Power Lunch - Dow drops 300 points as investors await trade deals, Fed decision 5/6/25

Episode Date: May 6, 2025

Stocks slipped Tuesday after President Trump’s shaky commentary on global trade deals, dashing hopes that progress will soon be made on the tariff front. We’ll tell you all you need to know.Plus, we’ll hear from the CEOs of some of the biggest companies in the world right now: Nvidia and Palantir. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

And welcome to Power Lunch everybody. I'm Brian Sullivan in Beverly Hills, California, the Milken Institute Global Conference. And we are about to be joined exclusively by CEO of Palantir Alex Carp, billionaire Thomas Tull. They just signed a big deal with Elon Musk, X-A-I, Kelly. They are here. We're going to figure out what's behind it. Talk more about the deal. It's all very exciting. It is. And we'll get right to it. Brian, let's just do a quick market check first. I'm Kelly Evans at CNBC headquarters. And that's not the only big interview coming this hour.

We'll also hear from NVIDIA's CEO Jensen Wong and Service Now's Bill McDermott. But let's start with a look at the markets, which are lower across the board right now. The Dow's down 406, the S&P, the Dow, the NASDAQ, all down about 1% a little less than that for the S&P 500. And we're 24 hours away from the Fed decision on rates, and there is a major change among our mock Fed. Five of our panelists say the Fed should not change rates tomorrow. Don Peoples once again calling for that 50 basis point rate cut. But now, Jeffries, Chief Markets, Chief Markets, strategist Dave Zervos joining him and calling for that upsized cut. And in fact, we've got Dave

standing by now to explain. Dave, really appreciate. Listen, if you saw a quarter point, I might even be on board, but about a half point. Kelly, I think we learned a lot in the last intermeeting period. We learned a lot about where tariffs are headed. We learned a lot about the rebalancing that this administration wants to take on, which is potentially disruptive in as big a way as we've seen. We just went through probably the most complicated April of my career and came out reasonably unscathed in the end, but it certainly was touch and go there in the second and third week. I think we have a lot of businesses expressing the word uncertainty at the Milken Conference, which I just left. And I think it's a time where you can move a little bit back toward a neutral

policy. I think you've got all the information. One of the reasons that I think we were all waiting a little bit is we didn't know the priorities and how aggressive this administration would move. And clearly, they moved very aggressively on tariffs and they moved very early. So I think we need to think about that when managing the downside. So in other words, you think the economic hit is actually going to be so large that a half point cut is necessary to more or less support the economy. I think it's a risk management. You know, Kelly, you always catch me in these because I'm really not a believer in talking about what I think the Fed should do. I always try to tell our clients what I think they're going to do. You put me in the position on this committee to talk a little bit

about what I would do if I was in that position. So take it with a grain of salt. I'm not saying that that's my prediction for the Fed. I don't think they will. But I think they could be quite dovetish. And I think they have a lot of reason to be devish. A lot of the intermediating movements in inflation expectations related to break evens came down substantially. And a lot of the sentiment that I've seen with our corporate clients, our CEOs, the companies we deal with is just pullback. And I think as a Fed, as an FOMC member, you're listening to that. As a Fed president, you're listening to that locally. And I think it's just a prudent risk management story in a position where we know tariffs are one-time hits to the price level. They're not sustained inflationary.

What about the ISM surveys where we've seen kind of those prices paid measures be sticky? And again, I'm trying to argue it from a point. of what will they do? Because I think they all lean, frankly, towards being slightly more hawkish and less inclined to cut rates right now, especially now that the jobs report and the ISM surveys have been okay.

I bet they go, why would we take on that risk? They never want inflation to come back again. So I just think, again, kind of from a real politic point of view, wouldn't the risk be on the side of hawkish for tomorrow? I think you could definitely make that case. You can definitely make the case, especially when the Fed kind of sat back in the second week of April, let the tenure know,

have one of its biggest weekly rises in 30 years yields rising more than they had risen since the 1980s. And they didn't really say anything. They didn't seem to care. So I think there is a disconnect between the administration and the Fed. I don't think there's a lot of love loss there, surely. And I think when the president is prodding Jay to cut, it makes it harder for him to do so as well in his own mind.

So there's a lot of things stacking up to say the market is probably predicting something more hawkish. So I don't think what you're saying would be a surprise. I think the surprise would be if he actually says, you know what, a lot of what we're hearing out there is a little bit worrying. And yes, the data are okay now. Yes, the consumer seems fine. But we're about to rebalance our fiscal and our trade deficits in a very big way.

