Palantir Surges as AI Demand Accelerates

22:11Ed Ludlow:Thank you, Yuhaira. The U.S.-China tech rivalry is moving beyond chips. According to sources, officials in Beijing are growing increasingly concerned about Anthropics frontier AI models, all ahead of a planned September meeting between President Trump and China's Xi Jinping. Blue Magazine senior tech editor Mike Shepard joins with the latest. I mean, the direction of travel on this story is slightly different to those we've been covering in recent weeks and months. But this is China being worried about the leading U.S. frontier labs and their models. Well, that's right. And what we are seeing is almost the other side of the coin.

22:48Some of the arguments that officials in Beijing are raising really echo some of the arguments that we've heard in previous years from U.S. officials when they've complained about certain tech products, including TikTok, for example, and the national security risks that they pose through data collection and other means of sharing information with an adversarial government. In this case, the concern from Chinese officials is that Anthropik's mythos and other powerful AI models could be used as offensive weapons against their economy in some fashion. And they didn't offer any specifics, but this is an issue that will be on the table very clearly.

23:27They have some misgivings over not having access to this model. But those are concerns that were shared by a number of U.S. allies in the weeks and months after Anthropic made that disclosure that its mythos system in April was so powerful that it could only be shared with a small handful of trusted partners. And access has since been expanded, but it still remains small. And it has continued to be denied to China, of course.

23:55Ed Ludlow:if you go back to when president trump visited china you know the hope of america's chip industry was that he would get some result for them so that they could more freely export lead edge chips to china the america's concern on the chips has been well we don't want china's military to have access to that technology um that will help them in their own national interest or in their own national security context what was the net result of any of that was it discussed in China? Will these topics be discussed in September here in the States? Well, there will be a meeting ahead of whatever summit takes place between President Donald Trump and his Chinese counterpart Xi Jinping.

24:34And at that, the lower level officials, probably cabinet level for the U.S., likely Treasury Secretary Scott Besson, who has taken an increasingly point position when it comes to matters of artificial intelligence and negotiations more broadly with China. They will discuss some of these parameters before the two leaders actually sit down here in Washington. And the concerns will all be on the table. It will be everything related to AI, including the chip curves that the U.S. has gradually been easing, but perhaps not to the extent that China would want. While the U.S. is allowing sales of NVIDIA's H200s, for example, to Chinese customers, China so far really has not given the green light in the other direction for those purchases.

25:18So we will see some questions raised there. There will also be questions about whether Chinese AI companies themselves are engaging in unfair practices. Anthropic, for example, has accused Alibaba, one of China's biggest tech companies, of using a process known as distillation to unfairly gain an advantage and essentially copy some of their models. And of course, it's important to note that in singling out Anthropic, Ed, China is pointing to a company that does not do business in China, mainly because the CEO, Dario Amadei, has singled out China as a U.S. adversary and a threat to national security.

25:54So all of this dynamic will be on the table as the two sides meet, even before the leaders sit down. We'll be wanting to see how they actually negotiate a lot of these intense differences over the way Chinese models have been developed and also some of the export and other curbs in place, including our rare earths.

26:15Ed Ludlow:Bloomberg's Mike Shepard thank you very much coming up we're going to get back to Wall Street's expectations for this week's big earnings SpaceX AMD both coming after the bell let's go back to Palantir the tech earnings story of the morning the stock up 27 % on track for its biggest jump its best day since February of 2024 $380 billion now in market clap big raise in sales and net income expectations for this year and it's the US commercial business that's driving absolutely all of it. Okay, it's halftime here in the beautiful city of San Francisco and the Bay Area. Don't go far. This is Bloomberg Tech.

27:09Ed Ludlow:Welcome back to Bloomberg Tech. There is some momentum in technology. I'm looking at the up for a fourth straight session, matching the run of gains that it saw in that first week of June. 2.6 % gain on the NASDAQ 100. A big driver of that is Palantir, literally the biggest points driver to the upside post earnings. Palantir itself having its best day since February 2024. The S &P 500, by the way, is at all time highs and technology is a big part of that story. But earnings continue, chips and rockets. Let's get to Bloomberg Tech's equity reporter, Carmen Reinecke in New York with what to expect later today.

