Odd Lots - Jerry Neumann on the Problem With Investing in AI Right Now

Episode Date: November 12, 2025

AI has made a lot of people fabulously wealthy. But sorry, it's probably not going to be the thing that makes you rich. And if history is any guide, we don't even know who the real AI winners are going to be. That's the thesis from longtime Venture Capitalist (now retired) Jerry Neumann. Earlier this year, Neumann published an article, "AI Will Not Make You Rich," putting the AI boom in the context of previous technological revolutions, such as the shipping container. He points out that a lot of the companies that were early to shipping containers didn't make much money, and that the real winners were the new businesses that emerged later and took advantage of the shipping container to build new business models (think about the likes of Walmart or Target). In this conversation, we talk about why it's so hard to invest in technological revolutions, where we are in the cycle, why he's getting out of VC, and when the big opportunities will eventually emerge. Read more:SoftBank Sells Nvidia Stake for $5.8 Billion to Fund AI BetsAI’s $5 Trillion Cost Needs Every Debt Market, JPMorgan Says Only Bloomberg - Business News, Stock Markets, Finance, Breaking & World News subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at  bloomberg.com/subscriptions/oddlots Join the conversation: discord.gg/oddlotsSee omnystudio.com/listener for privacy information.

The news doesn't stop on the weekends. Context changes constantly. And now Bloomberg is the place to stay on top of it all. Hi, I'm David Gurra. Join us every Saturday and Sunday for the new Bloomberg this weekend. I'm Christina Rafini. We'll bring you the latest headlines, in-depth analysis, and big interviews. All the stories that hit home on your days off.

And I'm Lisa Mateo. Watch and listen to Bloomberg this weekend for thoughtful, enlightening conversations about business, lifestyle, people, and culture. On Saturday mornings, we put the past week's, events into context, examining what happened in the markets and the world. That on Sundays, we speak with journalists, columnists, and key political figures to prepare you for the week ahead. Join us as soon as you wake up and bring us with you wherever your weekend plans take you. Watch us on Bloomberg Television. Listen on Bloomberg Radio, stream the show live on the Bloomberg business app, or listen to the podcast. That's Bloomberg this weekend.

Saturdays and Sundays starting at 7 a.m. Eastern. Make us part of your weekend routine on Bloomberg television, radio, and wherever you get your podcasts. Bloomberg Audio Studios, Podcasts, Radio, News. Hello and welcome to another episode of the Odd Lots podcast. I'm Joe Wisenthall. And I'm Tracy Allaway. Tracy, our colleague here at Bloomberg, Ed Harrison, had an interesting newsletter today. And it's actually something I've been thinking about a little bit lately,

which is that for all the talk of the AI boom or the A bubble driving the stock market, there's no A-I-Pure plays, really, that are publicly traded. Like, InVIDIA is probably the closest, but three years ago, people were excited because they were mining Ethereum. Before that, it was like video games. You know, this was only an AI company in people's mind for since late 2020. Google still is, you know, they're all investing in a ton. There's actually no, like, no AI company that people are excited about in the public markets. I mean, I think it's true.

Here you have this thing that a lot of people would say is revolutionary technology, right? But you kind of have to decide if you're going to understand. invest in it? Is it going to be like upstream or downstream? And there doesn't really seem to be that much pure play, as you say. As our colleagues, Sid Verma, might like to say, people are investing in the picks and the shovels, you know, this gold rush. Now, I don't think he actually said that it's just a million other people's head. But it is. It's picks and shovels, right? Stop making fun of completely reasonable commentary. That's what I say. You know, it turns out, incidentally, picks and shovels have been great. You know, you could have bought Caterpillar or you could

bought some old school HVAC company that's providing cooling or heating or whatever and made a ton of money. So actually, it turns out, at least for the last few years, all those awful cliches have actually been big moneymakers and I should not make fun of them. Well, here's the other thing I would say. It does feel like everyone kind of agrees at this moment in time that there is froth in the market. Maybe it's not a massive bubble, right? But there's some froth. And everyone is kind of admitting or saying that you're going to have some companies that emerge as big. winners, much like the dot-com era, and then a bunch of companies that, like, actually end up being losers.

And I think, again, like, that is consensus at this point. But it doesn't really necessarily translate into actual investment, because, of course, the trick is actually picking the winners and losers in the market. I don't know. It just seems like a weird point in time where people are like, oh, yeah, AI is great. But we all know that some of these companies are going to be massive losers, right? Right now they're all kind of being treated as winners, which is the other ones. Yeah. Anyway, it's a very weird time. We've got to do more episodes on this because it is sort of the central question on whether we're just talking about the market or talk about the economy, et cetera.

Someone I've wanted to talk to for a long time. Earlier this year, he wrote a essay for Colossus called AI Will Not Make You Rich. It came out in September. It seems like ages ago. It's very disappointing because I think a lot of people really are hoping to get rich on AI. So this is a very unwelcome message. It's also very much a sort of core odd lots thesis because in the essay he compares and contrasts AI with containerization, which is another favorite topic. So let's just get to the guest. Someone I've wanted to talk to for a very long time, someone who literally is the perfect guest, long time VC started venture investing in 1997. Also a professor at Columbia Business School, we're going to be talking to investor Jerry Newman. Also the co-author of a recent book, Founderverse Investor, the Honest Truth about venture capital from startup to IPO. Maybe he'll tell us when we'll see some of these. And the bringer of excess Halloween

candy. So he gets, he gets brownie points for that too. Literally the perfect guest. Jerry, thank you so much for coming in. Thrill to finally have you here. Thanks. I'm glad to be here. What does that mean AI won't make you rich? AI has made people super rich and it's making people rich every single day. You know, as an old mentor used to say, money's not money till it's cash. Okay. So is anybody really rich yet? Oh, come on. I mean, Jensen Wong bought an entire bar in Korea. He's like bought beer and fried chicken for everyone. He's right. I think it's smart to cash out early. Okay. That's what he's doing. Okay. Let's see more about this. Would he be selling his stock now? I mean? Okay. So what do you mean? Talk about this. Because you obviously have a lot of experience. Actually, that brings another line of question that I want to get into. But what does that mean? What does that mean? So look, I believe that AI is a revolutionary technology. Okay. I'm going to put that on the table. Which is important to say because not everyone agrees with that. So that's, yeah, totally. You know, I'm on blue sky and nobody agrees with it. But I do. I think so. But there's a difference between value creation and value capture. So even if AI creates a lot of value for society, who's going to get that value? Is it going to be the early investors? Is it going to be the core, you know, the foundation model companies? Is it going to be consumers? You know, they think that's the question people need to ask.

