In short
Squawk Pod Episode Summary: The Next Internet: Brad Gerstner & Mark Mahaney (10/15/25)
Episode Overview In this episode of Squawk Pod, the focus is on the intersection of technology, AI, and economics featuring insights from tech investor Brad Gerstner and internet analyst Mark Mahaney. Key discussions revolve around corporate partnerships with AI, the impacts of the U.S.-China trade tensions, and the future landscape of the internet.
Featured Guests
- Brad Gerstner: Founder and CEO of Altimeter Capital, talks about the tech sector, AI dynamics, and the state of the market.
- Mark Mahaney: Internet analyst who discusses the implications of companies integrating with OpenAI's technologies.
Key Topics Discussed
- Corporate Integration with AI
- Walmart's collaboration with OpenAI allows users to shop directly through ChatGPT.
- Mahaney notes that leading companies like Walmart, Spotify, and Expedia are recognizing the need to integrate AI into their platforms, which could reshape the internet landscape.
- Impacts of the U.S.-China Trade Tensions
- The episode highlights ongoing trade disputes, particularly regarding soybeans and cooking oil, amid broader U.S.-China relations.
- President Trump's remarks on tariffs and trade policies contribute to market uncertainty.
- AI's Role in Economic Transformation
- Gerstner emphasizes that the current AI wave, particularly with companies like NVIDIA, marks a significant shift comparable to the internet's initial growth.
- The discussion includes concerns about potential market bubbles as interest in AI investments surges.
- Insights from Jamie Dimon
- Jamie Dimon, CEO of JPMorgan, provides perspective on the financial industry's response to AI and its potential to disrupt traditional job markets.
- He emphasizes the importance of evaluating AI projects individually to distinguish between genuine innovation and speculative investments.
Key Takeaways
Future of Online Shopping
- The integration of AI into shopping experiences is viewed as a way to enhance user engagement and streamline transaction processes.
Trade and Market Dynamics
- The complexity of U.S.-China trade relations creates a precarious economic environment, affecting various sectors, particularly agriculture and technology.
AI Investment Landscape
- The current atmosphere is filled with both excitement and caution regarding AI investments, with the potential for both significant gains and volatility.
The Role of Major Tech Players
- Companies like NVIDIA are positioned as leaders in the AI space, while others, such as AMD, may need to innovate more aggressively to keep pace.
- The importance of "agentic commerce" is highlighted, indicating a shift toward more personalized and automated shopping experiences.
Discussion Points
Brad Gerstner's Perspective on AI and Tech
- Gerstner discusses the competitive landscape among tech giants, highlighting that only a fraction of companies will rise to the top due to evolving consumer expectations and AI capabilities.
Mark Mahaney's Insights on Market Reactions
- Mahaney expresses optimism about AI’s transformative potential but warns of the risks associated with speculative bubbles in tech investments.
Conclusion This episode of Squawk Pod explores the fast-evolving landscape of technology and AI, emphasizing the necessity for businesses to adapt and innovate. The discussions with Gerstner and Mahaney provide valuable insights into the future of commerce, investment, and the broader economic implications of AI integration.
---
For more episodes, follow Squawk Pod on your preferred podcast platform.
Written by AI. May contain mistakes. Listen to the episode to check what was said.
Transcript
Automatic transcript. May contain errors.0:00Bring in show music, please. This is Squawk Pod and I'm CNBC producer Cameron Costa. On today's episode, all angles of tech. Brad Gerstner, founder and Silicon Valley investor, maps out our future and the future of our biggest tech players. This is going to be a free for all. Nobody's going away. Microsoft, Amazon, Google, Meta. Think about the 50 companies that come after them. It's getting harder and harder to raise funding if you're the next foundation model. And the age of agentic commerce is upon us. You can now shop at Walmart directly through ChatGPT. Internet analyst Mark Mahaney on the companies like Walmart, Spotify, and Expedia that are jumping aboard the OpenAI train.
0:52If those companies are seeing early on that they need to be integrated with ChatGPT, I think you're going to see a lot more companies come on. And there's a chance here that this dramatically changes the internet landscape. Plus, J.P. Morgan's Jamie Dimon weighs in on AI. And cooking oil and soybeans are now caught in the middle of a trade war between the U.S. and China. Two biggest trading partners in the world. We're going to find a way through this, one way or another. But this is as penuous as it's been since March. It's Wednesday, October 15th, 2025, and Squawk Pod begins right now. Stand Becky by in three, two, one.