And if you're a prudent steward of monetary policy and you look at maximum employment and price stability over the long run, there are some real dangers during that rebalancing that you want to mitigate again. and you can mitigate against and you have the ability to mitigate it. I think going forward we need to ask everybody what they would do and then what they think the Fed will do. That gap would be fun to show everybody as well. That would be fun. David for now, thanks very, very much. David Zerbos of Jeffries.

Brian, back out to you. All right, Kelly, thank you very much. All right, welcome back to the Milken Global Conference here in Beverly Hills. We've got a big crowd behind it's here. Now, shares of Palantir, they're down today. They had their earnings last night. Most people said the earnings were good. So what exactly is going on?

We'll talk about that, but the real news is this. There is a big deal that is being made in AI today between Elon Musk's X-A-I, TWG Group, which is Thomas Tull, Mark Walter Investment Group, and also the aforementioned Palantir, and we're joined now by Thomas Tull and the CEO of Palantir Alex Carp in a Power Lunch exclusive guys.

Thanks for patience. Appreciate you being on. Why do this deal with TWGXAI? What does Palantir get out of this? Well, first of all, Thomas is a legend. in the investing community and almost all the successful defense tech startups have been financed by him and says a reputation second to none and I know his work and his reputation in the defense tech startup world which we know very well and without him many of the great startups

that come after Pallenture would not exist the mission of Pallentier is to give an unfair advantage to our friends and if you look at what regional banks have been satisfied with regulatory hurdles and deep technical issues, almost all which can be solved by the combination of his insights into banking, which he's one of the most successful people in the world at, our tech infrastructure and large language models and compute. And that combination will bring the unit economics

of a company like Pallentier to regional banks and allow them to do what big banks are claiming to do and grow their revenue at much lower cost. Those are very, he's very nice, Thomas. I mean, very complimentary. But here's the sad reality of AI. Every company wants to say, I've joked, Alex, we're going to change her name to CNBC AI because you get like a valuation, pop. AI is really good at taking money.

It's not been very good at making money. I think that's exactly right. You get pitched things like that all the time. And in our case, we hired Drew Kukor, who's a very well-known practitioner and artificial. intelligence, built J.P. Morgan's systems, and built out our team. And before we put this together, we use it at Guggenheim, we use our platform on all our systems, and it's made an impact, a material impact. And everything we talk about is, if you made a dollar before, are you now making more than a dollar, or are you just talking about it? So that's what we're...

Most people are just talking about it. This is a thing. You've got to judge a company or an institution by the fruits they bear. And the thing at Pallentier is we're quite good at making money. So we're highly profitable. There's a rule of financial hygiene called a rule of 40. I think the highest ever recorded or one of the highest ever recorded 83. We make money by being downstream of value creation. At Pallenture and with our partners, we own value and we take a part of that value and get paid on it. So we're not going to be making any big money unless the people we work with make even more money. And one of the problems of the AI revolution is it's a tale of two cities. It's a tale of people who take your

money and give you flowery language in. In the past, it was woke. Now it's like you get to feel good about something that's not true. Or you can partner, and we partner together in his bank. So you can already see the value creation. And we see this across every industry where the unit economics change, which is why Palantir is growing 71% in the U.S. commercial. We're going to get to that in a second, but I think I would say this. It's fair to say that AI, Thomas, is like the seed. We're planting a seed and hoping it grows into a tree. With this deal, you're kind of already buying the tree in a way. Pallantir's built out. XAI. They've got the infrastructure sort of built out. That giant processing center in Memphis, Tennessee.

Lots of land, lots of fresh water, by the way, lots of electricity and stuff it takes. How quickly does this put you sort of, quote, in the game to sell money-making AI to Wall Street and insurance companies and more? Right now. Because, again, we have been doing. this, you know, Mark built an incredible financial empire. Mark Walter, who... My partner, Mark Walter, for 30 years of doing this, has an incredible deep knowledge set around finance, structure, etc. We know that world deeply, and we know what's possible now

because we've done it. So we're not asking others to do anything that we haven't done ourselves. So by this year, we'll be out with major customers across banking, banking, insurance, money management, and I think by having Pallenteer their tool set. You already have customers interested? We absolutely do, yes. Already. And you just announced it this morning.

Yeah, I mean, we've been working on this for a while. You better have some substance by the time you announce this. We're very excited about it. And with XAI, not only do you have their models, but as you mentioned, their compute is unparalleled. You know, by the way, a slide pushback. back. It's not just making money. It's making money with less money invested. And what makes AI very special is not just that you grow the revenue in partnership, but the cost of getting the revenue should go down radically. And this, again, is particularly important for smaller

institutions because in the past, the way you would do this is by investing billions and billions of dollars into complexity. And the smaller institutions could not weave their way through that complexity. But what we can do, to use your metaphor, is supply the pre-built tree of your knowledge XAI models and our software infrastructure, and that pre-built tree lowers the cost of revenue and allows you to increase the amount of revenue you can take. Because you know what I'm talking about. Okay, when we talk about, you know, let's go back to what they call greenwashing, right? Every company was climate friendly, right? We make bleach and now we're climate- we make, you know, chemicals and we're climate-friendly.

