It was a pleasure hosting the second annual IA Summit in person on October 11, 2023. Over 250 founders, builders, investors, and thought leaders across the AI community guests and over 40 speakers dove into the world of intelligent and generative applications — from the open-source and closed models needed to run them to the emerging architectures and frameworks needed to build them to the battle emerging between Gen-Native and Gen-Enhanced applications. We’re excited to share the recording and transcript of the Public & Private Market Perspective in Evolving GenAI Ecosystem discussion with NYSE Group Head of Capital Markets Michael Harris, Greylock Partner Asheem Chandna, Altimeter Founder & CEO Brad Gerstner, and Madrona Managing Director S. Somasegar.
First, NYSE Group Head of Capital Markets Michael Harris walked the audience through the current state of the public capital markets, with a focus on IPOs and AI’s influence. He said that this year, we’ve raised roughly $12 billion worth of capital from across roughly 50 IPOs. In a typical year for the IPO market, we typically see about 100-125 deals that come to the marketplace, usually raising around $100 billion – $125 billion worth of proceeds. Following his discussion, the panel dove into the challenges and opportunities in GenAI, with a focus on strategic investments, successful AI strategies, and the impact of regulations on the industry.
You can see all the session recordings here.
This transcript was automatically generated.
Tim Porter: I’m Tim Porter, I’m one of the managing directors here at Madrona. We’re really excited to now move the discussion from all of the great discussions and talks we’ve had on technology and use cases and applications, to getting a financial market view of what’s happening in the GenAI landscape.
We’re really excited to welcome a great panel here. Michael Harris is going to share a few thoughts, to start. Micheal Harris is the Global Head of Capital Markets from the New York Stock Exchange. We have been delighted to have the NYSE as a partner here for the IA Summit highlighting many of our companies and the companies across a number of different ways.
Michael has had a career in capital markets on both the buy side and the sell side. Previously at Citadel, J.P. Morgan, Morgan Stanley and in Treasury. He’ll share views on the public markets first and then he’ll lead a great panel, where we’re fortunate to have Brad Gerstner, and you guys can come up. Brad Gerstner is the founder and CEO of Altimeter. Asheem Chandna, Senior Partner at Greylock. And our own Soma Somasegar, who will share their thoughts in discussions on what’s happening in the private markets.
Micheal, take it away. And thanks everyone for being here.
Michael Harris: Great. Thanks Tim. Appreciate it. So maybe just as a quick overview, we thought it would be helpful just to give a very quick overview in terms of perspectives within the public capital markets and then we’ll pivot to talking about the impact on private markets and also a little bit more about some issues that have been highlighted in today’s discussion.
First, how we got here. As a brief overview from a public market perspective, obviously not surprising if you look at the landscape for issuance overall, top performance is just looking at the S&P, Nasdaq and Dow over the last couple of years. And really what we’ve seen and I look at in two different parts. You have pre-COVID and post-COVID. Pre-COVID, when we had an interest rate environment that was essentially close to zero. Huge input and huge focus on the riskiest parts of the asset class.
And not surprisingly, as you look at the top right part of the graph in terms of IPO issuance, that’s where we saw investors allocate an enormous amount of capital, and effectively as we think about it within the capital market space, publicly every asset class was open, whether that’s IPO, SPACs, PIPEs, equity-linked issuance, private placements, everything, is essentially risk on.
Post-COVID as the Fed began increasing interest rates, all of that changed. So we’re now in a landscape where we’re in a much more risk averse environment. There’s much more of a shift that’s taken place, where now investors are more focused on companies that are profitable. Much more really focused on value versus growth. The SPAC market, which many of you have probably about, has effectively closed for issuance. And those that are still existing are having a very difficult time closing transactions.
And the higher marks and the higher evaluation and higher multiple companies that have been very prevalent in the marketplace pre-2021, you really haven’t seen that taken place. And so there’s much more of a focus towards cash flow, profitability and also companies that are on the pathway to profitability over time.