1:32Cue, please. Good morning, everybody. Welcome to Squawk Box right here on CNBC. We're live from the NASDAQ market site in Times Square. I'm Becky Quick, along with Joe Kernan and Andrew Ross Sorkin. And here we go. What day is it? Wednesday? It's been a heck of a week. It's a good thing to start with. I think it is. Where am I? Who was it? Admiral, what's his face? Why am I here? Admiral Halsey? No, from the vice presidential debate way back when. Oh, yeah. Yeah, that's right. That's right. Stansfield. Am I wearing underwear? U.S. equity futures at this hour are up by about 200 points. Where am I?
2:11Where am I? Now I'm channeling Grandpa Simpson. Right? I see. Where am I? Why am I here? You better watch it. We're going to put you in a rest home. I'm already interested. We'll put you in the one on 60 Minutes. Speaking of. Our own 60 Minutes star from this week. In the New York. And Leslie saw it there on Monday night. Page six. Could be like Barton. I didn't see page six. Page six. I didn't see page six. I didn't either. Page six. Okay. Page six. Andrew Ross Sorkin. Big party. Oh, wow. Major. Yeah. Didn't mention. I can't believe I wasn't mentioned. He's one of the top stars. I figure it's because we're related to you, sort of.
2:51It's sort of a... I don't have a... I gotta get the paper. Here, here, here. I can't believe I have to do it. Sorry. It's a lovely little article. What do you look at first in the morning? I like to look at page six. I haven't had a chance to, and that's a lovely little article. We had a little book party, and that was very nice of them. Thank you, page six. I don't always look. I'm kidding. I don't look for it. But I did today because I figured they didn't yesterday. So I figured they'd have something in it. And by the way, they tell the nice little story about the Waldorf Astoria, which has this very interesting role in 1929.
3:27They go into this sensational part about how, you know, history, doesn't repeat itself, but it might rhyme and we might, you know. You'd make a good ghost. I'll tell you what was scary. So I'm doing the exchange, doing it again today. I enjoy it. doing here at 1 o 'clock, Kelly Edmund. Kelly's out, so you're ready. Kelly's out. Yeah. So I get there, and I look over, and I see Jay Powell talking, and I go, wow, that person looks like Kathy Wood. And I go, no way Kathy Wood would be interviewing Jay Powell. I mean, that's like animatter in matter almost. Cats and dogs living together. Isn't it? It's like the universe is supposed to end, and it really was her.
4:08But then I just started thinking about it. You know what one of her targets for Bitcoin is, the bullish target? 2.4 million. What is Bitcoin? What are Bitcoin and gold? What are those called? The debasement trade. Who's debasing the dollar to get Bitcoin? It just was weird that those two are, I mean, they're just totally oppositional. Wait, if you think he's debasing the dollar, you think he's taking rates too low. I'm not saying he is. I'm saying to get to 2.4 million on Bitcoin. But the other thing, did she bring up, do you think he was, you know, he's a very sober, serious individual. Did she bring up flying taxis by next year?
4:48Self-driving flying taxis by next year. Can I just say something about the job Jay Powell was doing? I heard some of his comments yesterday. And I like Kathy Wood. I do. And I like Jay Powell. I like Jay Powell, too. I have to say he has done a pretty remarkable job of just sticking to the facts and saying, here's what's happening right now. He's been under an extreme amount of pressure. But watching him say, look, we looked at the jobs numbers. It was much worse than we had anticipated with the revisions. Now that puts us in an equal balance. And he's always just sounds so straightforward, stayed.
5:19He's looking at the facts. He's not letting other stuff influence him. He's not letting, he's not getting mad. No. He's pretty even keeled about all of this stuff. And I'm grateful to have him. He's got a cool job. I mean, who's the most powerful person in the world? That's why I was surprised that - He's probably the second most. Probably. And that's why I was surprised that Scott Besson wasn't interested in doing that job. Now, I do think that as Treasury Secretary, Besson has a much more important role than other Treasury secretaries have. He's involved with everything and anything. It's a more engaging role.
5:50Yeah. I think when you're the Treasury Secretary, you're flying around the world. You're negotiating with people. You're trying to make deals. You're trying to get stuff to happen. Not always. This Treasury Secretary is. Right. Well, you like him, too. In a perfect world, the Fed chair can actually, hopefully, just hang out, look at some data, talk to people, try to understand what's happening, but it's not... A little more ivory tower-ish. A little bit. Yeah. But not disconnected. I don't want to suggest disconnect. One of my favorite times is when they're standing in that new Fed building. Trump's going, you know, now I'm finding out there's an additional.
6:20And it was a building. And Jay Powell's like, no, no, no, no, no, no, no, no.