27:45Ed Ludlow:Hi, Carmen. Hey, Ed. So it's the moment we've all been waiting for since June at least. SpaceX will report its first quarterly earnings as a public company after the bell today. So we know the stock is down from that 135 IPO price. It's about 10 % below, even though there's some green on the screen here today. And so what we'll really be watching for after the bell today is what the company says, It's not only about what it saw in the quarter, but its forward guidance. So much of SpaceX's valuation hinges on what it sees in the future. One thing we'll be watching for in particular is what it says about AI.

28:21Analysts on Wall Street have projected huge growth for its AI unit and see it taking up more than 80 % of revenue over the next decade. So we'll see what the company has to say about this, if that matches up and if these models change at all. We'll also be watching for results from AMD. And really what I'm curious about here is if the stock can keep its rally going. So shares are up about 20 % in the last week. We saw some weakness in the semiconductor space about a month prior. And so AMD has a really high bar set for it. After its last earnings result, it had solid projections and the stock rocketed to a record high.

28:59So the setup here is super important. We're going to be looking to see if it's made any market share in sort of its competition with AI processing and NVIDIA. That's what's top of mind, at least for me, what I'll be watching tonight. Thanks, Ed.

29:12Ed Ludlow:Through most common, Ryanicki, thank you very much indeed. Our next guest says that cloud revenue growth last quarter by hyperscalers like Microsoft and Amazon justifies continued heavy spending for the AI build-out. And we'll probably learn more about that with AMD's earnings later today. Let's get more. Selim Wu, Portfolio Manager on the Global Robotics and Automation Team at Lazard Asset Management, also PM on Lazard's Next Gen Technology ETF ticker, TEKY. Welcome back to the show. Great to be back. I mean, at the end of the day, if the spending is justified and the spending is going to continue to grow, AMD will be a beneficiary of that.

29:49Ed Ludlow:You know, I find the chronology of earnings interesting. What do you think we'll learn from AMD about the future? Yeah, I mean, obviously, the hyperscaler earnings that we have seen so far has laid out the groundwork that all the relevant hardware semiconductor value chain will tremendously benefit because we continue to see the demand cannot, you know, basically there's a supply that cannot meet the demand. So definitely really interesting to see how they're evolving that in terms of their revenue and their addressable market. They're seeing, especially for the CPU, that can benefit as a core driver of the growth in the inference demand.

30:27Ed Ludlow:AWS CEO Matt Garman was on the show yesterday, and I tried to push him on what the CapEx number for next year will be, whether it will be higher in simple terms. That's on the spending side. On the growth side, also what's driving it? And he was pretty clear it's a shift to inference. Training load's still there, but inference is now the driver. Do you see that in your research? Yeah, absolutely. A lot of companies are highlighting that the mix shift to more inference work growth, especially on the back of the growth in the AI agency work growth, has been a boon for the demand in the growth. Not just AWS, I would like to highlight that the same pattern has been seen in the two other important hyperscalers, including Google and then Microsoft as well.

31:17I mean, it's truly impressive how the three companies have delivered combined 50 % rate of growth just this quarter, which was a double the rate that we have seen just five quarters ago. Very strong validation for AI ROI, continued commitment to spending, because they're seeing extended backlog of the customer visibility, which is a really important parameter for investing in AI.

31:40Ed Ludlow:All three of those, Alphabet, Amazon, Microsoft are members of TEKY. If you were going to do a scorecard from last week's earnings across the hyperscalers in the week prior, who came out on top of those three? Well, it's hard to pick, but I would say Microsoft and Amazon definitely relative to the expectation. this quarter was truly impressive because all the three delivered something much better than expectation that was pretty impressive in my view and then i think people have to be comfortable and understand that by nature this large spending commitment has to be taking place off front but over time you are going to see the tangible numbers flowing through their core businesses and this quarter was fantastic to prove that point selena if there's something in aggregate across all of the companies that you track and that you include, it seems to be that demand is still running ahead of supply, be that cloud capacity supply, literal chip supply on the basis of the compute.