Where we are now, is a much lower environment for IPO issuance overall. If you could take a look at the top part of the chart, we’ve raised roughly about $12 billion worth of capital. That’s historically much, much lower than what we’ve seen in the marketplace. And that’s been across roughly 50 transactions. To give you some perspective, in a typical year for the IPO market, we typically see about a 100, 125 deals that come to the marketplace, usually raising around a $100 billion, $125 billion worth of proceeds. So, still a lot lower. A bit of a recovery versus where we were last year, but still quite a bit of a retrenchment.
The bottom right-hand graph is kind of interesting, just because as you look at where the issuances come form in terms of proceeds, because it’s been such a slow market, it’s been very lumpy. And about roughly 60% of those proceeds are coming from essentially three transactions, Arm, Instacart, Kenvue. And that kind of makes up the bulk of issuance for the year. So very lumpy in terms of the overall pipeline.
Turning to AI, I think a couple of things that are just interesting takeaways. Obviously there’s been a big focus on the performance of companies like Nvidia. But that’s been really very concentrated overall in the marketplace as far as market returns, within a very small handful of companies. So, the left-hand part of the graph is looking at the S&P 500. But roughly 88% of the investment returns, almost 90% of the investment returns are focused on eight companies. And they’re very heavily skewed towards Nvidia, Google, Microsoft and a few others.
Not surprisingly, as you look at companies that are now looking at trying to pivot their strategies, incorporate AI into their overall corporate strategy going forward, you’re seeing the percentage of companies that are mentioning AI, or generative AI into their earnings announcement increase quite a bit.
The bottom graph if you look on the left-hand side, in the second quarter of 2022 there were no companies in the S&P 500 that mentioned the term generative AI. That’s increased substantially if you look at the second quarter of 2023, not surprising. And the other things you look at, the companies are just using the words AI, that’s increased dramatically.
Now, part of that, we kind of laugh in this room, there’s probably a bit of a focus on companies that are saying, “Well, I want to be a part of that trend as well.” But some of that I think is also very much a part of what you’re seeing companies trying to incorporate as a part of their general strategy. We heard from Marco over at Goldman Sachs, the way that they’re incorporating that as a part of just their overall platform across all parts of their business. I’ll talk in a moment about how Blackstone is using that across their portfolio. But I think it’s just interesting from an issuer’s perspective how much this has become an enormous focus in terms of corporate America’s viewpoint in terms of this space.
If you turn to the buy side, the traditional asset management part of the business, and by that I mean investors that are focused primarily on stocks and bonds, and this is just a quick snapshot, but BlackRock systematic and there’s another few folks in BlackRock that are here, they’ve been probably the most prolific at using large language models to try to look at the predictive elements of how the market’s going to react in terms of corporate earnings.
So if you look at when issuers go through the process of putting out an earnings report, that’s obviously something that they just can’t scan through the language, try to figure out whether or not they are positive sentiments, negative sentiments and whether or not that’s going to actually interpret into being a positive reaction to the marketplace, or a negative reaction as well.
The 60% corelation element is interesting. That’s the orange graph. That’s kind of their model that they’ve built over the course of about the last 15 years. GPT-3.5 Turbo, which they’ve been using as a use case just to try to look at the veracity of it. That’s actually starting to approach a pretty high degree of corelation, especially when you look at GPT-4, but nowhere close to approximating the performance of their own in-house models.
And that’s something that you’re seeing a lot of within the traditional asset management space, where most long-only managers are developing their own models in-house. And looking at possibly using GPT-4, possibly using other applications, but for the most part, they’re all developing their own in-house models. More importantly, as you look at the large language model industry for earnings, that’s become almost standard. Pretty much every long-only fund is using it.
The thing I thought was interesting, is if you look at hedge funds, just as recently as 2017, and this is coming from the folks at Preqin, which is the big publication that sources information from the alternative industry, only 9% of hedge funds actually reported using AI tools in any part of their investment process, whether that’s identifying assets, managing assets, or working with their portfolio managers to try to use that for portfolio optimization. 2023, it’s now expected to be 90%, so that’s a huge change in behavior.