6:29Let's tell everybody what's going on this morning, specifically with China, Trump tariffs. Here we go. President Trump firing another shot in the U.S.-China trade battle. On Truth Social, before the markets closed yesterday, the president called China's pullback on soybean purchases an economically hostile act. He said his administration is considering cutting off cooking oil-related trade with China, as well as trade covering other areas. Now, earlier in the day on CNBC, U.S. Trade Representative Jameson Greer said that the president could still slap China with a 100 percent tariff in two weeks or sooner, depending on Beijing's next move in a dispute over rare earth materials.
7:10Now, according to Greer, Trump and Chinese President Xi Jinping are still scheduled, still scheduled. This is apparently still going to happen to meet an economic summit later this month, though he said that could change. So we will see. So how much of this is posturing? Yes, but what I can't figure out is, so we had the Friday situation where the president said, well, first you had what China did. Then you had the president say, well, now we're going to go to 100 percent plus. I thought it was 150 percent, 100 percent on top of the 55 percent. I think this still might be that, though. I thought that's what then, of course, he seemed to walk it back a bit.
7:45The market seemed to get calm. But and now we're back. I'm not sure which is it. Well, talking, who was it? Tobin Marcus yesterday said, these things, the market thinks it's going to be like previously, it's the two biggest trading partners in the world. We're going to find a way through this one way or another. But this is as tenuous as it's been since March or April. And it could, you know, why are you laughing? It could get out of... You want to know why I'm laughing? No, I'm not sure. I'll tell you why I'm laughing. Is it anti-Trump? No, I'm going to tell you why I'm laughing. Okay, tell me why you're laughing.
8:17Because, as you may know, earlier when you were handing me the newspaper, I spilled a little coffee. I spilled the coffee all over the floor. And we just had a mop come out here. Over here. And I saw the mop. And so I figured, oh, here we go. Did you spill on your pants? No, I'm good. It's just all over the - You just destroyed the rest of the set. Yeah. Yeah. And we're going to have to do a little bit of work. You got so excited about paychecks. Send it back. He wants his paper. He wants his paper. Don't you have the post-it in? I can't find my post-it. Is that not on your reading? Oh, no. The Post is top reading.
8:49It is. Okay. 100%. Most fun newspaper to read in the morning. It's a serious publication. It was Alexander Hamilton, for God's sakes. Yeah. Okay. You think it's more fun? It's the most fun to read. Come on. Right. You don't think it's a fun paper to read? I do think it's a fun paper to read. I just know what you're implying. I'm not implying anything. You're implying it's kind of frivolous, not a serious... No, I'm not implying this at all. Right wing. It's Rupert. It's, you know, MAGA, MAGA, MAGA, MAGA. I wasn't saying any of these things. I wasn't even thinking these things. Okay. I enjoy it.
9:26Do you enjoy? I don't enjoy the New York Times because it's what I need for true news. It's not something that it's enjoyable. It's like castor oil. You guys know what next Thursday is? It's like castor oil. What's next Thursday? National Slap Your Coworker Day. Oh, my goodness. I'm not kidding. I'm calling in. I wondered. I'm calling in. Wow. I didn't even know that existed. It is. It's been around for five or six years at least. It probably would have happened by now. Don't you think? How long have you been here? 15 years, right? Yeah. I'm not saying who's the slapper or slappy. It's nice to see this one.
10:00Well, you remember, was it David Zaslav? Who had that show, The Slap? Yeah, that terrible show. Oh, that terrible. Yeah. Yeah. We're not going to do that. We're not fans of that. Don't blame it on Zaslav. We had people eating bugs and, you know, right? Wasn't that NBC? It was. It was. What was the name? Fear Factor. Fear Factor. JP Morgan CEO Jamie Dimon saying it's important, he says, now to look at AI projects individually when trying to decipher if there's a bubble-like activity in the sector. In a wide-ranging interview at the Fortune Most Powerful Women's Summit in Washington, Dimon was asked about what some have called a circular style of investments that big tech companies have been making in each other.
10:43It's not abnormal to have vendor finance, which is what you seal about by part of your company. You get the cash off financial purchase. You got to go one by one to say, is it pushing the bubble or is it real? Is it really going to develop stuff that will have product productive capability to pay off on the investment? Like I said, in total, I think they probably will. But one by one, I don't know. It's an interesting question. We've been talking about this sort of circular. Some people call it round tripping. And you're right. look, if an airline buys a Boeing jet, oftentimes it's finance. Vendor financing is a thing, and it can work.
11:20The question is, of course, these are companies that historically have not made money yet. And that's, I think, the harder part about getting your head around it. He did say that he's a little bit worried about some of the guardrails, though. That was one of the, just in terms of the AI, the regulations around AI. So that's another piece of the puzzle. And it ended with, but he was blunt when he said that technology will eliminate jobs and people should stop sticking their heads in the sand. It's true. Most of these big CEOs will admit that, and I think what he said recently is if J.P. Morgan is successful, they will have more employees five years from now, ten years from now, than they do right now.
