Byrne Hobart on What The First AI Agents Will Do + China, Google Analysis

Hey, Upstream listeners. Today, we're releasing this week's episode of The Riff, another show on the Turpentine Network. Bern Hobart and I discuss AI agents, the Chinese side of the internet, evidence that Google is hardcore about innovating, and more. Please enjoy.

So first I thought I'd ask you, you know, you wrote a bit about AI agents and where they could get deployed first. Why don't you share some of your thinking there?

Yeah. And so I think one of the things that we run into with a lot of these tools is that they can have, they can be kind of arbitrarily random in their outcomes. And, um, the more, the more thinking they're doing, just like people, like the more quiet contemplation they have and the more chain of thought they do, the weirder the outcome can be. And so if you want to deploy them in a useful way, you need some kind of real-world check on what's going on. And, um, that, that applies to a lot of different areas of generative AI. Um, you know, it's a lot easier to create synthetic data for, um, code than it is for natural language, because with code you can check and see if it runs and see if it produces the answers you expect. And, uh, with natural language, it's just harder to check that. You need a person to actually read it and decide if it makes sense. And, um, when I was, I was talking to some people about just where, where do natural, where, where does, uh, where did AI agents actually have that kind of feedback? And, um, one of the things I realized was that if you, if you look at tech support, um, you know, there are different levels of tech support and it's basically the simplest kind is this user did not actually read the documentation. So we will read the documentation to them. We will take their natural language question and answer it for them based on things that we were written, wrote down in a place that they could actually read them. And, um, that has, this is already being automated pretty quickly. And it's just a very natural use case for, for a generative AI tool where you dump the existing documentation into a RAG and then you query that and then you're, you're good, right? Um, but there is that other category of help that is kind of similar, but it's basically they're looking at the product, they're trying to use it, and they can't quite figure out how to do what they're trying to do, but it is probably possible to do within the application. And that's a case where a software agent that is just trying to use the application would be extremely helpful, especially if it does have comprehensive knowledge of what this application could do. So that is, that's, there's still room for human discretion and human judgment. The agent is not going to have perfect knowledge of what the user's actual intention is. And, um, it may, you may still need to step outside that context and really figure out what's going on. I think it does help automate a different layer of what customers are doing. And it automates it in a way, or sorry, it automates a different layer of what customer service people are doing. And it automates it in such a way that you can actually get a lot more data because if you find that a lot of your customer queries are asking to do the same thing, then you may find that there's a customer persona that needs to make this thing easier to do, or just that the application should make this thing more feasible to do. And, um, you, you know, you could have your human customer service people take very detailed notes on every step of the workflow, but it could be like the agents are way more willing to just jot down notes on everything they do. And, you know, so once, once you've done that, you could actually see that as a kind of layer of intelligence gathering. And that gets to a broader question, which we've talked about on the RIF and I've written about in the Diff, which is, um, how quickly do customer service and product management roles start to merge as a lot of the easy customer service stuff gets taken care of? And as a lot of the higher cost, higher touch customer service stuff actually becomes much more informative about how people use the product, what they think it should be able to do, and whether or not it can actually do those things.

Yeah, no, it's a good, it's a good overview. I want to segue to something else you wrote about as relates to this idea of return on equity and how it's a classic metric of evaluating company and a good way to evaluate Businesses across industries, but it sort of becomes a stranger signal when markets get more efficient at underwriting growth. Yeah.