But I think the other element is also when you look at the alternative space within private equity, that’s become where you’ve seen a lot of managers devote a significant amount of resources into using artificial intelligence. And part of that is just because the challenges with large private equity managers are unique. On one hand you have size issues that they’re dealing with. And the other part is just like the structured data aspect of the business is also very unique. They’re dealing with both traditional part of the data business, and also alternative parts of the data business.
And the other part is what we think of as the big four Vs. The volume is massive. They’re dealing with public data and also data that’s not public. The variety of data is enormous. And the velocity is always changing. And at the same time, the veracity of the data is also something that’s very unique to their business. I thought it’s interesting, it affects all of us because not only are we either directly or indirectly invested with all of these firms in some form or fashion, but the adoption rate is something that’s picked up dramatically and will only continue to escalate.
So looking ahead, a couple of things that are going to be top of mind. Obviously, the environment going into 2024, 2025, there’s a huge focus in terms of the recessionary outlook. We’ll talk a little bit about that on the panel today.
I think there’s another element that we look at even though we are primarily focused within the equity space, but the impact within areas of credit, high yield, leveraged loans. That’s something that is also going to be a notable impact. The convertible bond market is another area that we also tend to focus on. And then I think the evolving IPO market is going to be another area that we’re going to continue to have to take a look at.
We’ve seen issuance come down, but some of the big areas that we tend to look at as being signs in terms of the health of the overall marketplace. Aftermarket performance is big one. Investors that are allocating capital to the IPO market, often times they’re serial investors, to the extent that they make money in the marketplace, they’re more willing to commit capital to the next transactions. If they’re not performing, then they’re less willing to do so. Volatility overall in the marketplace is a huge part of the driver of their ability to continue to commit capital. And also liquidity as well. As rates continue to increase, their ability to apply additional leverage to their strategies becomes constrained, and so it becomes tougher for them to continue to generate alpha.
The cornerstone investor process is kind of interesting. Something that we’ve seen more recently in the IPO market, has been this element of attracting, whether it’s strategic investors, or specific investors, to a particular transactions, shrinking the deal size, using all that as a way of trying to maximize the success of a deal. I think we’re going to continue to see that being a strategy that’s going to be the case going forward.
But I think probably the biggest element and something that we’ll talk a little bit about on the panel today, is evaluation. And I always think of it as the IPO market, the follow-on market, the convertible market, those markets in the public market are almost always open. It’s just a question of price. And right now, what we’re seeing is a very large gap between where both investors are willing to sell, and market participants are willing to buy.
So with that, why don’t we turn to our panel today and talk a little bit about the impact on the private markets.
You can view Michael’s slides here.
Michael Harris: Why don’t we start with Asheem. As we think about the debate in public and private, just in terms of where the moats exist, talk to us a little bit about for intelligent applications, what do you believe are the keys in terms of just building wide moats?
Asheem Chandna: Yeah, I think for apps, first I’d say I feel like we’re very, very early in the journey that lies ahead. And it feels like just overall things are very unstable. It almost feels like you’re in quicksand every week, with just the rate at which things are changing.
I’d say, that said, it feels like for apps, the two primary moat areas are likely proprietary data, and then I would say, just understanding workflows. And so I think, end of the day, there are some amazing technologies already available. And this is just going to get more and more amazing. But end of the day, I feel like the winners are going to people who can take it to the user, and who really understand the workflows.
Michael Harris: Great. And as you think about the funding and evaluation environment, looking at the enabling technologies, how do you see that evolving? How do you see the current environment for that?
Asheem Chandna: Yeah, this has been… It’s interesting for those of us that track venture financings and the flow of dollars, 2021 was a record year. So staff ran up. And then 2022 came crashing down. And 2023 is overall going to be a much lower year than 2022. This will affect VC dollars overall, and that’s both early-stage dollars and late-stage dollars. This is going to be a slow year compared to last year, and last year is much slower compared to 2021, which is record. That’s kind of just overall.