12:04But there will be positions that have fewer headcount, that they will continue to grow. There may be more headcount, but not in some of these areas where you see headcount today. It's just in the past when there's disruptive technology and whatever comes about always employs more people. I don't. Andy Jassy said they'll have fewer employees. I don't think at this time. I don't think I think truly that Luddites could be right this time. This time is different. I do think this time is different. What I will say is this revolution is happening far faster than any industrial resolution. And we all might be, you know.
12:41A hundred years' work taking place. No, that's like the movie. We may all be on UBI. There may be so much productivity enhancement. That's not a bad life as long as you don't get so fat and lazy that you... As long as you don't just eat buttered popcorn. Right. That's what we were talking about. Cheese will be next. Next on Squawk Pod, internet analyst Mark Mahaney joins us. The future of open AI and the impact it'll have on the online world as we know it. This is like Mark Zuckerberg in private. Build out the user base, build out the engagement and charge them. But a lot later, I don't think they're going to really lean into monetization until they got 2 billion users.
13:18Don't panic. Mahaney says we still have a year or two before that happens. We'll be right back.
13:29You're listening to Squawk Pod from CNBC with Joe Kernan, Becky Quick, and Andrew Ross-Sorkin. Here's Joe. Walmart teaming up with OpenAI to allow users to make purchases directly on ChatGPT. Joining us now with more on the rise of agentic commerce and its impact on the tech sector, Mark Mahaney, Evercore ISI, head of internet research. We can talk about this. We can talk about a lot of the recent scuttlebutt surrounding AI. It's moving forward. It's unstoppable, obviously. You can talk about the Walmart deal or just all the bubble talk. I think there's a bubble with bubble talk right now. There may be.
14:08What I'm really interested in is you had a slew of announcements just in the last 10 days. You had Expedia, Uber, Spotify, Booking.com, Walmart. They're all integrating with ChatGPT. What a sign of the times. And so it is emerging. It's still small. $800 million is still small. You've got two to three billion person assets with Google and with Meta. But, you know, watch this space. And if those companies are seeing early on that they need to be integrated with ChatGPT, I think you're going to see a lot more companies come on. And there's a chance here that this dramatically changes the Internet landscape and the tech landscape, you know, in the space of the next two or three years.
14:50Well, we had that discussion, and we were told that it's going to be not just a zero-sum game for get involved so I can work with you. Oh, I will be. On Uber, that suddenly Uber's ride-hailing business is going to double because they're on ChatGPT. Why isn't it just people going to ChatGPT instead of doing it the old way for Uber? Why will the companies that are, maybe they need to be there to not lose business, but are they going to be able to add 20, 30, 40 percent to their business by just being on the AI? No, I think that's silly. No, but could it be? Who made that point? What's that dude's name?
15:28Remember the other day? No, but there could be incremental traffic. That's what's important. He's at your book party. I saw him at your book party. You were arguing with him about, what's the McKinsey guy? Oh, Michael Wolff. Yeah. Yeah. Remember? Yes, we were talking about all of his circular deals. You looked at him like he had three heads when he was saying it. Well, and then we had Jamie Dimon talk about it just yesterday in Washington, D.C. And what did he say? Did he say it's additive or it's just... Well, I think he was saying vendor financing unto itself is not... No, we're not talking about it.
15:56We're talking about whether Uber being on ChatGPT doubles Uber's business. Right. That's the ultimate question. No, I don't think it does. But does it... Actually, you should think about it more positively for a name like a booking, an Expedia, and eventually Amazon, too, probably, and a Walmart. This is just another customer acquisition cost channel. Who's paying who? Because we just had another retail analyst who was here saying, look, these retailers were all slow when the Internet first launched. They let all these e-retailers come in and eat their lunch. They don't want to get caught in the cold again.
16:26Are they paying ChatGPT for this, and is it a lot of money? Eventually they will. I think what ChatGPT is going to do is this is like Mark Zuckerberg in private. Build out the user base, build out the engagement, and charge them, but a lot later. I don't think they're going to really lean into monetization. They're just to say, yeah, we want to be everywhere. Until they've got 2 billion users, and I think that's when they'll start leaning into monetization. So that's still another year or two. If you were ChatGPT and you had unlimited access to capital, unlimited access to infrastructure, why would you try to screw anything up by trying to monetize?
16:55Now, wait, build up the traffic, build up the engagement. But the other side here is agentic commerce is coming. It doesn't mean it only occurs on ChatGPT. All of these, Amazon, all of these major platforms have to make— That sounds very monopolistic, what you were describing. We're going to let everybody come play with us. And as soon as we can, as soon as we have, as soon as they need us more than we need them, we're going to turn on the screws and ramp it up. Turn on the screws, charge people. I don't know. I sort of view that. I mean, this is kind of what a meta, think about this is what meta did.