I actually broadly agree with this thesis when it comes to AI, but maybe just to clarify the idea. here. Compare and contrast this current AI cycle with maybe previous technological breakthroughs. And, you know, I mentioned containers. I think a lot of people aren't used to thinking about boxes as this major advancement in technology, but at the time, they really were. Yeah, I mentioned containerization, and most of my peers think I'm talking about Docker. So I'm talking about shipping containers, right? The big boxes they put on ships, and then they can move from the ships onto the, you know, rail cars and now onto the back of trucks. And this was a revolutionary technology. I mean, it changed everything about the way we live.

I don't remember, I'm probably a little older than you all. But when I was a kid, my grandmother used to send up oranges from Florida at Christmas time, right? Because they were rare. You couldn't just go into a grocery store and buy them. Now you can buy oranges anywhere at any time. People don't really realize how much our lives have changed because of shipping containerization, because of these global logistics and the globalization of shipping. Now, this is revolutionary technology.

Who got rich from it? I mean, generally, if you look at the 1960s, say, how many people became wildly rich from technological innovation? Can you think of anyone? Because I've been asking this question for years. There are people who got rich in media and whatnot, but there wasn't a lot of technological innovation that made individuals rich. Didn't tech just start like 20 years ago? Did they have technology back then? Right.

I mean, I think this is the thing, right? So we talk about computer technology, the information and computer technology revolution as technology. but obviously this has always been technology. But only at certain times in this technological cycle do people seem to make money as investors and as inventors. Explain more, though, because I mean, I could argue that Maersk or someone like that got pretty rich off of containerization. Like maybe it took a while, but even though the shipping industry is highly cyclical, when they are in the boom period, they make a lot of money. Sure. I mean, the existing shipping companies got very large and made a lot of money.

They got larger and made more money. but is mayor, you know, who made money off of investing in marriage? You're talking about completely new entrance. Yeah. So just, I mean, as background, I'm a venture capitalist, right? Or have been a venture capitalist for a long time, recently retired. And I think about people investing and making money, new companies, inventors or entrepreneurs making money, not the existing incumbents making money.

And I think that people will make money on AI. It might be Microsoft making a ton of money on AI. It could be AI. You know, when containerization, shipping containerization came around, Sealand was the instigator of this. And the founder of Sealand made money, primarily because he sold early, right? He sold Sealand to RJR Nabisco, or sorry, it was just RGR at the time. And they thought they were diversifying, which is the big thing then, paid him a lot of money, and then they drove it into the ground. So what did Sealand do? What was, did they, I actually don't, I'm not familiar with this

company at all, which I think kind of speaks to your point. But what was Sealand? Yeah. So it was a truck, it started out as a trucking company. Okay. And the founder of Sealand was a trucking entrepreneur. And he said, it's silly. You go into a port. Your truck sits around all day while, you know, the Longshoreman put a cargo net into a container ship, load everything into it, pull it out, unload it, and then reload it back into your truck. This is not efficient. And this obviously is it's an obvious idea, right? Just put it all in a box. You can then put that box on a truck. The best ideas are always the obvious ones in retrospect. But the problem with it was it was a systems problem, right? The Longshoremen didn't want it. The ports didn't want it. The port authorities. didn't want it. The politicians didn't want it. Nobody wanted this to happen because this sort of enormous change would put a lot of people out of jobs. It would upset the existing order. And it did. I live in Hoboken. And in Hoboken, there's a lot of peers that nobody uses except to go running on now because back in the 60s, it was a longshorement town. And when I moved there in the early 90s, it was empty.

It was starting to gentrify, but because all of those people had lost their jobs and moved out. And I suppose, like, even in the case of Marisk, and I'm sure, you know, they obviously, have made a lot of money because the explosion of global trading volumes and containerization is part of that. Like, it wasn't overnight wealth, right? It wasn't like, it was not, it was not, people got richer, but it was not like some bubble get rich quick thing where they suddenly cashed in on the new thing. Yeah, I mean, I suppose whoever owns Maersk may have gotten richer.

Yeah. But it's not like you're going to look at the Forest 400 and see all these shipping magnates who became suddenly, you know, enormously wealthy. There are a few. Wait, okay. So if I think about a box. you know, part of the... We just keep this whole conversation on boxes, actually.

They get to AI at the very end. Well, one more question. And then we will be able to discuss AI. But I think about a box. And as you say, it's sort of an organizational structural problem. Like the box itself is not the huge technological advancement necessarily. So what exactly was it about containerization that prevented it from being disseminated, I guess, to new upstarts or new companies?

that could actually use that technology. Well, it was. So actually not sure I understand the question because containerization became widespread very quickly. Right. But what I mean is like why was that value seemingly captured by incumbents versus startups? Right. I think because it disseminated so quickly, right? It was an obvious idea.