32:43Ed Ludlow:Is that a good thing? And how long do you sort of see that lasting for? It's a great thing. But at the same time, it's a balancing act that entire supply chain has to work around to make sure, yes, when there's a strong demand visibility, they have to make sure that they can supply all the necessary components in time so that growth does not taper off. I think it's often underestimated how much amazing job this entire supply chain is doing just to make that happen. For example, the LTA that is taking place in memory space, for example, I think is a great example to make sure that the long trajectory for demand and the growth should be there.

33:24And that is something truly different and unprecedented this time.

33:28Ed Ludlow:So let's linger on that for a second. Long-term agreements in memory are atypical against history, right? What's happening, you know, based on the conversations I've had recently with SK Group and as well with Jensen Wang, is it's the customer saying, hey, let's do a five-year deal. That hasn't happened historically. Just, again, what does that signal to you? Why is that noteworthy? It is noteworthy because it basically proves the point that there's extended visibility from the customer demand and there's basic preemptive efforts to make that happen so that supply and demand go hand in hand in balance to make that this truly impressive one single generation growth opportunities coming from AI.

34:11We have to remember that all the visibility is actually coming from the upstream. If you think about TSMC, for example, the largest foundry in the world, I would argue that this is the company that has the best visibility when it comes to customers. And guess what? They just announced committing to additional$100 billion in CAPEX to expand in Arizona in the U.S. In conjunction with hyperscalers, I don't think they're spending money blindly. They're talking about they can actually see the demand. that can stretch out to the five years. That's why they're committing to spend today, and that's why they need to make sure all the supply chain, including the memory companies, to just pace together to make that supply and demand come in balance.

34:52Ed Ludlow:Selim Wu, Portfolio Manager at Lazard Asset Management. Great to have you back on the show here in San Francisco. Thank you. Now, coming up, we're going to speak with Palo Alto Network CEO, Nikesh Arora, about the next generation of AI-powered cyber defense, but actually I think a whole lot more than that too. The CEO of Hugging Face will also discuss the risks posed by rapidly evolving AI models. This is Bloomberg Tech. Sometimes we ask agents to think outside of the box, but we don't want them to think outside of the sandbox. This is the Bloomberg Tech Minute brought to you by ChatGPT. Now with ChatGPT Work, I'm Carol Masser.

35:34Globetrotters hunting for airfare bargains are in for a rude awakening, as the days of stumbling across a cheap seat on a popular flight could soon disappear. Bloomberg's Juan Ha reports that airlines from Delta to Virgin Atlantic are adopting artificial intelligence to change seat prices more quickly by weighing dozens of variables in real time, helping capture more revenue while shrinking pricing gaps that once allowed travelers to find bargain fares. Machine learning models can more accurately forecast demand by analyzing historical booking patterns, seat inventory, and seasonal trends, while also continuously tracking competitors' fares and capacity changes to update prices in near real time.

36:14The technology could lead to higher fares on busy routes as airlines pack flights closer to capacity, but may also result in lower fares on off-peak and lower-demand routes. That's the Bloomberg Tech Minute brought to you by ChatGPT. Put ChatGPT to work on your most ambitious ideas and projects. Get started at ChachiPT.com today by selecting work mode, available on Plus and Pro plans.

37:00you close deals faster with revenue agents and automations working around the clock. You'll have everything you need to scale your go-to market efforts. Elevate your wins with Adio. Start your free trial at adio.com slash iHeart. Aging is real, and so are the benefits of new Vital Proteins Collagen Sparkling Water. Because around the age of 30, your body needs backup to keep your collagen up. So get your daily glow up now in three fresh flavors. Strawberry blossom, lemon lime, and blood orange. improved skin health in as little as 30 days thanks to collagen peptides cheers to that so you can stay vital stay you visit vitalproteins.com to learn more and where to buy these statements have not been evaluated by the food and drug administration this product is not intended to diagnose treat cure or prevent any disease does ai make cyber attacks faster

37:47Ed Ludlow:more sophisticated palo alto networks is launching pan os 12.2 series a major software update featuring advanced virtual patching, an AI-powered tool designed to identify vulnerabilities and stop attacks before traditional software patchages are available. Joining us now, Palo Alto Network's chairman and CEO, Nikesh Arora. This is basically like an AI-powered series of network agents, right? Let's start with what they do. But I've found, over the last 12 months, where AI and cyber cross over. The best question is to ask, what is the problem you're trying to solve for here, Nikesh? Great. Well, nice to see that.