17:24You know, you use stories or you use threads and just build up the traffic and then later on monetize. But first, build up the traffic, build up the engagement. But I think what's imperative for all these companies, you have to build out agentic commerce capabilities. You have to make sure that when I go to Uber, it's personalized. And I don't have to always go through the same prompts. Like, they should be able to automate it a lot better, personalize it a lot better. All of these e-commerce experiences, these transaction experiences, need to be agentified. And I think they will be, by the way.
17:52So there's not just one agentic commerce winner. There can be multiple ones. But companies need to make sure they invest in that technology now. And there's some companies that are doing a great job of this. You look at Amazon with this Rufus technology that they have. Like, they're trying to personalize. Look how much information Amazon has about you. They should be able to much better personalize the shopping experience. And I think they will. So, anyway, I think that's the challenge that's been thrown down. But going back to the issue I think that we were talking to Michael Wolff about. Vendor financing, yes.
18:18No, no, we'll take the vendor financing piece out of that. We'll come back to that in a second. Just given that this is going to happen, how much can it grow the market for these players? Or is it just a replacement? Oh, I'm sure there's some incremental growth. I'll just step back. Google grew the information, the search, the advertising economy over the last 20 years. Google made it bigger because it just made it more efficient for consumers and merchants and small businesses to meet up. ChatGPT is going to do the same thing. It'll be an accelerator. It's not going to double the market. The question, I think, is an accelerator for Uber?
18:51Is an accelerator for Lyft? Is it accelerator for everybody? Because then you're saying the entire pie is getting bigger. See what I'm saying? Yes, I do. I assume you can either go to a restaurant and you can use OpenTable for that, or you could use DoorDash to bring you the food. Well, I'd use the Internet analogy. If you didn't figure out how to service customers, how to interact with customers on the Internet, you were going to lose market share. I think the Internet and agentic commerce and AI is going to increase the overall size of the market, but it's going to create a lot of winners and losers.
19:21And the losers are going to be the companies that don't develop better offerings. It's like if you didn't have a website, eventually you got screwed. I think the same thing is going to happen here. I'm not saying that for Uber to maintain its market share, it's got to go on ChatGPT. It's not going to double its market share. Maybe Expedia, because people probably, if you could just do ChatGPT, get me an airline or a hotel room, maybe that. I don't know. Well, then Expedia, what Expedia has to do is make sure that people go to Expedia and they say, this is where I want to go, and this is my budget.
19:51Put together the package for me. And by the way, they're doing that. They have this product called Expedia Romy. Priceline has something called Priceline Penny. All these companies have to create these agents. You need to create an agent in order to be successful. And where do you land on the vendor financing piece of this, the sort of round tripping? It's getting tricky.
20:13And I'm not sure how I come out. There's not red flags here, but there's some orange flags. And you better be careful about who you've set up your vendor financing deals with. There's a few companies where it should be relatively safe. I think OpenAI is. I think they've got access. I don't know. I think they've got almost limitless access to capital at this point. If they maintain the kind of growth that they're doing, and if they're rolling out the products, the products that they're rolling out are extraordinarily good. SOAR 2, for example. ChatGPT 6 is coming out before the end of the year, and I just think we're going to see these step function improvements in a product.
20:46If you do that and you're in the OpenAI ecosystem, you're okay in terms of vendor financing. All right. Mark Mahaney, thanks. Thanks, Jim. More tech is up next with Altimeter founder and CEO Brad Gerstner. In 2022, right, only two and a half years ago, Andrew, AMD and NVIDIA had the same revenue. This year, AMD will do just over$30 billion and NVIDIA will do$225 billion. A hundred percent of the incremental market share to AI basically has gone to NVIDIA. AMD missed a wave. He's a longstanding name in Silicon Valley, investing through hedge funds and venture funds. He's a four-time founder and an active player in Washington.
21:29You do not want to miss this next conversation. We're talking AI and chips, of course, but also China, interest rates, and much more.
21:40Welcome back to Squawk Pod. You're watching Squawk Box right here on CNBC. I'm Andrew Osorkin on a Wednesday morning, along with Joe Kernan and Becky Quick. That is today, right? We were discussing where are we, what day is it? It is Wednesday. We are super excited to get to our next guest for a wide-ranging conversation on the markets, AI, bubble fears, China, venture capital, policy, and so much more. He's got his glasses, the famous red glasses. Brad Gerstner is here, founder and CEO of Altimeter Capital, of course a leading tech investment firm with notable investments in Meta, NVIDIA, Snowflake, Uber, Amazon, Microsoft, and others.