Everybody who saw it said, okay, we need to do this, right? Everybody who is already in the business said, if this is going to happen, we have to do it. We can't be left behind. We will be left behind if we don't do it, which, you know, I think was also obvious. So the reason nobody else did it first was it was hard to do. It was hard to make happen. And technology has always come in these technological systems if they're worthwhile technologies. Right.

So the personal computer didn't change the world on its own, right? It changed the world alongside the Internet, you know, alongside a bunch of technologies that formed a system. So the hard part here was building the system, not the individual technologies. And this is true, I think, of computers as well. You know, the first microprocessors weren't considered revolutionary. Intel didn't consider the 4004 revolutionary. They considered it evolutionary. The engineers have said this.

And it wasn't revolutionary until people put it to use in ways that they didn't anticipate. Actually, can we go? I didn't know that. Like, I hadn't really thought about that with Intel, that at the time, it didn't feel to them that it wasn't a revolutionary technology. That's sort of mind-blowing. It is, right? This is, I think, from Michael Malone's, the Intel Trinity, the book. You know, he interviewed a bunch of Intel engineers.

And he said, you know, like, they thought they were building a better chip set to build pocket calculators or, desk calculators, I should say. So they had a client, BISICOM, who wanted to build a better desktop calculator. Calculators were big back then in 1970-ish. And one of the engineers said, well, why do we keep building custom chipsets for each different calculator? Why don't we just build a chip set that we can customize the software and change what it does? And Intel is kind of like, eh. And BISICOM was like, okay, we'll pay for that. And then BISICOM actually tried to back out, and they gave the rights back to Intel, so Intel owned the rights to this 4004. and then they started selling it.

And it wasn't Intel. Intel believed at the time it was going to be maybe used for dedicated hardware, you know, hardware controllers, that kind of thing, not by consumers. So it wasn't until people on the outside said, hey, you know, I love these IBM mainframes or these deck mini computers. I'd like to have my own, but obviously nobody can afford that. Why don't I just try to build my own, right? So it was these kind of outside inventors, these permissionless invention.

And then it really, the real revolution didn't happen until this, Everybody was like, oh, Intel. It was the 6502 where the price came down so dramatically that, you know, Steve Wozniak could walk into a computer fair. It gets them for free and go home and build a personal computer. That's crazy. I wonder who made the chips for the Commodore 64 computer that I had. They were 6502s. Those were Intel.

No, they weren't Intel. They were Moss Technologies. Oh, got it, got it. Those are the cheap ones. Yeah. Wait, what year was that when you had? Well, I think I like learned some basic and did some coding on it. I would have said maybe 1988, 1987, somewhere around there, made a few.

I miss those days. It's crazy that I didn't be, can I just say? Sometimes when I think about my life path, like, how did I not end up like a tech guy? Because I was like very into math. I was like one of those people who had a computer when I was six or seven. Wait, in 1987 you were coding. You were seven years old. Yeah, yeah.

I got that my dad got me this magazine that just, it was very crazy. They was literally just sent you pages of code, and then you just typed it in and you could like make a video game. I was doing that at 7. I could be like one of those like going. His parents gave him a computer when he was 7. Anyway. It's not too late, Joe. It's not too late.

Yeah, it isn't too late. Well, maybe it is too late because now we have AI doing all the coding, right? This is Caroline Hyde. And I'm Ed Ludlow inviting you to join us for Bloomberg Tech, a daily podcast focusing exclusively on technology, innovation and the future of business. Every weekday we bring you the top headlines from the world's biggest tech companies. From finance to defence, AI to entertainment and from startups to the magnificent seven. We highlight the latest stories of the people and companies pushing the tech sector to new frontiers and the politics that shape global tech markets. We do this all every weekday.

Then bring you the most important conversations and analysis in our podcast. Search for Bloomberg Tech on YouTube, Apple, Spotify, or anywhere else you listen. Join us every afternoon on your commute home and stay ahead of the tech news cycle. That's the Bloomberg Tech podcast. I'm Caroline Hyde in New York. And I'm Ed Ludlow in San Francisco. Subscribe today wherever you get your podcasts. Okay, just so I understand, when it comes to, I guess, the advantages of the incumbents to actually monetizing new technology or benefiting from new technology.

Is the moat around their business the network and their sort of role in the network? or is it the vast amounts of cash they have and the ability to sort of roll out massive investment to capture that value? I think it's the latter, right? So anybody can build a foundation model, right, if you have the money. I mean, the technology is not mysterious. It doesn't feel like the technology is really changing very quickly anymore. And of course, I don't have insight into what's happening inside of Open AI. But looking at it over the past couple of years, it's the same thing, but slightly better.

It's evolutionary now, right? The first part was revolutionary, and now it's evolutionary. So if you wanted to build one, you could build one. And I have friends who were running them on their laptops very slowly. But it's possible. So now the question is, do you have enough cash to build the data centers, to buy all the chips, to build something that is large enough that when you train it, it does something useful? And it's just a question of having that, the authority in the market to be able to raise that money.

By the way, speaking of ideas that were sort of really obvious that took a while, I'm always blown away that like how long it took them to put wheels on luggage. I don't think anyone like got super rich on that. But when I was a kid, I remember like we had these big suitcases. And that's the most obvious thing. It took a while. Anyway, I know, I don't think anyone got it. Technology is still not perfected, as you know, because you've been in airports with me and the

wheels on my luggage are broken. But I don't think anyone got like super rich off of wheels. That just seems like an obvious. That was just sitting there. You know, I want to jump ahead actually in the conversation a little bit because I don't to forget this point, but this is something I've become a little obsessed with, which is, I've been meaning to ask a VC about this, which is that there seems to be this blurring of private and public markets in various ways, retail participation in private markets, et cetera.

