Transcription
You guys are steeped in this stuff every day. Uh, the question was whether or not you believe AI is a bubble, which I think is the place we would probably start this conversation anyway, given that I know that's top of mind for everybody. Um, but maybe Bill, do you want to kick it off?
>> Sure. Oh, thanks for having me again, Leslie. It's great to be here with my friend Philipe. Um, uh, I'm going to say to be a little bit maybe controversial, not a bubble. And I say not a bubble because I think this is a very real tech cycle. I think back to the PC cycle of the 80s, the internet cycle of the '90s, the mobile cycle of the 2000s, cloud 2010s, and now AI. And this will be a very, very real market. Will create a lot of value. I think there will be a productivity payback long term. So I don't think a bubble, but I do think moments of overvaluation, moments of misallocation of capital. Sure.
>> And Philipe,
>> I'm always concerned about something. There's always something [laughter] that can go wrong in life. Um, I would say um, when you look sometimes at previous bubbles, it's not obvious necessarily who wins. So could we be making the mistake of thinking that someone who appears to win ends up not winning? That's one type of mistake. And the second one is that if AI becomes so successful, you could imagine the economy growing very fast, but a lot of job destruction along the way. And it's sort of odd to be in saying, listen, we love AI, we love everything it can do, but also does it lead to a lot of uh unemployment?
>> And so that's maybe more of a 10 or 15 year concern, but I sort of think about it already now a little. So maybe bubble isn't the right prism to be looking at the the broader concern as it pertains to AI, whether it's a bubble or not, but what it ultimately does down the road to.
>> It does feel, and and and Bill, interject too, but this wave feels like the changes are going to be dramatic in the future. Like in five and 10 and 15 years, the world is going to look potentially very different than today. Maybe people said the same thing with computers or with the mobile internet. The changes were maybe a little bit more incremental. This goes at the heart of replacing humans and both white collar and blue collar and uh drivers and cars and humanoids. There are going to be a lot of changes. I I I agree completely with Philipe. This is almost a super cycle and I think it's so fundamental. And one way I've started to think about it is that it's it's intelligence on demand. And you can imagine like, you know, like railroads, like electricity, like the internet. It's a it's a fundamental new technology and it will create huge markets. I think Philipe is right. It's very hard to tell all the winners and losers right now. We're too early to do it. But but one one thing I just want to want to say about it that's different than the internet cycle is that in this cycle, the incumbents have an edge right now. The big the big tech giants have an edge because it's capital intensive. It's engineering intensive and it's data intensive and they bring those assets to the table where the new entrance, well, there'll be some winners for sure, Open AI, Anthropic, others, but um, going into it and building these LLMs, building this intelligence on demand, it's the big guys who have really have an edge early on.
>> Do you think Philipe that gives them immortality, the Mag 7 or the hyperscalers, that they will be in that five-year future that that you see? They will still retain that dominance, or do you think there will be other that all of spending essentially will kind of get them to a place where they're ultimately replaced by the next Mag 7?
>> That's a good question. And I know that a lot of tech guys think they're immortal or they're going to [laughter] be spending money to become immortal. Me, I'm a European. I like a good glass of wine every now and then. [laughter] I realize I'm not going to be, you know, immortal. Uh, I remember when the Mag 7 used to be called FANG and FANG++. And there's always these acronyms. And usually by the time someone knows an acronym by heart, like the Mag 7, by the way, it's not really the Mag 7, it's the Mag 10 because there's 10 companies I think that are worth over a trillion dollar market cap. There will be another uh another index. And maybe Open AI or Anthropic or some of your companies, Bill, will also make it to that index someday. So I think the index will change. And then not just in the US, but you were talking to me early on, they might also be uh Mag 7 in other areas.
>> Because it's it's your jobs to to pick who those winners will be. And and given just all the uncertainties about how it does reshape our lives, how it does reshape the economy. How do you how do you diligence something like that? How do you decide what to invest in, how much to invest in them as you seek obviously that that next big win?