So this is a fun one. Like, there was this sort of return on equity revolution in corporate finance, which I think stemmed from a lot of— so for a long time, companies were— there was just not quite a metric they were thinking about or was indirectly a metric they were thinking about, but Not something they were explicitly targeting. You had this spate of empire-building companies in the 1960s where you'd have a high-growth company, they have a highly valued stock, they realize they can just mathematically increase their earnings per share. Like if they traded 30 times earnings, they use their stock to buy someone who trades at 10 times earnings, they have just increased their earnings per share. And so when they did that, they just kept acquiring slower and slower growing companies because they're going for low multiple. And for a while they were high-growth companies that were acquiring low multiple companies. And then when their own multiples came down, because people realized, you know, this company used to be an electronics manufacturer and that was an amazing business, but now they're mostly doing steel and meatpacking or whatever. And that's just not a hypergrowth business. So those stocks started to come down. And what we ended up with is this cohort of big companies where they had a lot of assets. They were in theory showing a lot of growth, but that growth was financially engineered. And now they're just not getting great returns on those assets. And so people started thinking a lot more about return on equity as just this measure of the quality of the business. You've put a dollar into this business. Do you get back enough money to justify the uncertainty of the return? And when companies started doing that and started charging their subsidiaries for capital and ranking them based not on growth or emergence, but on did you actually get a good return on the investments you made? That did encourage those, that did encourage different business units to actually think pretty seriously about how they're using their capital and to be better stewards of capital. And that's all well and good in a world where companies are mostly funding their growth from their internally generated profits. But as markets have just gotten larger and more efficient, and as we've built out this venture ecosystem, companies really don't have to do that. There are plenty of businesses that do, but One of the businesses that grow, it is just very rare for your growth rate, your ideal growth rate as a business to be precisely the growth rate you can fund. Either you have a very high return business, but it can't really scale. So think of like a service business that has one really exceptional service provider, like maybe it's a neighborhood barbershop and the barber knows everybody, everybody's the barber's friend. So their cost of customer acquisition is low, they get really high returns, but when they open up a new location, It's just some new random guy who got here and just as good as the Supercuts guy, but doesn't have quite the same brand, for better or for worse. Their marginal return on equity is low. So they have a high return on equity, but they can't really grow at that return. And then you have other companies where their return right now is low. So think of the classic enterprise SaaS company where they spend a lot to develop the product and they spend a lot on a very long sales cycle. Once they get their customers, they're going to be getting revenue from those customers for a long time. That revenue stream is probably going to grow over time. So you almost want those companies to be losing money as fast as they can rationally lose money. And in, I think it's Amp It Up, Frank Slootman talks about this with high-growth companies, that it's not that they're totally insensitive to costs, it's that they're very, very sensitive to the unit economics of here's what it takes to acquire a customer, here's what the customer is worth, and how can we acquire those customers as fast as possible? So at that point, at any given time, those metrics are just a really messy snapshot of the ultimate economics of the business. And that company is going to be continuously raising money in order to expand as quickly as possible until hopefully until they reach the point where that expansion is no longer justified, at which point their, their metrics are going to start drifting in a good direction because their existing customers are continuing to spend more money. They're not dumping so much money into marketing. They can kind of tamp down their R&D spending or at least not increase it in share of revenue over time. So their, their accounting economics start to look better. Their accounting numbers sort of look better, but their fundamental economics haven't really changed. So suddenly return on equity becomes this metric where it's really not an action-guiding metric, except in cases where the company has a very low return on equity and is continuing to reinvest in their business. Like in that case, you say this company is just a money pit and they are continuously destroying value by You know, if their cost of capital is say 9% and they can earn 5% on that capital, then every dollar they put back into the business is value destructive and they should be trying to liquidate and return assets to their shareholders or should be trying to fix up their business. But with other companies, if their return on equity goes from say 20% down to 15%, that probably means that they have identified this set of opportunities that do earn 15%, which is good. And they're pursuing those as aggressively as possible and they're saturating those opportunities. And so you kind of want your companies to have their return on equity rapidly approach roughly their return on capital, rapidly approach their cost of capital, because that means they are deploying money to every opportunity that has a better return than their shareholders can get elsewhere. And that's kind of what you want. So, um, it was, it was a useful exercise to think through why did return on equity make sense? And it made sense in more of a capital-starved, information-starved world where companies fund a lot their growth internally. And why is it that we can look at companies today and say, like, some of these companies have really high returns on equity, some of them have very low returns on equity? Doesn't really correspond to our intuitive sense of the quality of the business. So if you look at someone like Apple, and this is another thing that distorts it, is buybacks. So as companies mature and more of their revenue is locked in and very predictable, it is possible for them to borrow against that revenue and buy back stock. And it has For many companies, historically been a pretty good decision. So you look at a chart of Apple shares outstanding, you'll see they've bought a lot of them back. If you look at a chart of their stock price overlaid against that, you'll see that they bought them back at pretty good prices. So it was the right call, but every time a company borrows money and buys back stock, it's reducing its equity, but it's not really messing with the fundamentals of the business all that much. And so it's artificially increasing the return on equity, but nobody looks at Apple and says, They can definitely invest a marginal dollar in the business at the return on equity that they have on the historical investment they've made. It's just an accounting artifact that the numbers look the way they do. And then you can take that even further and look at a company like TransDigm, where technically they have negative equity. They have borrowed and bought back enough that in an accounting sense, the company's value is less than zero. But of course, the actual value of the company is the future cash flows and not the book value of the assets. So It's perfectly fine. Nobody worries that TransDigm is going to run out of money just because they have a negative equity. And then rewinding a bit, we've mentioned that return on equity does help compare businesses across different sectors. And that is one of the great things about it because many, many businesses require capital. And even if they don't initially require capital, some of them will require it to grow. And you can sort of kind of in your head try to capitalize some of their growth costs. You could say, Okay, this company spent $1 million this year on marketing. We can treat that as an expense, and for tax purposes it is, and for GAAP accounting purposes it is. But really, if they spent $1 million to acquire customers who are going to last for the next 5 years, you should think of that as they spent $200K a year on marketing over the life of that customer. And we capitalize that, and then we treat that as the asset and look at return on equity based on that. And you do get numbers that just let you basically ask what are otherwise incoherent questions like, should we invest one more dollar in steel or banks or insurance or SaaS? Very hard to compare them. But if you, if you do the accounting right and get the economic balance sheet measured, you can actually ask and answer that question in a useful way.

There's this famous interview over 10 years ago now between Peter Thiel and Eric Schmidt. Where Peter Thiel is saying, hey, you guys are sitting on all this cash. It means you have no idea how to deploy it. You have no idea what to spend it on. And this was a time where he was most making his case of technological slowdown and using sort of a corporate— Google as an example of a company that has all the money and is invested in innovation and claims that it wants to solve all our problems, but doesn't, just doesn't, is short of ideas of how to actually spend it.

Yeah, I, I've wondered about that speech because I think it was accurate. It was pretty accurate at the time, or to the extent that it was inaccurate, it was because Google was kind of trickling out money into a couple of these projects, but not talking about them a ton. I think some of this It was probably a PR thing where Google, at the time of that talk, I don't remember exactly how long ago, it was long ago enough that Google did not have to try that hard to make people, make the people it wanted to hire think it was a good place to work. There were a lot of people who agreed that if you are good with computers and you can get a job at Google, you should absolutely do that. And you've pretty much got it made. And that was roughly true. Over time, as Google has become one large company among many, and as it's gotten more slow moving and bureaucratic with its core business, and as that core business has matured to the point where, you know, the good news for shareholders is, if you spend $1 million to increase the click-through rate on ads by a basis point, it's a really, really good deal, very much pays for itself. The depressing thing for employees is like, they can feel like they worked really, really hard for a very long time. And all they did was increase the click-through rate on ads by 1 basis point. And is that really, you know, what you want on your tombstone? So if they can think to themselves, okay, that is, that is what I did. And what that supported was self-driving cars and curing aging and, you know, drone delivery and all kinds of other cool stuff, then it's probably in some sense worth it in the end. Um, so I've wondered if Google just wasn't talking about some of those projects as much, or if, um, that, that little debate might've been kind of the, the kick in the pants that the company needed to start really expanding these, these other bets and turning them into something, something real. And, um, I took my first Waymo ride last week. It was awesome. I don't want to go back. I'm on the waitlist here in Austin, but I was in San Francisco, so I was able to take one. And it was great. It was just, you know, everything you like about Uber or Lyft and none of the things you don't. And pretty nice. It was like a, you know, in some ways it was like, I was very curious about Google's decision because they tell you that you can play music from Spotify or whatever in a Waymo, but you have to download a special Google app and then you can only ask for music in a natural language interface, which was actually really, really clumsy and annoying. And I could almost feel the product manager responsible for the flow just apologizing to me that this is what they have to do. But it also felt like a little bit of an aggressive thing where Google is not content to get just some incremental revenue from one more ride in a self-driving car. They also wanted incremental install for Google Assistant. And I'm sure this decision that you can't just, you know, connect Bluetooth to, um, to your Waymo and play things through their speakers. I'm sure there were lots of internal debates over exactly how this would work and lots of internal turf wars and we're kind of seeing the output of that, but it also felt like the kind of slightly aggressive thing that companies do when they genuinely want to win. And that is, that is always one of the concerns that I've had about Google is they're, they're very nice and, um, they don't like to fire people. And, um, you know, they like to, they like to maintain their monopoly and, you know, make the web a better place, maybe, maybe more homogenized place, but a better place overall. But are they, are they really trying to win? Are they really trying to kneecap their competitors? And that was just a moment where it felt like, yeah, there's, there's a little bit of a little bit more kneecapping going on than I had anticipated.