With that backdrop, I’d say the AI companies are kind of… It’s a tale of two markets almost today. Anybody raising money with something AI-related, you’re in your own bubble. And so these companies move fast. They get valued in ways which are completely irrational at some level. I think that’s the reality. And I think everybody’s asking the question of, “Is this going to keep going, or is this going to stop at some point, or will this change?”
But as of now, I think for those of us that are in VC here, like every week’s a fire drill, because you come into the Monday thinking your calendar is a particular day, and then somebody meets somebody on Monday or Tuesday and then everything moves in like three, five, seven days in some of these companies. And then often they’re being prices at levels where you may like the entrepreneur and the team, but sometimes it’s challenging to do it.
But that said, I think that everybody is taking, all these firms, including Greylock where I work, we are very active, because it is the most important thing going on just across the landscape.
Michael Harris: Great. Maybe Soma, I’m turning to you. As you think about AI, GenAI funding, how does the landscape look now and how do you see that changing going into next year?
Soma: Yeah, I think like Asheem was saying, the funding for GenAI in general is hot. I actually think it was hotter earlier this year. And I’m starting to see maybe some signs of, “Hey, can I go from stratospheric levels to little below?” Part of it could be the economy, but I’d also say when we are talking about early stage investing, I always say this, no data is good. A lot of good data is great. Some data is where you run into problems. When you have some data and not enough.
So early-stage companies, you talk about GenAI, the most hyped thing on the face of the earth, like evaluations are rich, no matter how you look at it. But like anything else, the law of averages, the law of numbers will come into existence. Right now the question is, is it in 2024, is it in 2025? We can debate. But I think given the potential that people are seeing, the way that people are talking about it like, “Hey, this is the next platform wave.”
And the thing I wanted to remind our senses, is like go back to the previous platform waves, particularly take the mobile wave. The mobile devices started coming into some meaningful play in the early 2000s. It was 2007, 2008 when iPhone came on, that the world said, “Oh, this is really serious.” But it wasn’t until 2010, 2011 that people felt that like, “Hey, things like Uber, things like Airbnb could happen on the mobile platform.”
So Asheem mentioned this earlier, from an application perspective in GenAI it’s too early to know what is even possible. We need to give ourself some time to figure out how things are going to evolve. But meanwhile there is all this expectation of how much opportunity there is, that is causing all these evaluation rates even at early-stage of companies. And I hope for their goodness and for everybody’s goodness in the ecosystem, that there’s a little more grounded approach to that, that I foresee happening in the next year or two.
Michael Harris: Yeah. How do you think about it?
Brad Gerstner: Well, there’s a reason that everybody is excited and why the evaluations are what they are. As venture capitalists for 20 years, every great company has always been overvalued. The fact of the matter is, it’s overvalued for a reason. What makes these moments difficult, I think, is the opacity of the moment. We’re all analysts. Whether you’re a founder or you’re an investors, you’re trying to figure out where the future’s going.
And the reality is, today it’s very difficult to know just like it was in ’98 or ’99, we were looking at early signal. I talked to my team a lot about the early signal in search. You could’ve gotten the internet right, you could’ve gotten search right, and then you went and made a bet on early revenues that you saw at Infoseek, at Excite, at Lycos, go through the list of 15, that turned into zeros a few years later. Or you could’ve waited to invest in Google in 2004 in the IPO, and captured 90% of the upside that ever came from internet search.
Part of the reason we have this push and pull, at Altimeter we’re about $10 billion split between the public and the private markets, the venture markets. And so we think about this every day. We sit around the table and ask ourselves where is value accruing and where is the asymmetric risk reward at a moment in time. And when you have these moments of opacity like this, we’ve talked about ourselves as a research organization, we’re in research mode.
And because of the gross overvaluations, the distortions in the market caused by the Fed, we’ve largely been researching and not investing since October of ’21. But that’s not because of our lack of enthusiasm about AI, it’s about making sure that those early revenues are in fact defensible, durable, repeatable, because there is a lot of trial and experimentation going on.