>> Well, I mean, for us, it's about following the entrepreneurs. And I think they're already, we're going to see a whole another generation of, I think, great entrepreneurs. They're already at work building companies. And the thing that I'm I'm most excited about right now is this this new set of application companies who are going to basically harness uh the intelligence on demand and compute that's available to build new solutions, new applications that we can't even imagine. And I again, I go back to my railroad analogy where I think that we couldn't have imagined a lot of the applications around what was possible when you had transportation on demand that way. So I I think it's going to be I think it'll be an exciting time for us with some of the new companies that are going to come along.
>> I wish you had put a seat belt for us, you know,
>> because it's really what a time to be alive and what a time to sort of uh invest and observe. Along the line of what Bill says, as an example, Amber, this conference was called Seeking Alpha. So just to give out themes, you know, and and and ideas. I don't know if I can deliver alpha, but at least I can [laughter] you know, alpha. If you look at a company like Intuitive Surgical, they make robots, but those robots right now, they look really complicated, but they're a bit dumb because it's still a human that's like playing with the joysticks. How long is it before a computer re-records every single surgery that's ever been done on the robot, learns, and soon you don't need a surgeon? And then that would be incredible because think about the ability around the world in any area to be able to have an amazing surgeon. And let's say some great surgeon in Los Angeles invents a new procedure. I read the other day to do Achilles heel, uh, to fix the Achilles heel. Overnight, you change the software of the robot. Everybody around the world gets access to that new procedure. So one positive thing for me on the AI is it's also uh gonna make it that like maybe people who don't have the best access to everything, the cost is going to come down a lot and it's going to spread a lot around the world. And I I hope at least it will happen in the example of the surgery and the healthcare. I I I think Philipe, that that's I think great because um, two industries that did not get really as impacted as we thought during this last tech cycle was was education and healthcare. Like Philip was talking about, you know, about this robotic surgery. And I think this time, those two industries are going to be transformed by AI.
>> And you know, when when you start bringing that that kind of medical intelligence either to a robotic device or making it available, making, you know, every doctor more productive, more intelligent, Um, it's going to have a huge impact in education too. Another industry where you know, again, access to education, quality of education, quality of teacher talent enhanced by AI. I think we're going to see some incredible new solutions created.
>> Interesting. And coming off of the pandemic as well, where those two areas felt right for disintermediation.
>> I have a a kid who's going to high school. And yesterday we went to visit this new school in New York that is supposed to be an AI high school.
>> Oh, wow.
>> When you look at the program, they really make you believe in the demo. You're like, I completely agree. You're going to learn in three hours. What it would have taken six.
>> Yeah, exactly.
>> And against that, we said, well, how many kids are there in the high school? And they were like, six.
>> And I meant six per class or six in the high school is just starting?
>> And so we went like this and we're like, oh god, maybe education is not just AI, it's also socializing and being with kids. But in five or seven years, when that high school is full, maybe for my younger kid, I'll be like, "Okay, that's pretty interesting. Try that one."
>> What is an AI high school? Is it
>> It was one where uh kids learn on their own and there's programs that are extremely aware of their strength and weaknesses. And it's not really a teacher. It's more like a mentor that sort of gives you the confidence and the motivation to sort of learn uh this way. It's a new program. But even if that doesn't appeal to all of you, in five or seven years, there's going to be five different ways of thinking about uh how to teach. And so
>> Absolutely.
>> sort of another interesting example. As we think about opportunities, um, there's been a lot of discussion about how much is in the public markets versus how much is in the private markets. How do you think of that? Um, how do you think about that? I know obviously General Atlantic is primarily in the in the private markets. Uh, CO2, a little bit of both.
>> A lot of both. No,
>> a lot of lot of both. [laughter]
>> Um, do you do you agree with the notion that the private markets are are where the opportunity is right now, or do you feel like there's uh the the kind of there's a blurring of the lines between the two and that you can find opportunity anywhere?