38:31Look, first and foremost, as you've seen, there's a lot of conversation in the market of AI's capability in the space of cybersecurity. What we're doing is effectively sort of rebuilding the operating system of network security for the future. I mean, imagine having a car in the past and imagine wanting to create the next Tesla. We're basically creating the plumbing in our network products, which allows to build a Tesla-like experience for cybersecurity where our products will start to do a lot of things for you using agents, using AI. And that's kind of what the fundamental of Ciri's launch is.

39:06In addition to that, I think one of the biggest features we have as part of Ciri's is, as you've all heard about, all the AI models beginning to show vulnerabilities and start allowing attackers to attack faster. Well, the average time it takes to patch vulnerabilities is 55 days. We don't have 55 days when AI finds a vulnerability. AI wants to attack or exploit the vulnerability as soon as it can. So this release allows us to take a lot of the patches where vulnerabilities will be found and build patches in minutes and hours so we can provide patching capability to our customers in a short period of time to eliminate that 55-day wait that has traditionally been the industry average.

39:46Ed Ludlow:in what you published this morning there's a lot of background and research onto what you you were looking for and in one case in a review of open source software your team talked about sort of 14 000 vulnerabilities what you were just talking about is essentially zero day vulnerabilities right the window between um discovery and exploiting of vulnerability is now zero um that is a surmountable challenge yes yes yes please no i think look you're right the the possibility that AI shows you, whether it's the new models you've seen from OpenAI or from Anthropic, you see that AI agents are able to discover vulnerabilities, come up with attack paths, figure out how to concatenate a bunch of vulnerabilities and attack a customer's infrastructure.

40:30Now we need to figure out how to respond to that at the same speed at which AI is discovering these vulnerabilities. We don't have time to pack. So what we've done is we've basically built the capability in our software and we have a research team researching vulnerabilities that we build patches as fast as we can. We've gotten the patch process down to four hours. We expect to bring it even shorter with products like what allow you to create instant patches. But that shortening from 55 days to four hours is a big step, which is what part of the release that we just launched allows us to do.

41:02Ed Ludlow:The biggest story of recent weeks is two advanced open AI models escaping a sandboxed environment, gaining internet access and mistakenly accessing Hugging Faces platforms. And lots of people in your field... Mistakenly, you said? Mistakenly. I'm being careful with my choice of words. People in your industry basically put this as a watershed moment for AI in cyber. I really want to know what you feel about it. As you can see, a lot of frontier model companies are flexing. They're showing the immense capabilities of what AI models can do, and this is true. that we had an incident where an AI model using agentic capabilities was able to escape a sandbox and be able to go ahead and attack sort of infrastructure of effectively hugging face to go look for exploit-gen sort of capability.

41:54So I think what it tells us is one, AI models are going to be more and more capable over time from a cybersecurity perspective. I think, too, it's a lesson for our friends at Frontier LMS to make sure that before they create captures of flag exercises to go find something in their external infrastructure, they should probably make sure that their sandboxes are secure and they are not suffering from any vulnerabilities before they start pointing the models outside. I think from a more industry perspective, as you saw, what we have to do is we have to start fighting this notion of AI with AI on the defense side.

42:25So a lot of work has happened on the offense side, and offense is always easier. I think we have to spend a lot more effort on making sure that there's defense capabilities. And I said that in the past. I said four months ago that I expect this capability to be around. in six months, or looks like it showed up in four months. So we have now seen attackers can actually effectively take these models and the capabilities that Frontier LLMs provide, which I think is going to become commonplace over the next three months, where we'll be able to find open source models out there which will have similar capabilities as you see those models getting distilled.

