you know, what I just described like like sure, it's like four bullet points, but it's like that's like a Years of work. It's like impossibly [laughter] hard. So, you know, the question is can are you in a business where you have time [clears throat] for that to happen before someone who starts for the jump ball, someone else that starts

53:26from go beats you. Exactly. And and so it's it's good to have a like a um you know, a really good business. It it seems like one of the conclusions from that is it's the best time ever to start companies both because of capital market conditions that people will fund at fairly low cost of capital highly experimental things, but also just because with these tools if you start without the baggage, you can go so far

53:51so fast. You think that's roughly right? I mean, yeah, it's the best time to start a company until tomorrow, right? So, I agree. It's amazing. I'd also just say that like just thinking from my perspective, right? Like you know, people are such a pain in the butt, right? Like as a if you're running a company like they're such a pain in the butt. And the idea that you have all these agents and stuff it it makes it so you can do so much more like with so many fewer people. Of course, you know, one of the challenges always is like if you can sort of easily

54:18do something, then like so can everybody else. So, there has to be some wedge by which thing you build um produces like enduring value. Does it have a path where you end up with some sort of like block in or proprietary data in this world where all the agents are happily you know, changing providers you know, every day like what gives you any enduring um economics. But then of course then there's the further thing, right? Is it you know, obviously if you if you do the thing where it seems like there's an obvious path to enduring economics, then you that's a bad thing to do because

54:55everyone is trying to do that. So, it's very competitive and the really great outcomes are the ones that make no sense. You shouldn't do it at all, but then it turns out there is a path or the world turns. do are the most underrated sources of future competitive advantage then? Like if if ultimately good businesses need sustainable edge sustainable competitive

55:15advantage to have great profits. Anything underrated? Underrated. We talked about attention and data and physical proximity. These are examples of things that might be super or more valuable in the future than they are today because of cheap intelligence. Is there any type of it maybe it's regulatory capture. Um are there any types of edge

55:35that you think are underrated? I get the sense that a lot of these are appropriately weighted. Um you know, I guess one interesting question I've had I'll I'll two of these I was like one is does this make margins like if you think of corporate margins in aggregate or do they go up or down?

55:50to say like this. Right, do they go up or down? And you know, I I think everything we're talking about would sort of argue down, right? Uh there's many businesses where you know, you know, they're making margin because More consumer surplus. More consumer surplus. Everywhere. What's interesting is if you go back and read a lot of the research reports from the '90s when the internet was coming along. They said they said the same

56:10thing. Um so the base rate here might not be great. Like maybe we're wrong, but I think the main reason they were often wrong was like was this aggregation economics of Van Thompson aggregation economics, right? So maybe absent that maybe it wouldn't have been so wrong. And I do think that like a lot of aggregation theory breaks down when you know, cuz that basically relies on the idea that consumers only look at the first thing on Google, right? But it turns out if your agent looks at everything on on search then being ranked first doesn't matter and um you know, all the sort of core things

56:37like quality and and price and everything else um service become more salient. Um but you know, the other thing is that maybe like another sort of downstream effect of this is that it'll you'll be able to build more complicated businesses, right? So so like right now a lot of industries are more horizontal in structure because the complexity of running your slice is very high um and and that you know, makes it hard to to do what's above and

57:03below you and not have undue complexity. With um AI agents and and basically cheap and ubiquitous code and intelligence like you could build the whole vertical stack. People trying to do this I think in other countries but like you can have a full stack pet care business that does everything from make pet food all the way up through pet insurance and pet hospitals and pet vets and pet and all these things that historically were multiple layers but now because you can solve the coordination and reduce the principal agent problems by having um by having AI do it like maybe you those sorts of businesses are possible and

57:37then maybe those businesses are are higher margin in in the end and more instead of more efficient and more consumer surplus. You know, Carvana is sort of an example of that built before this was you know easy to do but maybe people can do it in other verticals in [snorts] which case maybe margins go up and there's consumer surplus. Um The other one is the fun one of things here is just how this affects like wealth and income distribution, right? I can tell tell two stories you know on the one hand I think a lot of people have talked about this idea that it concentrates income distribution, right? You've got

58:07all these people who made money on like various labs and and like whatever data Nvidia and um this idea that there are like increasing returns to small groups of people. Um but the counter argument is if we have cheap and ubiquitous intelligence um a couple of things happen. One is there's all these rent takers in the economy. Think you're the software company that charges bowling alleys like you know whatever, right? And those guys all get smoked which sucks for them but it does lower the cost of bowling. And so then you know that increases living

58:37standards for everyone who bowls, right? Um and that's flattening because the my my spending on bowling is small proportion to my income but like people who make less money spend more proportion to their income on on bowling. And so the it's it's a flattening effect on on income. If like all these like little rent pieces like go away the decline in corporate profits like all the money that the average American is giving to Bank of America goes from away from Bank of America the capitalists lose you know like the the average consumer wins, right? So that's