For the VC, in my mind, I feel like the exit was the IPO or the acquisition, right? So you buy, do VCs these days have to think a little bit more about market timing and selling early? So someone who is an early investor in Open AI or whatever, you know, in the past they might have just held or then sell it the IPO or the acquisition. Probably not going to happen in OpenAI's case. They're too big to be acquired. But do VCs, and in your experience these days, have to think a little bit more about this idea of selling early timing the exit? Well, I think VCs always had to think about timing. You know, I've done pretty well in VC and I attribute it entirely to being lucky at starting investing at the right time.

So the first time around I was starting in 97, which any any could make money. And the second time around, I started in 2007, which again, it was just, or 2008, it was just an easy time to buy in. But I mean, the timing of the sale. Like, is that something that where in the past it might have been automatic to exit? Now, that was not decayed. Now, I wrote this thing about VC in the 1980s a long time ago. It's on the blog. You can find it.

And the thing, because nobody talks about the 1980s, right? There was plenty of VC in the 80s, but nobody talks about it. I have people talk about the 60s and the 70s and the 90s. So I was like, all right, what happened then? One of the interesting things about it is there were the IPO windows then where, you know, 1983, the IPO window opened, a bunch of companies went public and then it closed again. And you can see it in the numbers when people went public. So people always had to think about timing.

The IPO is obviously the best exit because you want to sell to the greatest fool and nobody's greater fool than the public, right? So you look for the IPO window to open. When you can't, you have to sell it to somebody else. you know, VCs have this problem of their limited fund life. So I look at my portfolio and I'm a really early investor or have been a really early investor. So I'll look at companies and say, oh, I invested 10 years ago, they're going to have to sell. You know, it's the, so people look for the IPO window, but if they can't find it, they have to sell somewhere else.

Well, how much of the money flowing into AI startups now is just the expectation that a bunch of these little companies are eventually going to get bought by larger incumbents? and basically you're going to have consolidation and you will get that exit. I don't think anybody can predict when the IPO window opens. I mean, I wish I could, but I don't think, I've never seen anybody even say they could predict when the IPO window opens. So I think a smart VC invests in a company that can become self-sustaining to some degree. And then you wait for the timing to come. You don't invest and say, I'm going to flip this in three years. Now, you know, the other problem is VCs don't invest in all these IAA companies saying a bunch of them are going to become valuable.

They invest saying one of these is going to become value. Right. The lottery ticket theory. Yeah. The power law. Speaking of the IPO window, I'm never totally satisfied by a lot of the explanations for the drop off in IPOs generally. Do I know there was that law passed in 2001 or what was it? Sturbanes-Oxley.

Yeah, Sorbanes-Oxley. And I get that law. That law. That law that everyone hated for a really long time. I don't know. It doesn't, but then you see like, you know, in 2021, there was like a billion. garbage companies went public via SPACs, et cetera. How much is it about, okay, there are some disadvantages to being public versus there's just so much more private capital out there such that the imperative to perhaps ever go public and it's liquid and their rounds and like, what do you attribute?

There are these big companies that are private strife, but Open AI and Anthropic that choose to stay private. What do you think the main reason for that is? Well, it's because they can, right? I mean, being a public company is no party, right? I mean, it kind of sucks being a public company. What is it about it? What sucks? Well, you have to tell everybody what you're doing every three months.

Yeah. And then they come back and complain about it. So, sorry, I'm being a little facetious. But it is, it's hard to be a public company. Everybody, you know, anybody who runs a public company will tell you they spend a lot of time being a public company if they're running the company. So that's taking away from actually running the company. Yeah.

I think the flip side is you're liquid. And that's, you know. So if you can stay private, why wouldn't you stay private? or if you can go public and retain control of the company, you know, like Henry Ford or Mark Zuckerberg, then why wouldn't you do that? But I think it's because there is so much late-stage money, this isn't necessarily a good thing. It's because there's so much money out there that's not being invested in more revolutionary technologies earlier. With all this money being invested in AI, you may wonder if people are still going to want to invest late-stage stripe. Or the analogous stripe might be making money now, I'm not sure.

I know you brought up previous historic analogies like VC in the 1980s, but just to focus on the one that everyone else seems to be focused on at the moment, which is the dot-com bubble in the early 2000s. What are the key differences you're seeing in terms of the VC and financing environment now versus 20 or 25 years ago? I think the key difference is that most of the money is coming from people who aren't looking for much risk. Right? So, I mean, Open AI is primarily funded by bigger companies, right? Most of their money is coming from large companies. What happens if Open AI gets hit by a bus, right? So Microsoft's had a bunch of money. A whole bunch of big companies are out a bunch of money. I don't think much happens to the economy, I think which is different than in the dot-com bubble where a lot of consumers were in it. A lot of consumers were in it leveraged, right? The buying a margin or whatever. A lot of people had options, right, people, employees, and they were spending the money from their options before they were liquid. You know, it was, there was a much, I think it was a different dynamic with the economy. The wealth effect was a lot bigger. I don't know.

This is, I think this is a contrarian take on your part because you hear a lot about, I mean, in two dimensions, you hear a lot about the direct wealth effect from people's exposure to the stock market, which AI is a big part of the story. And then you also hear about, of course, the sort of real economy effects through all of the spending, which we will get into on, you know, the data centers and the catapult, the turbines for the gas generation, et cetera. I think many people would say there is a lot right now riding on the health and the sustainability of this particular sector. So I think we can separate the companies like Microsoft and Nvidia. Are they overvalued because of this? Maybe. Does it make a

huge difference to the economy? Probably not. I don't think so. And the companies who are, spending money on infrastructure, like building data centers, building power generation plants, those things, I think, are probably overbuilt, or not so much overbuilt as they are built. And I think in 10 years you're going to have a lot of extra compute, a lot of extra power generation, and people will be able to use that for other things. It'll also drive down the price of just using AI, probably. Yeah, Jevin's Paradox, bro, is going to be out back and forth. All right, so AI, where are we, you talk about cycles, and I think you use the word that