42:56So I think it's incumbent upon us to make sure we build the defensive capabilities while we have the time to be ready for these attacks to get faster, more efficient, and more quicker.

43:09Ed Ludlow:Nikesh, what is the token economics consideration where AI is increasingly used in the form of swarms of agents in defense? Tokens I said a few weeks ago that I think tokens are overpriced. I think part of the challenge is that the whole host of AI infrastructure for consumers is being paid for by enterprise. I think token prices need to come down to the tune of 80 or 90%, which will allow all of us to both both use AI effectively for products that we're building for our customers, or perhaps for doing drug research, or perhaps making organizations more efficient. So I think we'll see that. We already saw this past week that some of the frontier models have given us better pricing from a token economics perspective.

43:52Now, you will see a lot more tokens being used. You will see a lot of us use tokens in our products. You'll see us use them in our business processes. I started with somebody yesterday. I think it's not unreasonable to expect that 10 % to 15 % of our OPEX in the next 10 years is going to move towards AI or IT or tokens. So that's a big number. 15 % of OPEX will effectively move. And I think that's where we're headed from a token economics, token math perspective.

44:17Ed Ludlow:Nikesh, finally, before we let you go, there are reports that you are leading a group or a consortium to purchase an NBA franchise in London. What can you tell me about your plans for that, please? I think it's very important balance work in play. At not getting any younger, I need to make sure that while we're working hard to make sure we do what we do in our day life, we have to make sure we find time to be able to follow our passions. My son's passionate about basketball. I'm passionate about basketball. We have a cricket team in London, me and a bunch of my friends. I think part of the conversation was can we actually see if there's another opportunity to make sure that we can make basketball relevant in Europe as part of the NBA's efforts.

44:57So we have an aspirin bid in there. We'll see. there are many people who are interested in the asset and hopefully the best team will win

45:05Ed Ludlow:Nikesh just asking for I guess the Bloomberg Tech audience how much does an NBA franchise in London set one back I think what does it say what's the price of something what somebody's willing to pay for it so we'll find out Nikesh Arora Palo Alto Network CEO and chairman it's great to have you back on Bloomberg Tech thank you very much he didn't say no let's stick with cyber security Hugging Face CEO Clem DeLong told Bloomberg that the recent OpenAI linked cyber attack could have been much worse and that the use of open-weight models was crucial to defending the platform. We defended ourselves with an open model, right?

45:42And some of the guardrails prevented us from using Frontier APIs to defend ourselves. So kind of like improving the tools for defenders, you know, instead of obsessing about not giving them to attackers. I think it would be a good thing in the future to make sure these incidents are not too harmful.

46:02Ed Ludlow:Okay, coming up, another tech-focused hedge fund sinks as results of July's AI Link sell-off. We'll discuss what's weighing on KOTU's funds next. This is Bloomberg Tech.

46:26Ed Ludlow:We're still keeping an eye on shares of Palantir up, almost 27 % on track for their biggest jump since February of 2024. A big raise to the outlook, both on the top and bottom line for this year. Lots of strength in the commercial business in the United States. In fact, most of the growth coming from the United States. Internationally, a little slower, 33%. But there's a lot of potential here for Palantir and the market very excited about it. Let's talk about hedge funds. Last month's AI sell-off has hit another tech-focused hedge fund. KOTU Management's fund plunged over 8 % in July. It's worth a month in a year.

47:01Ed Ludlow:But the fund remains up 14.3 % year-to-date, according to a source. Let's get more on what's driving it and the impact from Bloomberg's hedge fund reporter, Hemma Palmer. You know, what happened in July? AI and chip stocks sold off pretty heavily. Is that the story here? Yes, exactly. It's a concentrated AI bet. CO2 has long been bullish on the sector, and four of its biggest holdings in the trade have worked against it. You're seeing huge declines on Taiwan Semiconductor, GE Vinova, LAM Research, and Applied Materials, some of them sinking as much as 29%. So this is part of what pulled down the fund's performance to the worst performance in over a year for the month.