We'll continue our interview in a moment after a word from our sponsors. How deep do you go to seek out an answer to a question? Maybe you've spent hours clicking the source links on an obscure Wikipedia page, or maybe you're even the type of person who checked out the entire shelf on the topic at your library. If you're nodding along, then check out GiveWell, an organization that researches questions about global health and philanthropy, even if a satisfying answer might require years of reviewing studies, talking to experts, and over 300 footnotes. GiveWell has now spent over 17 years researching charitable organizations and only directs funding to a few of the highest-impact opportunities they've found. Over 125,000 donors have used GiveWell to donate more than $2 billion. Rigorous evidence suggests that these donations will save over 200,000 lives and improve the lives of millions more. GiveWell wants as many donors as possible to make informed decisions about high-impact giving. You can find all of their research and recommendations on their site for free. You can make tax-deductible donations to their recommended funds or charities, and GiveWell doesn't take a cut. If you've never used GiveWell to donate, you can have your donation matched up to $100 before the end of the year, or as long as matching funds last. To claim your match, go to givewell.org and pick podcast and enter Econ 102 with Noah Smith and Erik Torenberg at checkout. Make sure they know that you heard about GiveWell from Econ 102. with Noah Smith and Erik Torenberg to get your donation matched. Again, that's givewell.org to donate or find out more. Hey everyone, Erik here. In this environment, founders need to become profitable faster and do more with smaller teams, especially when it comes to engineering. That's why Sean Lanahan started Squad, a specialized global talent firm for top engineers that will seamlessly integrate with your org. Squad offers rigorously vetted top 1% talent that will actually work hard for you every day. Their engineers work in your time zone, follow your processes, and use your tools. Squad has frontend engineers excelling in TypeScript, React, and Next.js, ready to onboard to your team today. For backend, Squad engineers are experts at Node.js, Python, Java, and a range of other languages and frameworks. While it may cost more than the freelancer on Upwork billing you for 40 hours but working only 2, Squad offers premium quality at a fraction of typical cost without the headache of assessing for skills and culture fit. Squad takes care of sourcing, legal compliance, and local HR for global talent. Increase your velocity without amping up burn. Visit choosquad.com and mention Turpentine to skip the waitlist.

Yeah, that makes sense. Relatedly, or speaking of Peter Thiel, he also had this other interview, and I would use this as a segue to change topics. He had this interview with Barry Weiss, where he was talking about tariffs and he was saying, yeah, you know, affect the American consumer, American worker, but it will affect China more than it will affect negatively, you know, negatively affect China more than it will negatively affect us. And thus it makes, you know, there's some logic to doing it. You wrote a little bit about tariffs and a little bit about sort of the trade imbalance. I'm curious if you— Think that that logic is sound or, and you think we're headed towards, like, is that, you know, I think he was complimenting sort of China's efforts the past couple of decades where they were able to build, you know, their own replacements, basically their own social media giants or tech giants that didn't rely on the US and thus became more independent that way. And people wonder if there's, you know, if we can do that to some degree too for some of these solutions. What are your thoughts on that?