And eruption? What was the word you used? Tell us how you see cycles and what cycle we're in, right? Right. So a lot of this is based on Carlotta Perez's work, which is pretty familiar to the venture capitalists in your audience, I'm sure. She wrote a book called technological revolutions and financial capital, where she explains the dynamics behind the Kondratia of waves that the Schumpeter talks about. So she has a theory about why these happen. And you look at the Industrial Revolution, the second Industrial Revolution. you can see these waves of technology, technological systems happening through the economy where they kind of start out, they grow really rapidly, and then there's usually some sort of adjustment, some sort of bubble bursting, and then things kind of level out and then start

to plateau, and then a new one starts. And this is, people have noticed this since at least, Candratiav, in 1926. She has a mechanism for explaining it, and her mechanism has these four phases. The first phase is eruption, which she spells with an eye, which I think is actually in the dictionary as a word. I don't know what the difference is between that in the eruption. But it is the part where... It makes you sound smarter.

Right. So, yeah, keep saying that. It is the start. It's when people have invented something and it is starting to catch on, but it hasn't caught on yet. There's a lot of people saying, is this the future, is it not? You look at personal computers in the late 1970s, early 1980s, and maybe even before IBM got involved. And people didn't think personal computers were. Some people thought they were their future.

And, you know, if you look back at computer history, everybody talks about the people who did think. that. They don't talk about the other 99.9% of smart people who said they weren't. This is the eruption phase where there's a lot of uncertainty about where this technology will go. It's starting to build connections to other technologies, starting to attract money, attract smart people because it's interesting and it might actually change things. So this is the beginning. And I think the connection here to AI is people wonder if we're in the eruption phase of AI or not. Is this the start of a new technological revolution? So which phase are we in?

I think we're not. I think we're in, I think this is the end of the information computer wave, the end of the computer wave, right? I think it is, this is the culmination of the computer wave, right? Why did we build computers? We build computers to help us think better, right? This is what they're for. They're knowledge machines. So now we've kind of reached the natural end stage of what they do. They're smart machines or smarter. So I think this is not a new technological revolution. I think it's the end of the old one. And this is why I compared it to containerization, because the previous wave was automobiles, mass production, and starting in 1915 or so up until 1970 was the previous wave.

And containerization was squarely at the end of that wave. And it was really kind of pulling together the technologies of that wave into something that increased productivity. Right. Like the final step of the global trade and, I guess, mobility revolution. Yeah, exactly. Okay. So how do you react as a angel investor?

Are you at the stage where you're looking for, I guess, the downstream winners, like the companies that are going to be able to apply or use AI most effectively? Or how are you actually deploying all these thoughts in terms of your own investment strategy? I retired. Okay. So you're standing it out. And I said, look, how am I going to invest in foundation lines? Right? I don't have a billion dollar fund.

I don't think that, you know, if you look at the big winners from the early big winners from globalization, the IKEA's, right? I mean, IKEA was a Scandinavian company until containerization, and then they became a global powerhouse, a hugely successful company, but they didn't need outside money. You know, Ingvar Comprad, I think he borrowed like a couple thousand dollars to start that company or to get that company to buy some inventory. He never took outside money. You look at, you know, Walmart, which had been around already, it was an incumbent and used this kind of globalization to bring a lot more variety of products to the stores. They didn't need outside money to do that. Yeah, I guess if you're IKEA and suddenly you're flatpacking everything and shipping it in containers and that's your big innovation, it's a money-saving technology, right? So you don't actually have to raise new capital in order to flat-pack everything.

Yeah, exactly. Okay. They were already flat-packing. Yeah. Your mention of the Walmarts and targets of the world was like in your essay, it was like it's sort of very light bulb thing. It's like, yeah, I don't know. I guess we'd take them for granted. But they're clear like massive containerization winners. The scale that they see. exceed ad is impossible to fathom in some prior era of Walmart is a logistics company. Change my mind. Literally. And many of you literally, literally, literally, literally that. But it doesn't feel like with AI that the equivalent has emerged yet, right? We're still at the age where people are building the container deploying. But the company that exists and is massive that couldn't exist prior to air, like does not feel, we haven't seen that yet.

Well, you've got to be a little patient. No, no, I get, no, I get it. No, seriously. So the first container ship sailed in, I think, 1956. Okay. So when did IKEA become a global powerhouse? It really wasn't until the 1970s that they started to expand that way. This is really important that it can really take a while.

Where would you expect it to show up? Like, what industries would you expect? Because obviously retail existed for a long time. Furniture existed for a long time. Then you get these behemoths. Are there industries that you think are ripe to produce? very tortured analogy, the IKEA of the AI way. Well, I think they have to be knowledge-intensive industries.

All right. I mean, this is what AI is doing, what it is making more efficient. I mean, I think there's people that ask me, like, well, then what should I invest in? Yeah, tell us. Yeah, this is, you know, I think, well, so. Give us the answer. I've been thinking that myself. But the answer is really that as an investor, I don't decide what to invest in.

I evaluate opportunities that come to me. And so I have built a box in which you can evaluate opportunities, right? They have to look like this. They have to look like an IKEA. And if IKEA came to you and said, I needed money at that time, you should have said, okay, I can see how shipping containerization is going to make you a much larger company where nobody else seemed to see that, right? Certainly the furniture makers in North Carolina didn't see it.

So I think this is the box that you evaluate things in. And as a long-time investor, I'm used to evaluating and I try not to come up with ideas. That said, it has to be a knowledge-intensive industry. And I think something that I said in the essay, which I wish I had said more about, was the companies that tried to use shipping containerization to cut costs so they can increase margins did poorly. Oh, this is key. Same more.