59:06leveling. The other piece is like there's a whole professional class, right? Who basically you could define their role as Intelligence. Intelligence, right? They take big corpuses of information and apply them to certain circumstances in kind of a fairly like, you know, straightforward way. Um, and and if those all get commoditized, one, everyone gets the best ubiquitous, which is one good for everybody, but two, um, particularly good for people who can't afford the best right now or

59:35any. Um, but then three, those individuals, they earned these extra rents because they were smarter. They then have to drop to their next best alternative. Uh, and in theory then that might flatten the curve. Now, the counterargument to that might be something like, well, yes, but returns to initiative and ambition and intelligence could go up because now I can do so much more because these models are like a um, you know, like a bicycle for the mind on steroids, right? Like they they give people who are smart and and have initiative like the ability to accomplish, so much more. Maybe that actually causes like increasing returns

1:00:18to that subset of people and it steepens um, the curve. Truman asked for a one-handed economist. I'm, you know, I see I seem to very distinctly have two hands on everything. Um, but uh, [laughter] you know, I I I think that's a big interesting question. They're super interesting questions and and like I I've found that a lot of situations I look at like the there's a lot of cross currents. Like on the one hand everything we started the conversation with the AI buildout. On the one hand, I can see all these incredible ways to use AI and I use a good gajillion tokens and I can see how

1:00:48I could use a gajillion more and obviously I am in the far right end of the distribution. If everyone uses as many tokens as me, we are so short. Like it's going to be this huge thing that's going on forever and everyone and and everyone who's in the buildout is going to make a fortune, right? On the other hand, right? The the amount I get per token is like going to the moon and it turns out I'm a pretty unusual guy in a lot of senses and people are going to adapt more slowly and the cost per token is going to collapse. And so maybe actually, uh, like we're totally

1:01:15overbuilt, right? And like I can tell I can tell both stories like with equal certainty. I have a good friend who has been running this very cool experiment where if you break investing down into picking what to own and then how much to own of each thing that you've picked overly simplified but kind of interesting interesting way of thinking about it that he's been running this experiment where he continues to pick the things to own but then runs his own portfolio construction as he would have in the past in in one lane and in the second lane is just completely handing that off to an AI harness thing that he's built

1:01:50and says, "You do all the portfolio construction but stick with the stuff that that I've picked." And what he's found so far is that that is drastically outperforming his version of portfolio construction which itself is quite sophisticated. And the simple argument that he would make is well, AIs are just really good at optimization and they've read every paper on portfolio construction and and all this kind of stuff and bet sizing and Kelly criterion and like they they know it all and that in the future perhaps we make that transition where the investor's job is just picking what to own and not how much of of each to own. And I ask you this

1:02:26question what you think about this idea and the future because you have such an unusual portfolio where again 80% of it is in one specific asset and I guess is if you were in lane two your portfolio would the AI would weight your portfolio very differently than you have chosen to do it. I'm underweight. [laughter] So I'm just I'm dying to know what you think about

1:02:52about this idea. You know, underweight Carvana. Should be 90. Time time scales probably are interesting on that to think about too. Um, you know, as I think about how AI is going to encroach into this business I think it's going to start at shorter time scales and work its way up to longer time

1:03:08scales. And so be be you know, cuz one of the challenges is that like the there's no verification loop. Like your your friend's ability to say that is based on results over I mean, certainly not 20-year results. I strongly suspect that like the march of AI into markets and I mean, they're started, right? Like, I don't know how Renaissance Technologies did it, but like I assume Seems like they might have discovered a transformer before everybody [laughter]

1:03:34else. It's like I'm not kidding. It wouldn't shock me if people who invest over shorter time scales starting with with seconds and working their way out to minutes, hours, weeks, months find themselves handing more and more delegating more and more to models. Makes perfect sense in that you could you could definitely imagine bifurcating that between trying to enrich you know, the model. I think other people do things like this already and this is just a more sophisticated version, but you can you can have a a fully sort of quantitative strategy and then you can enrich it with what you think is is qualitative kind of human-designed

1:04:08analysis and then you can have a quantitative strategy to to incorporate the qualitative. I think many of these multi-strat strats like do these sorts of things. And what he's doing is kind of a just a slightly version of different flavor of that. It makes makes perfect sense. I'm not There are so many ways, you know, to be successful and and models have just added more dimensions

1:04:26of freedom. But what about your own portfolio? Say a word more about how you feel comfortable having a position. Forget the company that it is Sure. that's that big in the face of like so much research that would suggest that's a crazy way to build a portfolio. A lot of that research I think is is more looking at price changes in terms of like securities prices

1:04:50Mhm. and less focused on like thinking about it the way someone who you know, owns a company might think about it. I mean, we're talking about Carvana. Were it not for my sense that like Carvana's infrastructure and brand and everything else is is just so differentiated that there are few scenarios where it works out really

1:05:10badly. Um you know, then I wouldn't feel so comfortable owning so much of it. It might just be as simple as that. Like like over thinking over you know, that that portfolio approach would have worked absolutely, but it's certainly if you discovered Walmart or Home Depot or Fastenal or something back in the day, you wouldn't have