I would look at tariffs from a macro and micro view. And the macro view is that tariffs are just affirmative action for American semi-skilled manufacturing workers. Like there's, there is a, you can just mess with the price points such that even if China is objectively more competitive in some category, if we put a high enough tax on things imported from China, then the US starts to be competitive and we'll do it instead. So it is indirectly a transfer from US consumers who are paying more for goods to US workers. And as with any kind of transfer payment like that, there is some deadweight loss. Like those workers might be doing something the US specializes in more. They might be doing nothing. Maybe they would impose some externalities if they didn't have jobs. But that is, that is roughly what we're doing. And that to me is just, it, you know, it's an economic question as to what is the cost. And then it's a values question as As to, is it better for this country to have a slightly higher employment rate and a little bit more manufacturing that is not really our competitive advantage, but does keep people busy, give them a job, and, you know, make them, make them A, feel like they're contributing, and B, make them actually contribute materially to the standard of living in this country. Is there, is there a fair trade between that and having a slightly lower standard of living? And that's, that's the question you ask about any kind of Transfer program, because any, any kind of transfer is going to create different incentives for working versus not working or working in one place, you know, working on one thing versus another thing. And we're, we're always just asking about the trade-offs. And then that, I think, you know, if the— most countries have, most countries that can do it have some form of welfare state. And, um, it's probably just better for social cohesion if the welfare state does involve actually working. And if you have tariffs, you are imposing some market discipline on that work. Um, there, there are places that have historically had welfare programs where there was nominally a work requirement. Um, and there's a book, uh, the title, but it's, um, William F. Buckley's book on running for the mayor of New York City, which, um, it was, it was not really a serious campaign. When he was asked in his first press conference after he announced he was running, he was asked what he would do. Uh, what's his first act? If he wins, and he said the first thing he would do is demand a recount. And he definitely did not win. But one of the things he pointed out was that at the time, in the 1960s, New York had a law on the books that if you receive welfare, you have to work for the city. And he noticed that that law was just not enforced at all. But even if you enforce it, it's probably much more expensive to hire supervisory people and figure out, okay, what can we get someone to do that is like a fake job, but has some kind of actual benefit? And Everyone involved in this would know this is fake. We are pretending to work and, um, it doesn't really matter if we accomplish anything or not. We are getting, we're getting the money either way. And that is, I think, a lot more demoralizing than just getting a check from the government every so often. But if you have a tariff such that there's a job that really only exists in the U.S. because we are keeping more cost competitive countries out, it is a job. And someone can actually look at your profit, they can look at their profit and loss and decide whether or not to hire someone. And they can say that even if this job only exists because of these artificial constraints we've set up, you can be good or bad at that job. And we can fire you for doing a bad job. And so it's more of a real job and less of just a pure make-work program. You know, every, every society sets up certain sets of rules that make certain things a job or not a job. And, um, In some societies, drug dealer is just a nonexistent profession. In some societies, it provides goods and services. And then there are externalities that we decide whether or not they're worth it. So we're always deciding in some implicit way what counts as a job, what counts as the job that we actually want to countenance in our system as a legitimate job. And I think of all the ways you could make that call, tariffs are probably one of the more prosocial ones. But it is, it is going to be annoying for, for those of us who are not really in the pool of potential workers in American low-value-added manufacturing. We will just pay more for stuff. But we were already going to pay more for stuff in order to support a social safety net anyway. It's just a question of exactly how you arrange that. So that's the macro piece. The micro piece is that, yeah, one of the ways you can use tariffs is you have an industry where it would work at scale. It won't get to scale unless it can generate some revenue now, but it has no hope of competing with foreign competitors who were already at scale. And that is the story of a number of countries' industrial development is that they, they see what industries work in different places. They see if they can replicate those industries at home in theory, and then they put up enough trade barriers that they can actually replicate them in practice. And then they sort of take down those trade barriers as the, as the industry gets up to scale. So this is part of how Japan ended up with a steel industry and then how Korea ended up with a steel industry is that It was supported and protected by the state until it got to the point where it was actually cost competitive with the rest of the world. And at that point, it's basically paying back the benefits it got through the positive externalities of one, it's a domestic manufacturing business, it employs people, it produces profits, et cetera. And two, it's just useful for countries to control resources like that and be able to produce them domestically. I don't think there's a very serious risk in the next 10 or 20 years that the US will just not have access to steel and really need it. But I don't think the risk is zero. And so I think it is useful to think about what strategic products we could hypothetically lose access to. What is the lead time for making sure that we can produce them in sufficient quantities at home? And then, yeah, potentially taking the economic hit in order to do so. So you can think of it as state-sponsored private equity or venture capital where you have this investment, That only makes sense if you write a sufficiently large check and does actually produce a return once you write that check. And maybe it's like ESG flavored for a different set of E, S, and G, but it is like an ESG flavored investment where you're saying this is not the thing we would invest in to get the maximum possible return, but between the financial return being better than zero and the social return being positive enough to make up the gap between that and the next best objective financial outcome, We can still agree that it's worth doing. I think that, that raises its own complications because you have to figure out what are those industries where the U.S. should compete and actually can compete. Because what you really don't want is a system where we have a set of targeted tariffs and there's a whole industry that is just devoted to exploiting those tariffs. But that is, uh, that is a very real risk.

Correct me if I'm wrong, but it seems like China did a pretty good job of replacing or building their own solutions to the services that they were blocking, you know, their own search engine, their own social networks in a way that didn't seem like a huge sacrifice for them, at least long-term. One, is that accurate? And 2, do you think the US could do something similar or will it not just be a short-term sacrifice, but also a long-term sacrifice and the products or services would just be worse?

I think the Chinese internet industry is actually extremely impressive. And from what I understand, Baidu was, you know, pretty far ahead. There's, There's an interesting kind of accidental linguistic protectionism that really applied to earlier search engines. It's less relevant now because, um, a lot of language models can translate pretty seamlessly across very different languages. But there was a time where, um, if you have a search algorithm that works just fine in English, it's just not going to handle, um, it's not going to handle Mandarin especially impressively. And, um, this was also true with Russian where Yandex was actually better at handling some of the nuances of the Russian language. And so it was for a long time a at its search engine compared to Google. Um, not, not because Russia was really trying to support their domestic champion. That wasn't the main cause, although they, they certainly cared about that after a while, but really because you, you had to choose which language you wanted to go after. And of course the, uh, the Russian speaking engineers who founded Yandex, they picked Russian. So, um, you do, like there, there was some just independent development of different search engines in different places and Baidu did. A really good job, and then yeah, after a while there was also a protectionist element to this where Baidu is the local company. If Google invests enough resources, they can win in China. The Chinese government feels like that that is actually a policy decision. It's not just a natural outcome of these totally unstoppable forces. It's something that someone decides whether or not should happen, and they decided it shouldn't really happen, more or less. Or I guess there's there's a near possible world where they decided that a a Google that is as compliant the CCP's wishes as Baidu. Is can be tolerated. But I think even that, it presents a risk to them because that does mean that they need Google to always feel that way. And it's just very hard to get rid of a search engine that people are using and like more than the competing one, sort of get rid of it on political grounds. Just look at various, various efforts to dislodge Google. They don't work very well. Even, even when you tell people, you know, if they're buying a new Android in Europe, they get to choose whatever search engine they want. Well, There's one search engine that almost all of them are used to and almost all of them prefer, and that's the one they choose.

That makes sense. Somewhat relatedly, although it's a segue, you wrote about this idea, what did you call it? One big shot approach to wealth creation where you basically talk about how for some countries, a lot of their sort of growth comes from a specific technological development. Why don't you unpack this idea?