>> Well, I'm glad I have friends like Philipe so I can get insight in what's happening in the public market because we do focus on privates. But I but I but I think you have to understand both. As I was saying earlier, I really believe right now that the the people driving the the the change in AI are the large public company part public tech companies. And I I think the incumbents have that advantage. So if you don't understand what Oracle is doing, what Google is doing, what Microsoft's doing, you really can't invest in the private market, you really you really can't make good decisions. So even if we're not making investment decisions on the public side, we we we have to be pretty fully aware of what they're doing and that's how we synthesize it into our into our decision-making.
>> Yeah. And uh, you know, in our company, we do everything from seed to privates to publics. On one hand, you could say, okay, maybe we're at an advantage because we know both about publics and privates. But on the other hand, the mindset to invest in publics and privates is different.
>> I would say sometimes the mindsets of public and venture uh can be the same around thinking about big ideas. But on the public side, it's not only you have to believe in the future, but you have to think, but is it already priced in or not? Like when Oracle goes from 50 to 350, at 350, it's like, maybe the idea is already priced.
>> And on the private side, because uh usually things take longer, you have to be more right. You have to be more patient. You have to be active in your companies. You have to support the founders or the CEOs a lot. So I would say that if you can do both, great. But they're definitely a different skill set. Sometimes I wish I was only in one or the other. Sometimes I'm happy to be involved.
>> Be involved. The the nice thing about private is just as we were saying before, I think this this tech cycle plays out over a longer period of time. And I think there is a there is a a duration that you get from being a private investor. you know, more of a five to seven year horizon that allows you to sort of play the cycle a little bit longer than sometimes I think the public markets are more demanding for shorter term performance.
>> Do you think it makes sense to fix the IPO process if the IPO process is broken, or does it make sense to democratize alternatives to give more retail investors access
>> Interesting.
>> to private companies, given just the leg of growth that they're in and that companies are staying private for longer, and you have Open AI valued at half a trillion dollars at this point as a private company?
>> I personally think that the IPO market is totally broken, like beyond repair, as measured by the fact that 20 or 30 years ago, there were so many more IPOs than today. Uh, today there's very few IPOs. Uh, we were talking earlier, this year there's been almost none. And I think it's not a great trend because at the end of the day, it's just easier for retail investors to be involved in IPOs and then onwards than everything before the IPO. So I find it, it's a little bit unfair. I do think that it's going to get fixed purely uh through competition more than through a regulatory action because when private assets get tokenized, the byproduct of being tokenized is you in essence become public because now you can trade at a premium to the price of the token. You could trade at a discount to whatever the token price was issued. And so all these people that are privatizing, tokenizing private assets, in essence, it's making them public. And so I think that one day maybe all assets are going to be uh public and tradable to the point that maybe some companies and they're going to miss like, oh, I wish
>> wish not tokenized. And I think there's going to be a battle between companies that choose to be tokenized and some say, I don't want to be tokenized. I don't really subject to the whim of uh investors uh putting a daily, maybe on a per second value.
>> I'm I'm a little more optimistic uh about the IPO market. I I I think we we had a we've had an exit recession in in uh in in private markets for the last three plus years. And some of that was regulatory, some of it was just some of the trends that Philippe was talking about. Um, but but I I think um, I think we started to see some some green shoots around the IPO market this year. And it was just unfortunate that um the wind was taken out of our sales a little bit when we had the the government shutdown. There were a lot of high-quality companies on file ready to go that would have gone public in Q4. I think built some momentum going into 2026 that got stunted. And I I hope it'll come back. I think it will come back in 2026. I think there's still um, I you know, while the the balance may have shifted towards privates, I think there's a desire among public investors to have high-quality pro you know, companies come public.
>> Yeah. Um, and of course now you've got Thanksgiving, Christmas, so this year it's tough.
>> Becomes a little more challenging. And they could have done so at market highs.
>> Yeah.
>> Um, you know, I want to ask you kind of about international opportunities as well. I know Bill, you just got back from uh, what was it, a couple weeks?