Read the full transcript
22:17Brad, welcome, welcome, welcome. It's great to be here. You're in town for the big Robin Hood conference. Thank you for joining us. amazing amazing organization you know I live in obviously Silicon Valley but what Paul Tudor Jones in that group has done to relieve poverty over 40 years in New York had an event last night over half a billion dollars announced for their new endowment so it's great they have an investment conference today I look forward to talking about it so here's the big question I imagine the word bubble is going to be raised at this conference multiple times if it hasn't already last night Are we in one?
22:50If we are, where are we? You mean AI? AI, some kind of economic bubble, market bubble. I don't know what you think it is. And I will say, you mentioned Paul Tudor Jones. He was on our air about a week and a half ago now. And he said he thought we were in some kind of bubble. And he put us on a comparable basis, he thought, of October of 1999, which was an interesting moment because the market still had 40 % to go. That's right. Well, first, you rarely have a bubble when the hottest book in New York is about a stock market crash, and everybody who comes on the show talks about a bubble. There's a bubble in bubbles.
23:28Howard Marks said, you know, it's too early because everybody's talking about a bubble. If you look at multiples as an example, you called out Bernstein yesterday I thought was great. He was talking about 50 times trailing PE for NVIDIA. It's 25 to 30 times on a forward PE. That's not the stuff that bubbles are generally made of. But certainly the further and further you go out on the risk curve, Andrew, you will find things that make you scratch your head. You say, I wouldn't have done that deal. And there are going to be meme stocks that certainly, you know, have no revenue, Becky, and, you know, in the quantum space or whatever, where you say that valuation doesn't make any sense.
24:04But we have this all the time. The bigger question is this. Is the phase shift associated with AI? is the transformation of all computing with now computer-assisted intelligence is as big as the Internet. And I would argue that it's as big as the Internet. And if you spend all your time hand-wringing, like many people did in 2001 and 2002 about the Internet, saying, I told you so, they missed the next 20 years of compounding in Amazon, in Google, and all these big windows. By the way, I would agree with you in that I think AI is going to change the world. It's a generational shift for all of us.
24:39The question is, could there be a hiccup along the way? Of course. What does that hiccup look like in terms of right now when you say to yourself the upside in this marketplace today versus the downside? We had an analyst on earlier this morning who said, you know what? It's going to be hard to reach materially higher numbers right now. But she kind of thought, you know, could the market be off 5 or 10 percent? She said that didn't seem totally unreasonable to her. We've had two 25 percent drawdowns in NVIDIA this year. Right. Remember the deep seek moment? stock was down 25%. And then on the tariff moment, the stock got down to 92 bucks a share.
25:15It's at 180 today. So volatility is the name of the game for two reasons. Number one, we have a president that's trying to reset global trade policy, national industrial policy in the middle of a super cycle. So that's just a lot of uncertainty. So I think that's healthy in a market, right? It's not straight up and to the right. We've had a lot of drawdowns this year. And if you look at these stocks and you say to me, does it make sense that NVIDIA is up 35 percent, given that their top line has grown way faster than 50 percent, the answer is yes. It's cheaper today than when the stock was$200 a share.
25:50But here's the real trick, and that is we can tell the difference between some crazy meme stock or a new crypto junk coin that we've never heard of and NVIDIA, right? But how do you as an investor find those gradations in the middle? How do you determine what's real, what's not, when it's not so extreme? I think it's a great question, Becky. If you look at how on the venture capital side of things, you see this barbell approach, and we're following it as well. We're really large investors in open AI, clear breakout market winners. On the other hand, we're investing in very small companies where we're buying 10 % to 20 % of the company.
26:31In the middle are these inflection growth assets, really high valuations, but haven't yet proven that they're the multi-year compounder. There are over a thousand of those in Silicon Valley, over a billion dollars that are trying to raise capital right now. That's where I think you're going to see a lot of the chop, a lot of the washing out. You'll also see that in the public markets, right? This is a moment of, you know, peak creative destruction, right? There are winners and there are losers. And that's why we see the volatility. How can Oracle, Joe, be up 25 % on earnings? It's a head-scratcher when they announce these deals because they're gaining hundreds of billions of market cap in a single morning.
27:10You only have that when you have this level of uncertainty about the path forward. Okay, so let's talk about the certainty that you think you have about the path forward. Yeah, let's do it. So you would say, who are going to be the ultimate winners in all this right now? Yeah. Well, I mean, I think on the consumer— NVIDIA, you think, is obviously a winner. Would you touch an AMD? Would you touch a Broadcom? Sure. We own Broadcom. We've owned AMD. What I would say is if you go back to 2022, and I showed this graphic on the podcast recently. In 2022, only two and a half years ago, Andrew, AMD and NVIDIA had the same revenue.