So the idea was not just— so there's, I think a lot of the national wealth does come from the deployment of these big general purpose technologies, but a lot of the individual wealth accumulation is actually adjacent to those new technologies. So if you look at the railroads, for example, there were people who made money by starting a railroad and just slowly expanding it, but there was a lot more money made from promoting railroads or taking enough control of a railroad. This is 19th century, like taking enough control of a railroad that you can manipulate the stock and trade on the things that you were doing to make the stock price go up and down. Like that's where a lot of the big money was. And then, um, in railroad adjacent, um, physical industries, there were, you know, steel fortunes that were made from selling steel initially to finance the railroad buildout and then shipping steel on railroads in order to finance, uh, or in order to, to supply things like The development of skyscrapers and of just steel in everything. And similarly, oil benefited a lot from there being a railroad network where Standard Oil was a very big customer for the railroads and was able to get very, very good pricing treatment from them. So the actual core business often doesn't create as much wealth as the consequences of that business existing. And if you look at the automotive sector, You do see something similar where, yes, Henry Ford made a ton of money, but if you look at the aggregate wealth that was made from the rise of suburbia, increase in demand for oil from cars, the rise of suburban-centric business models like grocery stores, that just swamps the money that was directly made from owning stock in a company that made cars. So, and I think even if you look at more recently, the the world of software, um, there, there were a lot of people who, there was a lot more money made from noticing that chips get better and better, and therefore we can make software that's a complement to this thing whose price is asymptotic towards zero. Um, a lot more money was made from that observation than from actually making the chips better, even though, again, some, some decent-sized fortunes were made from that. And, uh, you know, Jensen, Jensen is doing just fine right now. Even Nvidia's case, part of their, big part of their moat is the software side of things. So there's, you still, like, if you find some really, really big trend where some kind of technology is improving rapidly, declining in cost, you think it will be deployed everywhere. The tempting thing to do is just bet directly on that. And then the wealth maximizing thing to do is bet on the indirect consequences of that and just In fact, in that piece, one of the claims I made, which I'm, I don't take entirely seriously, but it's fun to think about, is that the success of value investing in the last half of the 20th century was partly a bet on cheaper information transmission, like cheaper, cheaper communications, cheaper information processing, where if you're Warren Buffett and it's 1951, you have to drive to the state insurance office to get, um, get the financials from these insurance companies. And then you have to put an ad in the paper saying, I will buy your shares for $30 whatever. And then you have to go door to door and find people who bought the stock 20 years ago and still have those and buy it from them for cash. So it's a lot of work, but now you can open up a stock screener and find everything that looks even slightly cheap and buy it. And it's very, very straightforward. You don't even have to pick up a phone and call a broker. So, uh, the, the drop in the cost of information was a subsidy to anyone, any investor who was willing to do extra work to get that information first. And that was like a one-time subsidy where markets were very inefficient because data was scarce. And as they got more efficient, the people who were exploiting those inefficiencies made this one-time gain from everything snapping into the right valuations. And now it's just a lot harder to repeat that.

Yeah, it's interesting. What is the right mental model to have when thinking about which types of technology innovations show up in the Statistics versus the ones that don't really lead to any observable growth developments?

So I would say there are 2 pieces of that. One is that if technology has a very low marginal cost, like software or communications, then there will be a period where it's being deployed more and it's being used more, but it's underpriced. And so at that point, it doesn't contribute as much directly to GDP as it does indirectly through people using it in order to do more real-world measurable economic activities. And then over time, it starts contributing more to GDP because it's fully deployed. We've saturated all the customers who could use it, and now our growth comes from getting more money out of those customers. So that is GDP accretive. So I think you'll, you may see something where software was deflationary until everyone had a Facebook account and an Instagram account and everyone used Google all the time and everyone had an iPhone, et cetera. And, you know, everyone's subscribed to all these things. And then at that point, The subscriptions keep rising in price and the ad loads keep creeping up and the prices for those ads keep rising. And so software actually becomes inflationary because the product isn't improving all that fast, but the cost of that product, the economic activity that's driven through it is rising. So I think that's one part of it. Another part of it is that sometimes technologies will make a contribution that actually shows up on the input side rather than the output side. So, um, There was a really interesting piece on The Great Suck Nation that was written in early 2020 that talked about aspirin as an example of this. And it pointed out that aspirin, we think objectively, just increases productivity in the sense of you have the same set of people, the same set of capital equipment, but it is less likely that a headache means you can't work that day. But when you're compiling macro statistics, you would look at things like number of days people work or percentage of sick days people take, et cetera, when you're figuring out what the labor input is. So if the net effect of aspirin is everyone has one more workday because they're not miserable with a headache on average one day a year, then what you actually get is a roughly 0.2% increase in total labor input given the same set of laborers. So you have an increase in efficiency. You definitely have an increase in just human happiness, like the headache goes away, but you don't actually see that as an increase in productivity. You see that as an increase in hours worked.

You know, Aaron Levy, the CEO of Box, was on the More or Less podcast with Sam Lessin. And Sam basically tried to get Aaron to accept the idea or the possibility that AI might be deflationary in that there's no natural guarantee that it will lead to more usage. And so maybe if it just makes it, you know, cheaper, maybe there's actually less profit Uh, to go around. And Aaron was saying, hey, there's no way we're at peak software. You know, there's like making things, you know, cheaper and easier and just introducing new workflows will likely lead to, or will most certainly lead to more usage and thus more, more profits to be generated. Is this, is this your sort of view as well? This kind of this broader question as to when new technological innovations could be, if it in fact, you know, doesn't lead to more usage.