>> Two weeks. Yeah.
>> Abroad, including China. Um, and just real quick before we get into just the the broader global environment.
>> On TikTok,
>> is the deal done?
>> We it we we see the press releases, we've seen executive order. Like, is there a bow on this deal at this point?
>> I I think we're I think we're on track to get the deal done as it's been reported. I mean, there was when when the framework agreement was announced in September, it was a 120-day clock on it. That puts us into a closing in January. And I have every reason to believe it's going to it's going to work out on schedule. And uh, so I I don't think there's any reason to be concerned.
>> So January, right, when the IPO market opens back up, all happen at the same time.
>> Let's hope it opens.
>> 2026 is our year. Um, Phippe, I took a look at your 13F uh going into, which I know is very delayed because the deadline is tomorrow, so we'll get a fresh one tomorrow. Um, but you had pretty limited international exposure at least on the long side. I'm curious if that was by design, if that was a reflection on kind of your view of international visa via opportunities on the public side here.
>> Our our our view of the world is that the US is the greatest country and the greatest place to live. And even more so with AI. We have all the engineers. We have all the new great AI companies and we're dominating the world of AI. In Europe for a while, and I can say this being French, you know, I thought that MAGA would lead to a mega, you know, make Europe great again. Europe is completely stuck in reverse. Uh, it has this sort of European government. It's really the United Nations of Europe. It's not the United States of Europe. They can't make any decision. It's very bureaucratic. Plus, they can't invest in AI because they have to invest so much in energy and defense. So, it's difficult to make an encouraging. Uh, now, that said, the European markets were very strong this year, so I could be totally wrong. Bill and I share that I think China is an incredible market uh just because they have very good engineers and it's a big captive market.
>> Me personally, I still prefer for the time being to invest on the public side of China. I respect all the geopolitics and at least on the public side, you could choose to change your mind. As some of these big Chinese private companies go public, I think it will give investors confidence to reinvest in China. But we also have to invest in China very much with the rules and regulations that the US government imposes on all of us. And it's definitely a set of constraints. And then in Asia, I think Japan, Taiwan, and Korea, especially on the semiconductor side, have a lot of uh interesting uh things going on. But I really at the end of the day, would be like 80% plus of the action is in the US.
>> Yeah.
>> And Bill, what did you learn from from those travels?
>> Um, well, I I I China remains a strong innovation economy and and um, they have half this world's STEM engineers. They've got a a very large home market. They've got a billion plus internet users. Um, so I I I think it's a very exciting place. Um, I think um, if you if we were here a year ago, people would have said China was uninvestable. This year, China is investable again. That their equity market's up over 30%. Their trading bonds are up 100%. They they're the they're the market leader right now in IPOs this year with 80 IPOs and high volume. So I I think you're seeing a real uh, you know, reacceptance of China or, you know, investors becoming more comfortable investing there. And I think we're going to see tremendous innovations. I actually agree with Philipe on Europe, unfortunately. I but I think in innovation, it's a two-horse race. I think it's US and China.
>> Who's winning right now?
>> Well, I I think they're both going to be strong. I I think both have very large home markets. Both have significant engineering talent. Um, and and if you look at areas like AI, they've got their leaders as well. Alibaba, Tencent, ByteDance are very strong AI competitors. I'd say in robotics and industrial automation, I think they're ahead of us right now to some extent. So I think it's mo but where I net it all out though is that at the end of the day, two incredible centers of innovation that investors should be exposed to for sure, private investors.
>> Uh, I believe we have an audience question.
>> Yeah, we have a couple on the app, but I'm going to go the old-fashioned way. We have a question right here.
>> Hi, this is uh Greg. Thank you for your insights. Could you talk about some of the trends in the cost of compute? And whether, you know, in terms of bringing on these applications, but certainly it seems like the costs are just succeeding at this point where those inlect and, you know, what does that ultimately mean for the the cost and the return on investment for the applications going forward? And what's the risk to the companies that may be running business models at losses today, uh, thinking about the future cost coming down, whether they'll be viable in the future? How do you think about that in terms of investing in those companies?