But I think that you can, whether you’re a founder or a funder, you have to manage these two simultaneous truths that are in tension. One is that this is going to be the biggest thing of all of our lives. And on the second hand, 90% of this stuff won’t be here in five years. And these early revenue indications are head fakes. And that some of this stuff, of that 90%, it’s all overvalued. But for the stuff that emerges, we saw some people on stage earlier, Harrison, Nikita, et cetera, there are really interesting things going on solving fundamental problems to enable solutions that I think are going to push humanity forward.
I’m really excited, but yes, there are a lot of reasons to be concerned about the world at large. And there are a lot of reasons to have a healthy degree of caution. If you’re a founder building a company and you have $3 million of revenue, and somebody is offering to do you a round at $700 million, it feels good in the moment. I started three businesses. I know how good it feels. But the next round you have to do at a billion and a half.
And so don’t snatch defeat from the jaws of victory by having too much success early on. Understand that it’s a journey and the goal of every round is to optimize your chance for ultimate success.
Michael Harris: Yeah, we were talking about that earlier. The same thing applies in the public markets as well. A lot of it, companies always focus on IPO day as the biggest event of their lives. And it is in many respects. It’s an enormous transition, but it is the start of a very long journey. And you want to be able to actually deliver on performance and deliver for the management team going forward, and for your investors. So it’s a great point.
Speaking of, as you kind of think about, Brad, a lot of folks here are working in the application space. And talk a little bit about the funding, the evaluations environment. How do you see that?
Brad Gerstner: Well, first I think that, again, I put myself in the shoes of a founder today going out and raising money, whether you’re an application writer or even up, you’re in the legal space, and the medical space, some productivity app, everybody’s going to want to talk to you. Don’t mistake them wanting to talk to you as over anticipating what that means in terms of actually collecting a bag of money at the end of it. Because every one of us is having to deal with the reflexivity of interest rates, geopolitical risk, all the things happening in the world, there is no doubt exiting summer that we’ve entered the zone of disillusionment that you heard referenced up here.
I think if I’m a founder, there is a lot of excitement about what you’re building. But I think you have to find that equitable distribution of risk and reward for the stage that you’re at. Be really sober. I would encourage you… We just went through this period of gross excess. And this year Mark wrote about the age of efficiency at Meta. I’ve talked a lot about time to get fit.
We have a 1,000 unicorns that are going to have effectively do what Instacart did in the public market. They’re all going to do down rounds. Even the best ones striked at a 50% down run. This is going to happen. We have a three-year process, not to mention the A pipeline, the B pipeline, the C pipeline that’s all going to have to go through this.
So when you’re starting from scratch, your seed, series A, series B, do not create these problems for yourself, because you have a bunch of over exuberant VCs who are giving you evaluation that you know is too good to be true. It is too good to be true. Set yourself up for long-term success with partners who are going to be there when the funding environment’s not so friendly.
Michael Harris: Yeah, good point. Soma, maybe turning to M&A, maybe talk a little bit about what you’re seeing in the M&A environment, what you expect it to be going forward and how do you see that today?
Soma: Yeah, that’s a great question, Michael. But before we talk about M&A, the thing that’s fascinating for me is and even when I look at the AI Fortune list of companies. Look at it and like, “Hey, who are the investors that have invested in most of these companies?” I think Sequoia shows up at the top of the list. I don’t know if people want to guess who number two on the list is or not. It’s Salesforce Ventures. Okay? And I’m sure that if we do it six months later, Nvidia will be giving Salesforce a run for their money.
Between those companies like Salesforce, Nvidia and more and more we are hearing that from companies like Google and Microsoft, and even the last week, announcement from Amazon who bought $4 billion into Anthropic. You are starting to see people taking a variety of, what I call [inaudible 00:26:15] practices, come to play. Some may be M&A, but investments taking a stake and all doing strategic partnership, relationship, whatever it is.