Whereas the companies that use the efficiencies and passed the efficiencies onto the consumers so that they could become larger, became larger, right? I mean, you look at these people saying, oh, we have the AI. We're going to fire people. I mean, that's, I think, is the exact wrong move. And I think it's probably just every new thing comes along. People are like, oh, we're going to fire people. You know, it's just an excuse.

But if you're firing people because of AI, you're doing it wrong. Right? You should be using AI to say, I can use my people to do more. I can grow my company. I can vary my products. I can take more market share. So the value goes to the consumer, and I guess you capture the value by selling more, right? More knowledge.

Yeah. I mean, I think Walmart never tried to maximize margins. Right. It was volume. Yeah. Okay. On April 4th, 2023, around 2 in the morning, A man was found stabbed multiple times on a sidewalk in downtown San Francisco.

Hey, we did this to you. What happened next turned the story into a political firestorm. Reports have identified the victim as Bob Lee, the founder of Cash App. From Bloomberg Podcasts, this is Foundering, the Killing of Bob Lee, beginning April 16. From a consumer standpoint, why do I care if your workflow internally at your company? is become more efficient things to AI. Either the product is better or new or something. You know, I was thinking about this.

I don't mean to pick on anyone. But it's such you made it like open door meme stock. They had a new CEO. He sent out this memo. He's like, everyone has to do, start using AI more. It's clearly press release in the form of an internal memo because he got a, I think he tweeted. Waving a flag going, we are using AI.

And it's like, yeah, but like, does the economics of buying homes via whatever get better? like, does it actually make the business better? This strikes me as very interesting, this idea that, like, it's not good. It's just not that exciting, especially any sort of customer-oriented company, which I guess is all companies, the fact that AI has become part of their workflow. It's the electric knife effect, right? So I don't know that effect. We, I just call it that.

But, I mean, electric knives were one of the fastest growing consumer products in the late 1960s. Because people are like, oh, we are electrifying things. Like, let's electrify the knife. And within like three years, they were in some massive percentage of households. Like 80% of American households had an electric knife. And things like, you know, the blender didn't get adopted that quickly. But who has an electric knife now? Right.

I think people take whatever it is the technology is and say, this is everything. I mean, back in, you know, the early 2000s, late 1990s, every company was an internet company. Right. Do people walk around saying that's an internet company now? Are you an internet company? I mean, everybody's an internet company. Everybody uses it. It's the baseline.

And that's, I think what a revolutionary technology does is it becomes part of everything. So, yeah, you're right. Consumers don't care that your lawyers using AI. They wanted to make sure that they have good legal advice. Yeah. Our producer just messaged Dash saying that he had an electric knife. Oh. Does he have one now? Probably not in the 1960s, though, Dash, do you have one now? My mom might still have one.

In case that didn't get picked up on the audio, his mom might still have one. His mom might still have one. Dash, will you bring it in for us? and we can use it as a prop on the table. I'm just imagining, like, going to a restaurant at the time. And it's like, oh, we slice your bread with electric knives. Like, how unexciting that would be from a consumer standpoint. I don't care, you know. Yeah, I guess that's true.

But, okay, but things are different from a shareholder standpoint, right? And one of the reasons we see when companies say that they are using AI, the share price goes up, it's because shareholders expect all these easy cost-cutting gains. Do we have any indication that investors are actually going to be patient and wait for, I guess, the value to spread to the consumer? Or are they just going to demand basically these fast cuts? Well, and I mean, I'm not a stock market investor. Yeah, I know you're private. I'm going to make fun of stock market investors, which is, you know, I think they have a very short attention span.

There's a great quote in that article about VC in the 80s where the Wall Street Journal said, you know, there was beginning of the 80s, there was a craze in anything that ended with in, Onyx, right? Oh, and I-C-S. Right? It was people were, right? But it was a short-lived one because it didn't really deliver and then people were like, all right, let's move on and do something else or invest in something else. So funny. If I were writing a fiction about the 80s, I would immediately, I would say like applied photronics or something. Yeah, right? Yeah, I hadn't thought about that. That's amazing. So I think they want results, but I think it's too, people are impatient. I mean, I was kidding, but I'm not kidding. People are impatient. They want to see results today. I don't think AI is going to show you the real revolutionary results for a decade.

Taking AI and just retrofitting it onto the way you do things now is only going to add a little bit to your efficiency. You have to actually re-engineer everything around this, change your processes, hire different people or train people, and then you're going to see big efficiencies. You've got to have prompt training in schools, right, in high schools or something. Well, we were talking about this with Tyler Cowan earlier, and he had, you know, a similar point. Are big companies, setting aside, okay, like, yeah, it's not very exciting that a big company is able to marginally reduce their workforce. I think a lot of those are fake.

I have a feeling a lot of these companies are going to end up having to hire people back when these things don't work. Or they had nothing to do with AI and they just wanted to do layoffs. Can legacy institutions, like, there's some, like, inherent roadblock to the degree to which legacy institutions can incorporate on AI? or is this the type of thing where it's just going to be entities that didn't exist before becoming really big household names? No, I think it will be legacy. I mean, I think, you know, Walmart was legacy. IKEA was legacy.

It may not be the names you expect, right? Sears and Woolworths didn't really benefit from shipping containerization. They ended up going out of business. I think you have to have the right attitude of how you're going to utilize the efficiencies it brings within your business. Again, to grow your business, not to, you know, grow your. CEO's salary, right? You have to be spreading this to the consumers for your company to be successful. All right. In the intro, Joe made fun of Cid saying something completely rational.

Well, Sid didn't even say it. Yeah, I know, I know. But this is your joke, right? It's an internal joke. Right. Right. Okay. So I'm going to ask the cliched question in that theme. Are we in a bubble? Can you define bubble? No. No. I'm sorry. That's a not fair question. I think the question should be asked liberally. It's like, I don't. Are things overvalued? Are things overvalued? Yes, but there's a difference between things being overvalued in a bubble, right? And I think things are overvalued. And I think there may be a infrastructure bubble. In the article I wrote for Colossus, there's a chart of container ships being built, of shipping, ships being built.