So I think there's a J-curve with respect to revenue and the inflationary impact is really tricky. So the J-curve for revenue is as you're deploying, say, AI customer service agents, and as you are, I don't know, replacing your direct-to-Kindle cover illustrator with just a call to Midjourney or DALL-E or whatever, um, you are reducing costs and you're eliminating some jobs. And so that's deflationary. And then over time, my strong suspicion is that the vast majority of consumer-facing AI products will be primarily monetized through ads, and that those ads will just get better and better because the AI can constantly steer the conversation it's having in more and more commercial directions. It's collecting more data because you're just interacting with it more. You're giving it more tokens of data and just giving it more feedback. And meanwhile, tech companies are just getting better at collecting data all the time, getting larger and larger samples. So over time, They do generate more revenue, probably drive more purchases through these chatbots or, you know, through AI-generated search results and things like that. But then, you know, the question of, are those, are those incremental purchases actually inflationary? And a lot of the things that consumers buy are just not very capacity constrained. Like we can make more iPhones without the price of iPhones going up and, you know, make more of just a lot of electronics, furniture, clothes, lots of consumable, uh, like lots of consumer durables, lots of consumable products without this increase in costs. So, um, so probably, and in fact, we can probably get some economies of scale. So it's entirely possible that it ends up being just accidentally deflationary in the sense of we have a lot more control over spikes and declines in demand because we have these tools that are very, very good at targeting ads. And if there is something that's been overproduced and is a little bit, um, has a little bit of a glut of it. And so it's taking up very expensive warehouse space and we want to get rid of it. Um, you can probably drive a lot more consumer interest towards that thing because your chatbot will steer you towards the idea of buying this thing. And, um, so maybe, maybe that actually reduces the overall cost of manufacturing goods across any category where they can be marketed to consumers and ends up being net deflationary. But It's also, um, I think it's one of those cases where a world where these tools are fully deployed, where that is in effect, is a world that is sufficiently different from the current one that there's just going to be a lot of other uncertain questions. And there's also the question of, does AI end up being inflationary because it is, um, it can convert, it can turn electricity into economic activity very efficiently and very, maybe linearly or even super linearly. And so. We would actually see deflation across the manufacturing stage of the supply chain and across the transportation stage and storage stage and all that stuff. But then inflation in the raw material inputs because some of those inputs are energy sources and we'll use a lot more of those. But then you have the, the other open question of how much energy efficiency can we drive? How much solar will we deploy if we have just radically higher expectations for power demand? How much You know, how cheap does nuclear get if we treat it as something which should be mass-produced rather than something where every power plant is this bespoke thing that has to be designed and redesigned over and over again at great cost, and you can't break ground till many, many years of debate and discussion and, you know, creating giant reports have passed, et cetera. So lots, lots of uncertainty there, but that's, that's how I would think about the, the direct impact of AI on just macro variables is. J-curve where at first it's destroying some demand, but that's offset by increasing demand because you have to actually build the data centers. You have to actually hire people. You have to spend a lot of money on marketing to get these tools in front of consumers. And then, um, at some point, as, as they, uh, start monetizing more by driving purchases rather than through subscriptions or some other revenue source, and they start to just increase the overall, uh, level of economic activity.

Yeah, no, that's a helpful, uh, Mental framework for it. I want to segue to something else you wrote about, which was passive investing, I think in a couple of different places as well as capital gains. Why don't you share about some of your thinking there? Yeah.

So I've always found the phenomenon of passive investing really interesting because there is like, I framed this in Capital Gains by saying, here's a really, really easy argument that passive investing can't work. So let's say you have 2 investors. One of them spends all of their time researching companies and trying to figure out which one is good. The other one picks companies at random. How could the one who's doing research not actually win over time? And then the counterargument to that is, well, if you look at the set of all active investors, necessarily they in the aggregate form at the average level, less whatever transaction costs or taxes they pay. And so half of them, you know, even in a perfectly efficient market where there are no transaction costs and no taxes, Passive active investors have to underperform on a dollar-weighted basis. So you have this knockdown intuitive argument that passive investing can't work, and then this very strong kind of first principles argument that it has to work. And the first principles argument is generally winning. Passive is still taking share. But when people talk about kind of apocalyptic scenarios of what if the entire world goes passive, I think they are thinking through the mechanics of that because Investing is always on this continuum where it's always some form of market making, where someone is buying something cheap in order to sell it dear to somebody else. And when you do literal market making, you're just doing that continuously, instantaneously. You're making a market in SPY or S&P futures or Apple shares or whatever. And you just expect to be buying at a slightly lower price than you sell and repeatedly buying and selling and On average, you have a slight edge and that edge accumulates, or that edge means you're accumulating money and occasionally losing some, but mostly making money on slightly more than half the trades and making a little bit that just adds up to a lot if you're doing this at scale.

Yeah.

That, but then if you look at what someone like a multi-manager fund where they're very catalyst-driven, they're essentially doing a slightly slower motion market making where they're saying, We think that this company is, we know this company is presenting at a conference next week and we have a bunch of reads on how the data is looking. So we think they'll be slightly more optimistic than investors. So we're going to accumulate an inventory of this stock now in order to sell it in advance of slightly higher demand in a week or two. And so they are making a market like they are, they're accumulating inventory now in order to sell it later in response to demand that they anticipate. So they're doing market making on a slower scale. And even someone like a long-term buy-and-hold investor, or even a private equity firm, is also implicitly doing that. They are making a market in bulk in a weird way, but they're still doing that. And so that market-making is necessary for markets to function. You need a buyer and a seller, and it is just very rare for a buyer and seller to show up like a natural buyer and natural seller, someone who owns a stock and wants to get rid of it forever, someone who doesn't own it and wants to buy it and hold it forever. Very unlikely that they happen to want to trade at exactly this, the right time and exactly the same price. So you want some intermediation between the two, and that's what market making is. And everything is on that continuum. So when people talk about passive dominating the market, at some point that has to mean that the market makers don't make markets anymore. But at that point, passive can't function. Like there's no, no seller or no buyer. There's just one universal buyer or universal seller. And so prices have to go to infinity or zero because there's no counterparty. So you can't actually get to that point. And meanwhile, when we talk about passive, we're implicitly saying that we know what the market is. We've defined the market and that is what the market is. But passive doesn't really cover private companies. There have been efforts to do something that is kind of like an index fund for various kinds of privates, but it's very early stages and kind of hard to pull off. So if you have a setup where there's the market and it's mostly passive, and then there's not the market and it's active, then passive investing's returns actually start to get weaker because people aren't directly analyzing, or the prices of public companies aren't being set as accurately, which means that someone who's a PE firm, for example, they can systematically buy low and sell high. And so they're capturing more of the overall upside from equity ownership by taking something in or out of the sample of public companies that are the market at strategically optimal times. So what I ended up concluding from this is that one, the idea of passive share of overall investing is just a very messy, fuzzy concept. It gets messier and fuzzier as the share goes up. But 2, there is probably a cycle where Sometimes passive gets really big, and then at that point there are so many opportunities for active investment that passive returns actually start to go down. And at that point, more investor money shifts to active, and then we're kind of back where we started. The reason it doesn't look like a cycle is that we're just in the first wave of it. But I would guess that over my lifetime, it's entirely possible that passive peaks, and then we say, well, passive was a cool trend and it's kind of passé now. And then it reaches a low point, and then people start to realize, hey, when I look at the returns from private equity firms, they're actually not that great compared to returns of just buying an index fund. So I have to go back into passive. And we'll probably see that oscillation happen multiple times.