There’s a lot more activity that we are seeing from the hyperscalers today than before. I think part of the reason is you look at the balance sheet for these top five, six, ten companies, they’re very healthy, a lot of cash around, so they have resources to be able to figure out how to deploy these dollars in the best possible way. And M&A is one part of it.
Like Brad mentioned, $3 million company, $700 million evaluations, it’s no different than when Databricks decided to buy MosaicML, for a $1.2 billion. And I don’t want to talk about the revenue here, but you can all guess, you can think about or speculate what it could be. So there are some acquisitions that are starting to happen. And I think in ’24 and ’25, I think the bigger companies are going to continue thinking about how best to deploy their capital and their resources.
Particularly, we know that GenAI is the next big thing and everybody is looking for, “Hey, where do I make a bet? Where do I buy? Where do I partner? Where do I build?” And part of it is also the fact that the evaluations are somewhat crazy. All the acquiring companies are being thoughtful about, “Hey should I wait six more months? Will I get more data? Will we receive more reasonable evaluations?” So they’re trying to figure out what is the best for them. But I think the activity overall, is going to continue heating up in the coming year or two.
Michael Harris: And what do you think the impact on evaluation’s going to be? Linear?
Soma: It’s one of those things where the best companies and the best teams are going to command rich evaluations. And you just have to sort of… And the sooner you realize that, you are going to be able to make appropriate bets.
Michael Harris: Yeah, good point. Asheem, there’s been this increase in strategic investors. That’s been a big focus of the conversation over the last year. How do you think about that going into 2024? Do you think it’s going to continue?
Asheem Chandna: It’s hard to know, but I would say, assuming this remains the main theme, which in technology, which is likely to remain for well beyond the foreseeable future. So I think if that’s the backdrop, then I’d say every technology CEO and every board, this is the number item on their mind, so they’re going to deploy everything they have to build that capability set.
I’ll share a quick anecdote that, I was talking to a friend recently who runs a large publicly traded software company, you guys all know the name. I won’t mention the name of the company. He was telling me that he was recently meeting with a large customer. He’s meeting with a large customer of his. And he goes in with his team in the morning. And there’s a two, three hour conversation. And basically, the customer basically beats them down on price. So on the annual contract, they beat them down after two, three hours of conversation.
Just given the backdrop on everything that’s going on. Then he said, in the afternoon I go back into the same customer organization, and I’m going in with some of my team members, and I meet with somebody that starts the AI initiatives for the company. And the meeting then is like, it’s like night and day. The person showed up at the meeting, checkbook open, saying, “What do you guys have in AI for me? And I’ve got an unlimited checkbook.” All the way up to the CEO. He was like, “Hey, same customer.” That’s a little bit of what’s going on. It’s actually, it’s the vendors, but it’s all the way back from the customers.
But we also had cases, we’re early-stage VCs at Greylock, we don’t spend our life primarily talking to, let’s say, Fortune 50, Fortune 100 CEOs, but we’ve had many of those continue to reach out to us in the period now where the CEOs directly are reaching out and asking for Zoom calls.
And sometimes they get on these calls and we sign up. We’re not sure if we’re want to sign up. Some other exec shows up with like, “Hey, this is a [inaudible 00:30:29] switch.” But three months later the CEO shows up and they sit and listen and want to talk for 45 minutes. It’s on everybody’s mind, are they going to get disrupted and what should they be doing.
Michael Harris: That’s great. Maybe a question for everyone on the panel just as you think about the public markets, who do you see that’s executing well in terms of their AI strategy, generally speaking? Or partnerships that they’re working on, who do you see in the public markets that’s doing a great job? Maybe starting with you, Asheem.
Asheem Chandna: I’d say one is, we’re in Seattle and so just to start with, the obvious name here will be Microsoft and Amazon just to pick two. And I think it’s fascinating to see with Copilot, how Microsoft, how they’ve set pricing. And it will be very interesting to see where this goes and is this really going to increase TAM, or are you going to get for the same price point, are you going to get a lot more in the product. But it looks like there could be some potential to dramatically increase TAM. I would say maybe start there.