27:47This year, AMD will do just over$30 billion, and NVIDIA will do$225 billion. 100 % of the incremental market share to AI basically has gone to NVIDIA. AMD missed the wave. But to Lisa's credit, she raised her hand and said, I'm betting the farm. I'm going to give 10 percent of my company to open AI, and I'm going to bet that the MI450 is going to work and it's going to be competitive with NVIDIA. That's not exactly the position of strength that you want to be coming at it. But it's also a recognition. Lisa said we're in year two of a 10 year super cycle to build out the compute in America. If you're not on the AI wave, then you're not going to participate in all that upside.
28:30When you think about the big LLMs, OpenAI clearly seems like they're going to be a winner. Google, I don't know where you think Anthropic lands in all of this. I know you own Meta. They're trying to play this game, but playing in an open source way. Does that work? Yeah, so let's bucket it into consumer and enterprise. I think that's the easiest way. On the consumer side, I've invested in three or four super cycles now. the internet super cycle, the social, mobile, et cetera, the winner tends to be winner take most, right? So Google had a lot of competitors, 20 competitors in search. They all have pretty good technologies, but Google consumed 98 % of all the upside ever associated with search.
29:10Meta in the case of social, Apple obviously with the app store. I think chat GPT at one, you know, well over a billion and a half monthly users, over 800 million weekly users. And what most people don't talk about is the engagement. engagement is going up. So if you look at the cohorts under those numbers, they're extraordinary. And so I think they're breakaway now in consumer. Can Google catch them? I think it's very challenging. Can Meta catch them? I think challenging. Will they build good businesses around AI? Yes. These are money machines, the best businesses in the history of capitalism.
29:44They're going nowhere. I'm impressed by what Sundar is doing. I'm also impressed by what Alexander Wang and the team at Meta are doing, but they know how hard it's going to be to catch up in consumer. On the enterprise side, this is going to be a free-for-all. Nobody's going away. Microsoft, Amazon, Google, Meta, they're all, you know, Anthropic is probably one of the market leaders in enterprise today, certainly with Claude Code and the work they're doing with Cursor. So there are going to be a lot of players that are left in the game, but those that we just mentioned, Those are going to be in the hunt.
30:17Think about the 50 companies that come after them. It's getting harder and harder to raise funding if you're the next foundation model, right? If you're the next application to do coding. I think the market's sorting itself out pretty quickly into the market leaders. Okay, so what do you do about Apple? Yeah, listen, what's extraordinary about Apple is we all love these devices so much. Love them, love them. Right. And, you know, the staying power, they can miss a generation or two of AI like they have. And we still buy the phone. Right. So I think that what we all want is a super powered Siri on our phone.
30:53It needs to be lighter weight. Battery life is a real issue on these devices. In fact, some complaints about the new thin phones, the battery life, you know, isn't as great as it should have been. I think there's still this is really heavy. Right. It's the new one. Right. I think there's plenty of time for them to get in the race. I will tell you this. It's DEF CON 1 at Apple. It's existential at Apple. They're worried. OpenAI is coming out with a device next year. Right? Johnny Ivan team are now working with Samsung. And so what do you think the answer is for them? Can they build it themselves?
31:25Do they have to be people talk about buying an Anthropic? Does the government allow something like that to happen? So I would say two things. Number one, I think they've looked at everything acquisitively. The two market leaders are now too expensive probably for them to buy. The crazy thing is I actually think that the government would allow one of these players to buy Anthropic because it increases competition. Right. The level you just had two banks on Goldman's M &A revenue up 60 percent, Bank of America, 45 percent. We have not had M &A revenue in five or six years. So we've had a total transformation.
31:58There's a lot of consolidation in M &A coming. But I would say for Apple, don't count them out. They can bake their own stuff. They can use open source or they could do do an acquisition. One related question back to Meta with the open, not just the open source, but what they're trying to do with devices now, the glasses and everything else that Mark is pursuing. Do you believe that he can become a new hardware effectively platform or even a software platform with the LLM where you go to them for that purpose as opposed to what I think is being used for now, which is to power a lot of it? I think there's a kind of a meme running around right now that Zuckerberg is hitting the panic button and paying these people a billion or a billion and a half dollars to go work at Meta.
32:36I think that's a total mischaracterization. I think the core business at Meta is on fire, better than it's ever been. Huge beneficiaries of AI, when you think about Reels competing with TikTok, when you think about what's going on on Facebook. And if I'm him, I love the fact that you have a founder who's going all in, doubling down, tripling down on AI. And by the way, is it rational to pay a person a billion dollars? You know these companies do hundreds of tuck-in acquisitions you never hear of, right? Under a billion dollars. Now these super empowered individuals who can build a transformer model, think of Noam Brown, who invented inference time reasoning in the 01 model.