>> Yeah, so if you think of the cost of compute, you can measure it by, you know, tokens and the cost of a token. I think of a token to AI is like a gasoline to let's say, you know, an engine. And it's sort of strange because if you were to say, well, as price goes down, P times Q should go to zero. P goes to zero, but actually P times Q can go to infinity even though P goes to zero. And that's just the elasticity. I think if the price of tokens goes down, and it's going to go down very dramatically, the elasticity and the things that we can do with the lower price of token is almost infinite. We just mentioned Bill had a few examples around education and other spaces. I mentioned surgery, but you can imagine as the price of token comes down, so many things are going to be able to be done. And not just with the intelligence in the software, but intelligence in cars, intelligence in humanoids, intelligence in machines. So I'm sort of fairly optimistic that for a long period, a decade plus, at any decrease in the price of a token, the overall P times Q will still be extremely growing, extremely. Yeah, we we've had um, we have about 200 companies in our portfolio and we've been applying AI even at this level of cost um to to them pretty aggressively and we've seen very high paybacks already in areas like customer care, coding, digital marketing. And I think those are just the front edge of, you know, applications or value add that's going to be available by applying AI. And I think as Philipe said, as as the cost inevitably will come down, I think it's going to open up more solutions, more applications that are going to have productivity paybacks. [clears throat]
>> So, it's a related question, but are you concerned at all about the the complicated lending and kind of the circular economy going on between the big AI companies and some of the semiconductor companies that we've heard a lot? There's always the headline every day about this billion, you know, billions of dollars going towards compute and that those costs. Concerned at all about the maybe there's less transparency around some of those moves?
>> I think um on the public market side, Bill, you can if you want to address the private side, you should definitely study uh what's going on. And I think there's sort of like the naysayers have a little something in the debate. Um, but if you compare to year 2000, in year 2000, all of the capital was fueled by IPOs and was fueled by new companies with sort of fairly dubious business models. Today, the capital and if you think the the wealth of the world right these Mag 10, and I'm exaggerating a little bit, but are producing let's say almost a trillion of free cash flow, maybe it's 600, 700, growing to a tr close to a trillion. They have no debt. >> Everybody else by the Mag 10 probably produces no more than another one or two trillion of free cash flow with a ton of debt. They're very unincome, very encumbered. So the world equity is really concentrated in these 10 companies. And these are not investments that are made by the government or that are made by companies with debt or that are made by the capital markets. These are investments made by companies with real boards, return on capital requirements and and things like that. So I think the system is pretty healthy and the actual implied leverage in the system is small. Like these little things that Nvidia is doing is small. The fact that AMD wants to sell a little piece of the company, but they get to be potentially the number two player. I could see how that actually is a good move, that makes sense. So, I'm watchful. But if you tell me like, am I worried about it? Uh, not yet.
>> I I I think there I think what it's telling you is that these large companies that are taking that capital and and and investing in each other or investing is because they think there is a real significant opportunity at the other end. And Philip's right, the earnings are following through. They're getting revenue now. They're getting earnings now. And yet, I think they're all fighting for a very, very big prize. I mean, the ones who win the intelligence on demand game are going to sell a heck of a lot of compute and a heck of a lot of capability. And I think they're right to say we've got to invest now to win that. And uh, and there I think they're thinking amazing about they're getting a payback. I mean, you know, the amazing thing about the valuation increase of of the MAG 7 is the earnings follow through. This wasn't a case where you, you know, doubled or tripled, you know, PE ratio. It was a case where the earnings were there, which is what's been so remarkable to me.
>> Yeah, fantastic questions. And thank you both, Philipe and Bill. So great to have you on a day like today where this is front and center for everyone. I feel like I learned a lot from the conversation. I hope you all do, too. We covered a lot of ground, so really appreciate your time.