Adobe is another one. Maybe I’ll mention just I think the announcement they did yesterday or a couple of days ago. But they really seem to have plugged in generative AI into a number of their products. I feel like for Salesforce, ServiceNow, it’s still a little bit of wait and see. There have been some announcements, but still maybe waiting to see what really comes out of it. And maybe I’ll pause there, yeah.
Michael Harris: Soma?
Soma: I’m not biased, but I’ll start by saying Microsoft has truly done, I think, a remarkable job. And particularly when you look back at what they’ve done in the last five years or seven years. I think they’ve anticipated, partnered and executed really, really well. But if you look at broadly, the public companies, the two companies, for me at least, that I’ll put above the line, are Microsoft and Nvidia.
The thing about Nvidia that’s interesting is the last 10 years or so, anything new that the world thinks is going to be the next big things, Nvidia is right there. It started with more cloud computing capabilities. Then autonomous vehicles. Then Metaverse. Then AI. Anything you talk about like, “Hey, Nvidia is now the player.”
So either they have done a phenomenal job of anticipating what’s happening in the world, and they’ve executed really, really well. Some combination that’s working really well. And it’s been super impressive to see how they’ve embraced AI in the best possible way, particularly from the silicon, from the chip set, from a software perspective.
Then for me, slightly below the line, I actually think as much as I love the other technology company in Seattle, I think Amazon, Google and of course the ServiceNows and the Salesforces of the world, I think relative to what they could do, what they should do, what I think they’re capable of doing, I think they are all starting to execute better and better with every year that goes by. But your question is, who’s embraced so far. I think I wish they had done a little more in this time period. But ask me the question six months from now, some of these guys will absolutely be above the line.
Michael Harris: Okay, great. Brad, what do you think?
Brad Gerstner: Well, it was music other my ears this morning when I heard from the stage that maybe the best job to have over the next five years, would be a hedge fund manager because of the obvious winners and losers to come out of this age of AI. Part of why I love our crossover function, is I was talking with Mustafa Suleyman last year, I remember not being able to get my head around inflections evaluations, and he said to me, “I know we’re overpriced for you, so you should just go put everything in Nvidia.” He said, “Like I did.” And fortunately, we’ve had our best year, first or second best year in the 15-year history of our firm this year on the public side.
And yet, if you look at the world, the S&P ex-tech is down 4% on the year. The Russell is flat on the year. Our short book is up on the year. And our long book, which is comprised basically of Meta, Microsoft, Nvidia, Uber, Amazon, is up 80 plus percent. So it’s been an extraordinary year of dispersion for a lot of reasons. But I actually think we’re early and just getting started. And I don’t think the same winners this year, will be the winners as we roll through this.
The first layer, it’s not surprising, is Nvidia because they’re the enablers’ enabler. As we heard last night, and I was talking with my good friend Feroza up here, if you add up the value of all the hyperscalers today, maybe a trillion, like the cloud components of their business, maybe a trillion and a half dollars of enterprise value, maybe two trillion. Today you have Nvidia valued at a trillion or a trillion and a half. Does that relative valuation make sense? Sacha told us last night, probably not.
And so, I definitely think we’re having Microsoft as performing incredibly well. But this is Amazon, the boat, the tide’s going to rise there. I think this will be good for Google in a lot of ways. I think it’s existential for them in a lot of ways. The two companies, when it comes to AI and then on the infrastructure side, I’d be remiss not to say Databricks and Snowflake as well, who I think are both doing really interesting things.
But the two companies, two of the biggest, probably the most profitable AI applications in the world, Meta and ByteDance. ByteDance will do 25, 30 billion in net income this year, growing at 55%, it’s content and monetization flywheel is made. I first heard, I asked Yaoming in 2015 how he was going to build and application that targeted content of consumers without a social graph, and he laughed at me and he said, “Brad, AI is way more powerful than a social graph.”