And you can see this huge rise right after containerization started for a bunch of years. And then it dropped back off because now people had their ships. But everybody had to get in at the same time. Everybody had to go order ships. A ton of ships were being built. And then the CAP-X, you know, and then they were like, okay, now we have ships. It's a little different with chips because chips aren't going to last, you know, hour or along a container ship lasts 30, 50 years.

But I do think that you're not going to need as many chips in the future as you are buying today. Really? Well, you know, you're going to have more use of AI, probably. There's going to be more use. But I also would think they would become more efficient in compute. That's just the history of compute. Does your definition of a bubble, does that have to include a, a buildup of leverage of some sort?

Well, I lived through the dot-com bubble. So, yeah, I would say so. Okay. And you're not seeing that right now. Well, because I think a bubble popping has to hurt. And if Microsoft loses a billion dollars, that doesn't hurt. It doesn't hurt any. I mean, I'm sure it hurts.

Whoever's invested that money at Microsoft, but it doesn't hurt societyally. Which is everyone who has a 401K. Well, I guess. But, I mean, really, Microsoft loses a billion dollars. Yeah, sure. How much does that affect their stock price? And even if the stock went down 10%, that's not a, it's not like the dot com bubble, right? When that popped, people were laid off. The economy went to a recession. I mean,

it was pretty deep recession. It took a while to kind of pull ourselves back out of that. Talk to us more about the dot-com bubble. I try to bring it up in every conversation because that's when I got interested in markets. I did a little day trading in those days when I was in college, et cetera, and I just remember that period very fondly because I was young. What do people get wrong in their memories of the dot-com bubble? So here's the thing I remember most about the bubble. I was a corporate VC. I worked for a big company here in New York, Fortune my friend of company. And one of the companies I had invested in was a public company and they were raising more money. This was in January, right before the bubble popped, right?

The peak was in March 2000. Then it was January 2000. So the company was selling stock. They said, hey, you know, if you want to buy some more stock and our company, we'll sell to you without the underwriter discount or before the underwriter discounts. We'd get it at a 7% below the market price. And I went to the CFO of this giant company. and I said, hey, we can get a good deal on this stock. We can get 7% off, right?

Who doesn't love a bargain? And he said, well, do you think the company is worth that price? I said, no, nowhere near. He's like, so why are you buying it? I'm like, oh, I guess that's a good point. He's like, so why would we still own it? And it's like, that's also a good point, right? Which, why aren't we selling this if it's overvalued?

I mean, the thing that people forget is everybody knew it was overvalued. They were all just waiting for it to go up more before they sold. And he said this and I'm like, yeah, I guess better to sell early than late. And we ended up selling that entire position, which luckily paid for the whole portfolio before the bubble popped. This is a, I mean, I think this is, there's sort of the difference between John authors had a really good newsletter. I think about a year ago, which is that. Well, you're just promoting all the newsletters today. I'm doing, I'm doing my job to Bloomberg.

But it was basically like, you know, you get these situations like dot com where everyone said, this is massively overvalued. We all know, we all, it's ridiculous. Then you have bubbles that are more like the housing bubble in which I don't think on any sort of traditional metrics the banks were overvalued. I think probably the PEs are probably normal. It's just that the earning stream was entirely unsustainable. And I guess that's the question with AI. Like, I don't know. In video is probably expensive.

But like I don't think people think it's like crazy stretched on PEs. It's more the question of like, is this chip demand at this pace sustainable? like are the earnings estimates realistic? Right. Well, I mean, the economist said the housing, housing was overpriced in 2005, right? The homes were, yeah. Yeah. The banks, you know.

But the banks got obliterated, and it was not because the ratios of the banks were completely out of whack because it's just that the profits could not be sustained by any stretch at that point. Anyway, I think it's an interesting distinction. Well, I think there's still an open question with AI about the network of relationships that are sort of driving a lot of this business. Like that's where I would see some of the maybe 2007-2008 analogy actually being true, this idea that like you have this whole system of funding with banks

that's keeping the whole machine going. But when the collateral that's underpinning that system suddenly loses value, the whole thing falls apart, you could maybe make a sort of similar argument for AI where you have this network of companies that are sort of investing and selling to each other. if the value of the underlying asset, which I guess would be compute in this case, starts to fall really precipitously, the whole thing kind of collapses. But I'm stretching that. It's amazing to me how much of the lessons we learned in the 90s, people just don't know or remember. I mean, the whole blank check company thing.

We did that in 1999, right? And it all the SPACs, right? And it totally failed. And then people did it again. It was crazy to me. And now it's kind of circular. Yeah, I know, right? I mean, it's still a bad idea.

And I think the circular revenue has also happened in the 90s, and that was also a bad idea. You'd think that people could see that and factor that out. God, I have so many questions. Why are spec's a bad idea? On paper, it seems like a totally fine way to go public. In practice, it only seems like total garbage companies take that route. Yeah, it feels self-selecting. Yeah, well, of course, because the way they're structured to get people to invest in a company, you don't even know what it is yet.

it means that it's not great for the companies, right? You get a ton of... Yeah. Oh, right. Right? That's just leakage. We just explain the mechanism again? So people, you know, if you put money into a SPAC, when they decide they're going to do a deal, you're allowed to take your money back out of the SPAC.

I can't remember all the details. All right. Right. So if you're like, well, it's a good deal. I'll leave my money in. But if you're the company being acquired, you don't know if you're going to be acquired or not because people could take their money out. So it just, why wouldn't, if you could go public on your own, you would prefer to do that.