47:48But yes, as you mentioned earlier, the fund is still up 14 % on the year.

47:55Ed Ludlow:The Philippe Lafont thesis is that there is a super cycle in AI coming. Any deviation from that after July? I can't imagine that he would have changed his perspective, which he's held on to for quite some time. He has experienced volatility earlier this year. They were down 5 % in March, given the tumult in the Middle East. They also posted a return in May, its best return in like 25 years when they gained 14 % in that month alone. So, look, it's been a roller coaster for investors in this fund. But you do know what you're getting when you invest in Co2. And it is this sort of long-held vision that the future will be AI-driven, that there is this super cycle to come.

48:41And I can't imagine that this one month would have changed his perspective. Who knows? He may have bought the dip as well. Right.

48:48Ed Ludlow:Let's go back to the year-to-date game, 14.3%. Like, where does KOTU sit in the sort of group of other hedge funds that you've been able to track? Yeah, good question. So when we look at the spread of the sector, we have Tiger Global, KOTU, Viking, Lone Pine. We call these firms the Tiger Cubs because they came from Julian Robinson's Tiger management and have spawned their own firms. And so when we look at, say, there is some divergence, right? Viking Global, for example, a lot more cautious on the AI sector, really pulling back those shares, not as exposed. That would have helped them in a month like last month.

49:26But they've also said that that's meant some missing out on some gains that other firms have really taken advantage of, like KOTU, like Tiger Global, and to some extent, like Lone Pine.

49:38Ed Ludlow:A conversation for another day, but I'll also point out we see KOTU a lot on the cap table and some private rounds and private companies too. Bloomberg's Hammer Palmer. Great job. Thank you very much. That does it for this edition of Bloomberg Tech. We have more earnings after the bell today. AMD, SpaceX, its first quarterly earnings as a public company and first call. A real big focus on Elon Musk and whether the question of a SpaceX Tesla merger gets posed. Also Paramount and then tomorrow Uber and Disney are ones that we're watching as well. It is a session that's kind of been dominated by Palantir's earnings, but basically stocks at all time highs on the S &P 500 technology is leading the way.

50:20Ed Ludlow:How does earnings impact that? So much to recap on the pod. You know where to find it on the Bloomberg platforms as well as online on Apple, iHeart and on Spotify. Let's keep it going. This is Bloomberg Tech.

50:37When you're running a business, the best days are the ones where priorities stay on track. For midsize and large companies, risk can affect multiple parts of the organization at once, from property and liability to cyber and regulatory challenges. At that level, managing risk becomes an ongoing discipline. At the Hartford, the focus is on helping businesses manage risk before it turns into something more disruptive. And when losses do happen, that work is paired with insurance coverage shaped by years of underwriting, risk engineering, and claims experience. Learn more at thehartford.com slash risk mitigation.

51:12Policies provided by Hartford Fire Insurance Company and its property and casualty affiliates, Hartford, Connecticut. Aging doesn't stop, and neither should you, with Vital Proteins Collagen and Protein Shakes. Because around the age of 30, your body needs more support for movement and recovery. On workout and rest days, reach for a 30-gram total protein shake or go with our classic collagen peptides. Help support healthy hair, skin, nails, bones, and joints so you can stay vital, stay you. Visit VitalProteins.com to learn more and where to buy. These statements have not been evaluated by the Food and Drug Administration.

51:44This product is not intended to diagnose, treat, cure, or prevent any disease. Everyone's talking about how AI is transforming work, especially in sales. While the landscape shifts, one thing remains the same. The thrill of closing a deal. Whether it's a gong or a confetti machine, every team has its celebration rituals. Adio is designed for that moment. It's the agentic CRM that turns customer signals into actionable insights, helping you close deals faster with revenue agents and automations working around the clock. You'll have everything you need to scale your go-to market efforts. Elevate your wins with Adio.

52:15Start your free trial at adio.com slash iHeart.