And so I think that Meta is incredibly well-positioned. I think that ByteDance is incredibly well-positioned. Obviously political challenges there. The trillion-dollar struggle that I get really excited in thinking about, I spent the first decade of my career in consumer. And Google has at an absolute lock for 20 years on the top of the funnel. They have collected the largest tax, the largest toll on the internet.
And the first breach of the moat that I’ve seen in 20 years, I thought mobile might be it. Desktop search volumes went negative in 2012. And I thought, “Here we go.” But Google figured out how to increase monetization on every SERP, how to push organic results down, and magically grew earnings 20% for the next decade. I think that that string is really taut at this point.
And whoever can capture the imagination of the consumer for the general purpose utility agent, that agent in my pocket, that helps me find and book my trip, deal with restaurants, answer questions that I’m going to give information to, that’s going to make them uniquely positioned to answer those questions.
I was at the Javits Center last week. I asked an audience of a 1,000 people, how many have used ChatGPT, half the audience raised their hand. I said, “How many use it instead of Google?” The same half raised their hand. That hasn’t happened in 20 years, okay? So to me, watching how that’s going to play out, I think the contenders really are Meta, Google, Microsoft slash ChatGPT. But the sleepers here are really Apple and Amazon, who have incredible consumer footprints and chops. And so I’m excited to see what they’re going to do as well.
But I think that that change is coming. And what I said last week on CNBC, is that the monopoly that Google has today on search, the idea that they can cross this chasm and replicate this monopoly in the age of AI with a general purpose agent, if they do that, it will be a case study of one. Even Sacha said last night, rarely do you make change between these moments of tectonic transition where the incumbent emerges as the leader on the other side. So I think that’s going to be, that’s the trillion dollar battle that I’m excited to watch.
Michael Harris: Fantastic.
Soma: And like you said, we were lucky. I shouldn’t say we. Sorry. Microsoft was lucky to have caught the last train out, because the one other example, not necessarily consumer but more enterprise, is like, “Hey, Microsoft owned, at some point in time, or maybe even 80% or so of their server workloads.” Servers around the world was running on a Windows server. Being able to say, “Hey, I need to move from that to cloud computing,” where the margins are going to be completely different. And Microsoft caught the train a little late, but they’re in a good position to [inaudible 00:40:25], but that rarely happens in the history of business.
Michael Harris: Yeah, good point. I think we might have time for one question.
Question from crowd: Yeah, my question is about how the regulatory environment impact on private and process evaluations for GenAI, particularly with congress looking to regulate the capabilities, the FTC looking into preventing big tech companies from acquiring competitive companies.
Brad Gerstner: I think Lina Khan is an unmitigated disaster for American capitalism.
Michael Harris: But tell us how you feel.
Brad Gerstner: Yes, for everybody in this room. But that’s going to change. The great thing about this system that we have, is she will come and go. And sanity will prevail. And the fact of the matter is, founders who are risking everything deserve access to bigger companies. It’s good for consumers, it’s good for the capital system, it’s good for the entrepreneurial ecosystem.
Where did we lose our minds, that actually doing something that’s beneficial to consumers, should be blocked by Washington. That’s going to change. But in the interim, it’s caused a moratorium really in boardrooms and amongst CEOs focused on buying things. And it’s caused me, as a capital provider, to look at companies I would’ve otherwise funded, and said, “Oh, well. My downside protection is Meta buys it, or Google buys it, or Amazon buys it.” And now I have to discount at a much higher rate than I would’ve otherwise discounted that downside protection. It’s not a good thing for the funding environment today.
I think the regulatory environment, I’m quite sanguine having talked with many members of congress and the administration. I think Elon talked about this after his visit to the White House. I’m quite sanguine that the US understands the national security implications, the value of progress and maintaining our leadership in AI. So I think we’re going to get a sane set of regulations or a lack thereof out of Washington. And hopefully, over the course of the next four years, we’ll see a change at the FTC.
Michael Harris: Great. I’d like to thank everyone on the panel. It’s been a great discussion, we’re all clear on. Thank you.