Well, on the more extreme side, or at least one angle of active investing, you know, as a segue, you wrote about meme coins and you also did a Capital Gains piece about, you know, what happens when an asset's price follows a self-referential feedback loop with the case study of MicroStrategy and their Bitcoin bet. Why don't you talk a little bit about that?

Yeah. So meme coins, I kind of, I think they fit nicely into the concept of just, it's a fun way to gamble. I don't think there's really, there's not that much else there. Like we have this new way that you can create a fun thing to gamble on. And that's exactly what people are doing with it. They'll do it. They'll have fun. They were probably going to gamble on something else anyway. And there are just a lot of opportunities to gamble right now for better or for worse. And I think probably net for worse. I think that In some ways, it is more pro-social, better for the world for gambling to happen on products like Robinhood rather than on sports betting apps or Pump.fun. Because if people are gambling on Robinhood, their counterparties are either investors who are market makers who are also trading with hedge funds and institutions who are trying to accurately set prices, or their counterparty is pretty much that accurate price setter. So Gambling in equity markets is basically a subsidy for price discovery. It makes it harder in the short term, makes it more lucrative in the long term, and so ends up making the market more efficient and transferring— it's basically transferring wealth from individual gamblers to, in the aggregate, to retirement savers who benefit from more efficiently priced passive assets. Whereas the sports betting stuff, it just means there are a lot of annoying ads out there and periodically people get really damaging addictions and ruin their lives through that stuff. So That said, I'm, I'm pretty sure that there will be a combination of regulatory and social response. Either the social status of gambling will decline in the same way that for a lot of, just a lot of addictive behaviors, we go through this cycle where at first not that many people are doing it. And so it's just not a big social problem. And then it becomes ubiquitous and we start to discover that some people just aren't able to handle it the way that other people can. And then we set up new norms around this. There's a really, really fun essay. By Clay Shirky about this and television, and he's using addiction to screen time. That's what he's talking about, but he uses the rise of gin in England as a case study in how this works. And his basic narrative is England had had a drinking culture of sorts for a very long time. They had beer and some people had wine. And so they had a lot of norms that kind of assumed that you have fairly low alcohol by volume. when you're drinking. And if most people are manual laborers, they're working in agriculture, et cetera, it's just not a big deal for them to have a high-calorie but very low-alcohol beer or 2 for breakfast and then another beer or 2 for lunch. Like, they need the energy, they're going to work it off, and they're really not getting drunk. They're getting buzzed at best. When you introduce gin, which has maybe 10 to 20 times the alcohol content of beer, suddenly the norm of That you can have a drink or 2 with breakfast is a very, very dysfunctional norm. So that's when you start to get this understanding of actually, if you are drinking at lunch, you probably have a problem. If you're drinking at breakfast, you definitely have a problem and you should cut it out. And they sort of developed this idea that there are kind of higher class ways to consume alcohol and they are, they're going to be more expensive. We're going to wrap them in some expensive labor and we're going to do this in particular settings. So. They ended up with a set of norms that were pretty much compatible with alcohol being available and not being just this massive destructive force. But the way they got there was by it being a massive destructive force for pretty much a generation. Gambling, hopefully we don't have 30 years of this stuff. Hopefully we develop better norms or rules sooner. My money is on the norms because a lot of the norms can be essentially developed and enforced by the gambling platforms themselves. And to the extent that they're going to be regulations written, I'm sure they would all prefer that the regulations be written by redesigning the apps in such a way that they are just not exploiting their customers and bankrupting some of them, versus having the rules written by a very reactive legislature that has a lot of really, really depressing stories of gambling utterly ruining lives. It's probably just way better for DraftKings' long-term value if they try to really put a stop to the extreme gambling as soon as possible.

Interesting. I identify as having like a latent gambling problem, which is to say I don't have it now, but if the tools were much better and it was just so low friction to do a little sports betting or something else, and I know it's basically there. like expanding or keeps improving, I could see myself getting into it. And so, yeah, you know, creating or leaving just a little bit of friction, whether it's on the product side or the regulatory side, will keep out people like me for whom, if it was just dead simple, you know, might get into it. Yeah.

And I think another piece of this is that the gambling companies are able to do things that a lot of financial intermediaries wish they could do, which is just absolutely refuse to do business with anyone who knows what they're So if you are, you know, if you're a market maker, you, you can try to avoid trading with hedge funds, but you can't actually avoid doing it. Like you, you do have to trade at the market price. Like you do have to accept this order. And there are venues where, you know, there are venues where it's easier for someone to just more anonymously make a trade. And with gambling, yeah, they can, they can look at your performance in your account and they, they know who you are. Like you had to put in your financial information, et cetera. So they, they know your name and you, It's harder for you to have fake accounts or it's, uh, it probably ends up being criminal. Um, probably ends up being some, some AML thing or KYC issue on your end if you're using somebody else's information to gamble. And yeah, if they see that you're actually good at what you do, they just don't let you bet very much. And that is one of the proposals I've heard, which I kind of like, is that we could actually separate the business into retail-facing interfaces and then a wholesale piece that has to be separate where Anyone can take either side of any bet. And all of these bets are determined by the market. So instead of one of the gambling sites trying to figure out, okay, what is, you know, what is this weird parlay bet where people are just going to mess up the conditional probabilities and wildly overestimate the odds of some extreme tail event happening? And we just know they'll systematically put money into this and always lose. If there's actual competition to take the other side of that bet, Then maybe the pricing starts to reflect what reality really is, and it just becomes hard to lose that much that fast on these really tail-risky bets. It is probably, it might actually be healthy for the ecosystem if people started to get this sense that the people who make a lot of money in sports gambling are not the people wearing team jerseys. They're people who like, they don't know the rules of the sport. They don't know anything about the sport. They just look at large volumes of historical data, create models, come up with pricing, and that's it. Now, it's really not that extreme. You do actually have to have someone with real-world knowledge of what's going on because otherwise you are just constantly on the wrong side of real-world events like player injuries or the weather or whatever else can affect the outcome and range of outcomes at a given event. People get the sense that actually the real winners are just boring nerds who care a lot about data and would happily be trading corn futures if there were a little bit more money in it, then maybe that tells them that this is a business where you're always taking the other side of some really, really sharp people and you'll probably lose.