1:05:30taken full advantage of it. Um and that's kind of my sense, you know, with with Carvana. If you read history, of course, all the people who did really great tended to be very concentrated. But of course, the history is lit that you don't read the history of Yeah, you don't read the history of you didn't. I also think that a lot of that literature is is sort of what most people who are looking for average or slightly better than average average results should do and it's totally the

1:05:53right thing to do. Um but that's not what I'm looking for. I like Carvana a lot. It wouldn't upset me any if the stock went up a bunch and I felt compelled to diversify. That would also be okay. It's just more of a a trade-off um between expected returns and sort of knowingly taking a um an unusual risk. What do you think about education? You and I have kids of roughly similar ages. Comment a little bit on how you're thinking about their educations, the institutions that have driven education in this country that have been really important in that whole

1:06:27path-dependent story you've described. What do you think the present future of education looks like or should look like? I've been like talking to the head of my kids' school about this and it more and more it seems to be that his thinking and you know, and and my thinking are heading in this direction of like the way you're educating kids is getting more bimodal in terms of what you're looking for. Right. Let me step way back. What is the benefit of

1:06:53intelligence? [snorts] Personal intelligence, like your own intelligence, in a world where there's infinite intelligence from models. And you might say, "Oh, it's useless. Like doesn't matter. Like someone with a 50 IQ can navigate the world just as well as someone with a 150 IQ." And I don't think that's quite right. And intuitively, it doesn't feel right. But I think just to nail down where it is is that the more intelligence you have and and more crystallized intelligence you have, which would be sort of education, the better you are at communicating with a model and telling a model what you want. So, like I would be have a very tough time like

1:07:28having a deep conversation with the model to like create leading like quantum physics experiments because I don't know anything about quantum physics. And the model could try very hard to explain things to me, but in the end it would be explaining things in crayon and Nerf. And like it just like I we just like it would just be very hard to get to leading edge and like have a and get like the best out of what the model has

1:07:48to offer. Um whereas in an area where I do have some expertise, um the model and I can communicate kind of effectively and I can direct the model to my to my will like much better. And so, there's like a super important like you to direct a model you you to do your will, which I think is going to be an important skill in the future, you do have to understand something um about the world. You have to be you know and and smarts help for that. And education helps for that. Like you can't ask it to you know, assume this is a stochastic process unless you know what a

1:08:22stochastic process is. Now, you you certainly it could teach you what one is, but that's a lot of friction instead of this figuring this all out. You don't have to work out the math, but you have to know what it is, right? And so, when I've been thinking about education, there's like some this idea that like you're teaching your kids a bunch of foundational skills that build on one another where I think in the end the sum total of those they have to be able to do them kind of without models. Not because they're going to do them without models in the future, but because if they don't

1:08:51know how to do multiplication, it's going to be really hard to understand matrix multiplication. And if they don't know what matrix multiplication is at a at a high enough level, they don't have to be able to do it really well, but they have to be able to understand what the model is doing when it's doing it.

1:09:02And that's so they can direct the model. And so, increasingly I'm of the view that like we should spend a bunch of time with kids basically separated from these models teaching them basic skills that we think are important and what those are is a really hard thing, but the idea being that yes, model can do it, but you need to know it so that you

1:09:16can tell the model what to do. And then on the flip side, some English like I don't like models because like the the just cheat on their Shakespeare essays. I'm like you're not being ambitious enough. Right? If the kids are cheating on their Shakespeare essay, Give them something harder to do. Give them something 100 times harder and more ambitious and have them use models to their fullest extent. Right? Have them bring the work product back and then make sure that they didn't just give a prompt to the model do the homework. Make sure they own it. Ask them why it's like this, right? Force them to like justify it, right? And

1:09:45I think that's another really important thing cuz kids need to be good at using them and they also have frankly just need to be taught they can do like anything. You and I get to talk every so often on the phone every time I wish it was recorded so that other people could hear it cuz I learn so much and I think in this particular conversation today, which is sort of a version of a call we had several months ago, it's just such a valuable set of food for thought about the future of intelligence and what that means for markets and companies and uh is a little bit not in

1:10:13opposition to but a different perspective than the most common one that you're hearing, which is sort of unlimited, you know, unlimited build out forever, short everything forever, um no cap on the demand for intelligence [music] and uh I've loved talking to you about it both on the phone and today, so thank you so much for your time.

1:10:31Thank you. You know how small advantages compound over time? [music] That's true in investing and just as true in how you run your company. Your spending system is your capital allocation strategy. Ramp makes [music] it smarter by default. Better data, better decisions, better economics over time. See how at ramp.com/invest. As your business grows, Vanta scales with you, automating compliance and giving you a single source [music] of truth for security and risk. Learn more

1:10:57at vanta.com/invest. The best AI and software companies from OpenAI to Cursor to Perplexity [music] use WorkOS to become enterprise ready overnight, not in months. Visit workos.com [music] to skip the unglamorous infrastructure work and focus on your product. Ridgeline is redefining asset management technology as a true partner, not [music] just a software vendor. They've helped firms 5x in scale, enabling faster growth, smarter operations, [music] and a competitive edge. Visit ridgelineapps.com to see what they can unlock for your

1:11:25firm.