33:14One person can have that level of impact. Of course it makes sense to pay him that kind of money, right? Because it's not the legions of engineers that are inventing AI. It's the architects at the top who are helping you break away. And here's my guess for Meta. I don't think they're battling to be the leader in open source. I think they want to be the leading frontier model. I think it will ultimately be a closed source model that will benefit Meta. And I think they're making some smart moves. Andrew, we never got to where you said the promise of M &A and investment banking and IPOs. Remember how many times we said there's a promise, but it's not here.
33:53And maybe it's going to come. And it's coming. But it's only been 10 months. So it really did happen. We never, we crossed the Rubicon. you never really acknowledge we are in like it's party it's 1990 we're partying like it's 1990 we're partying like it's 1999 you never I will I will acknowledge right here and right now then it happened we are it's happening I don't by the way if you look at M &A activity I mean it's happening I think the kind of activity that you'd imagine but for so long I think you might have even been on saying you may not see it, but everybody's been called back to work that's in investment bank.
34:32On that point, is there a clock ticking where you feel like we better hurry up and do these deals if we're going to do them? Yes. Particularly the big deals because you've got the Trump administration and you don't know what comes next. Yes. I don't know if it's as much about that, but I do think that there's been a ton of pent up demand. Nothing like they sent the corp dev teams on these companies to the beach for the last six years. They're all back from the beach. So they pulled them back from the beach. They went to work. And now there are lots of deals in the pipeline, lots of deals, as we're seeing from these banks, right, that are being announced.
35:05So I think there is a sense of urgency, Becky. Before somebody else buys the stuff you want to buy or before regulators say, hey, hold I think the sense of urgency is because AI is moving at 10x the speed of the Internet, right? OpenAI got to a billion users radically faster than Google did. It got to$10 billion in revenue radically faster than Google is. I think there's a sense for every company that wants to compete, whether you're a chip company like AMD or whether you're a frontier model like Anthropic, now is the time. So it's going to continue to heat up. When we throw around all those numbers in these deals that we get, even for gig of power, it's a trillion dollars.
35:46And then we see what NVIDIA, how many chips. The money doesn't exist yet. And that's a question you keep asking. Yeah, that is part of the issue. Even the biggest companies, OpenAI, doesn't technically have the cash to do this. But they're promising it. But they're making the commitments to do it. We need Jay Powell. So I think there's a lot of hyperventilation about this particular issue. So let's distinguish kind of fact from fiction here. They are laying out a five to 10 year framework for what they're going to build. It gives you the T. But it takes a long time, Joe. If you deconstruct this by my math, this is like$150 billion of CapEx for OpenAI in 2030.
36:28In 2030. So now we have to forecast the revenue for OpenAI in 2030. Is it plausible or probable that they'll have$150 billion of revenue in 2030? I think it is, right? I think it could be more than that. And so the numbers that you're hearing, again, on trillions is a framework for a decade-long investment cycle. And it only gets spent under two conditions. Number one, the demand shows up, so consumers and enterprises all use it. And number two, the power companies and the chip companies deliver on their promise. Brad, I want to thank you. It's a very cool project. You've got to come back when that happens.
37:05Please do. Okay, thank you so much for this. that's the podcast for today thank you for tuning in squawk box is hosted by joe kernan becky quick and andrew ross orkin what's next thursday national slap your co-worker day oh my goodness not kidding i'm calling it you can catch this very happy family every weekday morning on cnbc starting at 6 eastern and you can get the highlights right here on squawk pod just don't forget to follow us wherever you're listening now so we're automatically on your podcast feed we'll meet you right back here tomorrow have a great day we are clear thanks guys
From the publisher
In the latest corporate integration with OpenAI, consumers can now shop at Walmart directly through ChatGPT. Internet analyst Mark Mahaney discusses the impact of corporate and AI partnerships and the future of the online landscape. In a wide-ranging interview, Silicon Valley investor and founder Brad Gerstner, CEO of Altimeter Capital, shares his perspective on China, the AI and chips arms races, and interest rates. Plus, soybeans and cooking oil are caught in the middle of the U.S.-China trade war, and Jamie Dimon has weighed in on AI.
Mark Mahaney - 16:06
Brad Gerstner - 25:24
In this episode:
Brad Gerstner, @altcap
Joe Kernen, @JoeSquawk
Becky Quick, @BeckyQuick
Andrew Ross Sorkin, @andrewrsorkin
Cameron Costa, @CameronCostaNY
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