Speaking of really sharp people, you actually mentioned him earlier. You mentioned Buffett. You read a book about Buffett. Did it update your views a little bit on how we should think about The Oracle of Omaha.

Yeah. Yeah. So this was a really, really fun book. I'm going to really quickly look up the title. Yeah. So Buffett's Early Investments, the title, Brett Gardner is the author. I highly recommend if you are, if you are a finance nerd, especially a finance nerd who's read Buffett biographies before, really, really recommend it. The first thing to say about this is I am always very appreciative when someone writes a book about business history and they include financial statements. Because sometimes they're writing a narrative and it's really interesting. And in one chapter, the company is doing $1 million a year in revenue. And in the next chapter, it's doing $75 million a year in revenue. And we don't know what happened in the meantime or when they started making money or anything like that. And this book tends to have 10-year financial statements for pretty much every company it talks about. So that's the first thing. Um, just really good move. I'm very glad they did that. I would probably pay double the price for every business book I read if it just included the financials. But the other piece is, yeah, it's just a really deep look at how did Buffett think, what was the actual opportunity. And for a lot of these, for some of the case studies, it was things I hadn't heard about before or things that were maybe historically more obscure. But for some of them, they're widely known, widely talked about things like Buffett bought a bunch of Amex right after JFK got shot in the midst of this big Amex financial scandal, which is not the company's responsibility. They got ripped off, but There was this question of would they survive or not? And, um, there was actually an interesting detail, which is a show up in other biographies, which is that, um, Amex for a while was organized as some kind of special category of company where it was not limited liability. Shareholders could be held personally liable if the company's liabilities exceeded their assets. And a lot of Amex shareholders were big trust departments, big banks, financial institutions, and they really did not want this unknown open-ended liability. So you couldn't really buy Amex shares. You could kind of buy Amex futures or like futures on Amex future cash flows, positive or negative. And that was a little scary for people. But also in the book, it digs through the financial statements. And when you look at Amex, it's actually kind of a leap of faith on the credit card business at the time that Buffett bought in 1963, because they have this core traveler's check business, very profitable franchise. Amex is kind of synonymous with this business, but also— It's a business that is not really growing that much and is fairly gradually on the way out. They have this credit card business, but credit cards are a fairly new technology at this time, and there are plenty of horror stories about banks getting into the business, losing a ton of money on fraud, and having to lick their wounds and tighten up their underwriting standards. And the core, like the main business was still the traveler's checks. So if you were investing in Amex in 1963, you were buying a combination of this static low-growth, post-growth business and this fairly uncertain startupy business that happened to have a well-capitalized parent company and happened to have some good growth prospects. And then it turned out the credit card business did actually do quite well and Buffett did quite well. So it was good to get that read and feel like this was not actually just a really obvious next thing for Buffett to do. It was a pretty hard thing to underwrite. One of the other transactions that the book talks about is Studebaker, which was a car company, had some other subsidiaries, and One of them was a motor oil additive company. And what Buffett actually did in this case was he learned what the main ingredient was, who made the main ingredient, and where the manufacturing was, and then just hung out at a rail yard in Kansas City for a while, counting how many Union Carbide tankers were showing up because that was his proxy for how much of this stuff is getting made. And that allowed him to see that this additive is actually being manufactured at scale. It's growing. And I was able to tease out the impact of that from the financials, which to me feels almost exactly like the stories I've heard about hedge fund people just trying to figure out before, before, um, Amazon broke out AWS, trying to figure out, okay, how big is this AWS thing? You know, where, what do its margins look like? Like, how can we actually track its growth? That was a hard problem for a while. And people found lots of different datasets that helped kind of give them a peek into that. Um, and yeah, it turns out that There were investors doing that kind of thing long ago. And then the last one I want to talk about is Disney because Disney was actually, you know, it is that kind of classic case where it's the transition from Buffett, who's just looking at balance sheets, adding up numbers, maybe finding some hidden asset and figuring out what it's really worth, et cetera, to Buffett as someone who is investing in quality companies at a good price. Disney was definitely in the quality company bucket, but there were lots of corporate governance issues because Walt Disney had had a very bad time in Hollywood doing deals he shouldn't have done, losing control of his company in the past or losing control of his favorite intellectual property. And he wanted to avoid that. So there was Walt Disney, the company, there's Walt Disney, the guy, there's Walt Disney, the guy's holding company, which does business with Walt Disney, the company, and does it on pretty onerous terms. So like Walt Disney, the name was licensed by the holding company run by Walt Disney, the guy, to Walt Disney, the company that you could buy stock in. And That like today, if you looked at a company like that and you're like, the CEO is literally licensing his name to the company he named after himself and he's charging a lot of money for that, you'd say, okay, one, egomaniac. 2, he's robbing the shareholders. And so therefore, 3, I do not want to be a shareholder who's betting on an egomaniac who's picking my pocket all the time. And I think Buffett just had to get comfortable with the fact that one, it was cheap enough to compensate for that. And 2, that Walt Disney was doing this kind of shady thing as a defensive move against other people's shadiness and not because he intrinsically wanted to take a company public in order to have a vehicle that he could periodically loot.

That all makes sense. Well said. And we're right at time. There's a couple other things I want to get to about Bitcoin being a strategic reserve, but perhaps let's save it for next week. Brent, this was a great overview of some of the Uh, topics you've written about the last couple of weeks. And as always, until next time.

Yes, indeed. Talk to you next time.

Upstream with Erik Torenberg is a show from Turpentine, the podcast network behind Moment of Zen and Cognitive Revolution. If you liked the episode, please leave a review in the Apple Store. Hey everyone, Erik here. At Turpentine, we're building the first media outlet for tech people by tech people. We're the network behind the show you're listening to right now. We have a slate of hit shows across a range of topics and industries, from our AI and investing cluster of podcasts to shows that drive the conversation in tech with the most interesting thinkers, founders, investors, and influencers like Econ 102, with Noah Smith. We're launching new shows every week and we're looking for industry-leading sponsors. If you think that might be you and your company, email me at erik@turpentine.co. That's E-R-I-K at turpentine.co, and let's partner together.