So by a definition, the SPACs are buying companies that couldn't go public on their own. Right, right. Is VC investing fun again? I got the impression, like, two years ago, everyone was pretty depressed. And I'm quite sure we did a few episodes on it at the time. But are people having fun again? Well, again, I retired, so no. Nobody's having fun.

I wasn't having fun. I think if you have a billion dollars, it's probably fun. Yeah. You know, if you're doing the big deals, I think if you're in AI, it's fun. if you're doing anything else, it should be fun if you're doing something besides AI, right? Because now everybody is distracted by the shiny new thing. You can go find companies that are interesting to you and invest. The problem is you have to worry about what happens a year from now when they need the next round.

Is anybody going to be paying attention? I think it's probably pretty hard to be an early stage investor unless you're investing in AI. And if you're investing in AI, you're probably not writing small checks, low valuations. and you can't control the outcome at the end. You can only control what you do at the beginning. So you probably won't be making money. I have this theory that actually nobody likes bubbles or booms even. But let's say bubbles in part like if I missed it, I'm upset because someone else is getting rich.

If I'm in it, I'm like really anxious. Am I going to like, I'm anxious about two things. I'm anxious am I going to sell at the right time? I'm also kicking myself for not investing more. No matter how much I invested, I'm upset with myself. for not having invested more. Does that write, this is just my impression when I read history,

which is that everyone, even in the boom times, there's like, this like din of like stress underneath. Is that true? Am I just fantasizing projecting my own neuroses from birth onto other people? Well,

everyone thinks they're going to time it right. But it's fun. The bubbles are fun, especially if you're young and stupid, right? There's the New Yorker cartoon. I want my bubble back, right? Yeah, yeah. The flip side is, it is stressful.

I remember, well, it's stressful both after, obviously. I still have my, Razorfish Stock Certificate, I had it certificated because I couldn't sell it for anything, you know, for any real money. And that at one point had been worth quite a bit of money, more than my house. I remember, what did that company do? Is it like an ad network or something? No, no, they built websites.

Oh, yeah, right. Yeah, right. Yeah, right. They were great. I mean, they were a great company when nobody else knew how to build websites. I had some friends who worked for Razerfish even as like recently as like 2010. They sort of hung on for a little while. Jerry Newman, thank you so much for coming on odd lots. It's been wanting to chat with you for a long time and really appreciate you joining us. Yeah, thank you both.

Tracy, I love that conversation. There's a lot there. I mean, obviously, I like any time we can talk about the 90s bubble, but I had never really thought about or come anywhere close to thinking about the AI analogy with containerization. It's a little embarrassing because that's such a core topic for us. We've talked about boxes so many times. But to sort of reorient my thinking of AI in these terms is very helpful. Yeah. Well, I mean, this just kind of proves the point that no one thinks of containerization. I mean, I said before, it is a technology story. And one of the reasons I do think of it that way is because I read that book, The Box, which is really good. But no one thinks about it as an investment story because of the reasons that Jerry just laid out. Yeah, no, that's really interesting. And, you know, again, like, it does feel at some point that non-tech businesses, non-A.I. businesses eventually.

someone, hopefully, for the industry, like, makes a lot of money actually using these tools. Because we've been in the Pics and Shovels phase or whatever. But at some point, maybe it's an existing healthcare company or et cetera. Or maybe it's a new kind of law firm or an incumbent law firm where it's like, okay, we have found a way to use this technology in a manner that is very profitable, productive, and market expanses. So to me, that's the key thing. key thing is it's not that we're going to use this technology necessarily to cut costs and boost profit margins. It's that we will actually expand our customer base and make it up in volume by selling more knowledge. You know, it's an interesting thing that Jerry said,

and he put into words something that I hadn't really thought about before, but this idea about being at the end of this sort of computer revolution. And there is something about AI specifically where people, it's like, and it can't really literally be this where it's like, well, this is the last technology, right? Well, no, because we're going to get robots next. Yeah, right. And I don't know if like other booms or technological revolutions had this feeling where it's like, this is the last one.

Theoretically, if you get AGI or whatever, maybe robots, you don't need any further technological innovation, et cetera. It creates, I think, a very weird, uncomfortable dynamic. But the idea of AI is the end of what we do with computers. rather than the start of like something genuinely new. Like, that actually like snaps into place a lot of thoughts for me. Once we invent God, we're done. Yeah, what's we invent God?

We're done. Everything else takes care of itself. Yeah. All right. Shall we leave it there? Let's leave it there. All right. This has been another episode of the All Thoughts podcast.

I'm Tracy Alloway. You can follow me at Tracy Alloway. And I'm Joe Wisenthall. You can follow me at the stalwart. Follow our guest Jerry Newman. He's at G.A. Newman. Follow our producers, Carmen Roderiguez at Carmen Armin. Dashel Bennett at Dashbot and Kill

Brooks at Kail Brooks. For more oddlots content, go to Bloomberg.com slash oddlots where we've been daily newsletter and all of our episodes. And you can chat about all of these topics with fellow listeners in our Discord. Discord.g.g. slash oddlots. And if you enjoy Oddlots, if you like it, when we talk to you about why you're not going to get rich from AI, then please leave us a positive review on your favorite podcast platform. And remember, if you are a Bloomberg subscriber, you can listen to all of our episodes absolutely ad-free. All you need to do is find the Bloomberg channel on Apple Podcasts and follow the instructions there. Thanks for listening.

Bloomberg Tech returns to San Francisco, June 3rd and 4th. Join me, Emily Chang, along with Tom Giles, as we convene the CEOs, investors, and innovators shaping what's next in technology. Powered by Bloomberg's Global Newsroom and unrivaled terminal data will break down the capital, connectivity, and big ideas driving the industry forward.

Be part of the conversation. Register now at BloombergLive.com slash tech. That's BloombergLive.com Tech.