There was a lot, every company wanted you to say they were green, they were whatever. There's a lot of that going on in AI right now, Alex, you know that, right? You know that, right? Companies like, oh, we used to do this, but now we're AI and everything's great. How much of it is just BS? I was gonna say something else. You've gotta judge the,

you've gotta judge the products by what they bear. So again, is the revenue, cost of revenue less? Is the revenue going up? Are you able to deal with complexities you couldn't, like regulatory complexities? Are you able to impose unique business strategies onto your business, and those businesses are going to flourish. And yes, there's a lot of BS, but you can really, really see it in the numbers.

One of the things, by the way, I would say that's very contradictory is the people who are most interested in making money are going to gravitate to the things that actually work. And it's very similar to the battlefield. It's like there's not BS on the battlefield, and if you're going to partner with Thomas and his partners, you have to deliver real financial results or they're not interested. Because you're not doing magic. This is about success. And you better have metrics to weigh and measure everything you're doing

and to be hyper honest about the results. That's it. And look, I believe artificial intelligence is ultimately going to have an unbelievable impact on almost every aspect of life. But it's early, right? And it's not magic. It's going to take time to develop, develop use cases, find out what works.

And, you know, that's what we're diving into. A billionaire hedge fund manager, Paul Tudor Jones, who you probably know said on CBC this morning that AI was like going to kill half the human population or something like that. It's an existential threat to humanity. Sounds like you would disagree with that. Well, I'd have to have the context.

He's obviously a very smart guy. But I do know this. We have adversaries on this planet that are pouring everything into artificial intelligence, quantum computing, space and if the United States is not willing to compete and go all in, I think we're going to have very big problems. One, to just tie these, the biggest danger to your company or to our country is not being

able to know what to invest in and what's BS. The stuff you can invest in on the battlefield all over the world has already changed the course of history. And right next to it are a bunch of complete BS products where the unit economics are completely failed relations. Like what? Give me an exam. What's... I mean, look, I'm not... It's garbage. We're focused on making our business great, but every business is going to have to reach out. And one of the things that makes

Thomas so special here is you can't expect everyone to have the taste to pick out the thing that's BS or the thing that's not. Thomas went and found the world's leader in battlefield operational AI and hired him. So you don't have to trust your taste on AI. You can trust his. You don't have to trust your taste in what's going to end up price. You can go look at his. By the way, every government in the world is doing this.

They're looking at what works and what doesn't. He told me yesterday that Palantir was pretty good at this. Most people in the world who are making money by lowering their unit economics and growing their revenue and or are alive on the battlefield have a partnership that looks a little bit like the three of us. Very quick. Well, I want to ask about Aaron Rogers, but it's Steelers. I know he's going to walk off. Very quickly, your numbers look, I got to ask you this.

Stock's down today, but you've printed money. You've made a lot of people millionaires, a lot of our viewers and listeners, hopefully, very rich. What's the market getting wrong about today? Hey, if you don't like 71% growth in USCOM, 45% growth in USG, a rule of 83, 55% growth in 90% of our business, in 55% growth in the US, 49% growth in 90% of our business. You don't like that? Go look for something else. We're pretty happy.

You don't have to buy our shares. We're happy. We're going to partner with the world's best people, and we're going to dominate. And you can be along for the ride, or you don't have to be. Are we going to see Aaron Rogers in a Steelers uniform this year? I'm here to talk about AI, and that's a more complex issue than artificial intelligence. I don't know if that's an existential threat.

threat, but we'll get it. We won't get into it. Don't worry. Thank you. I've got a panel later with Peyton Manning. Maybe I'll ask him. He's a good man. Thomas Tall, Alex Karp, great conversation, big deal, TWG, TWG, Palantir, XAI, change in financial services and insurance and it's not BS. But Kelly, I am BS, and I'm going to send it back to you. Great stuff, Brian. Thank you so much. You really appreciate it. Brian Sullivan. I will hear from him again shortly. We have some news close to home. The group of media brands, including C. NBC that are being spun off from Comcast and NBC Universal later this year, now has a name. Versant, the name chosen as a blueprint for versatility, growth, and innovation.

This is the company's new logo. Cnbomboc, MSNBC, Gulf Channel, USA, Oxygen, and More will all be a part of Versant now, as well as digital assets, including Fandango, Golf Now, and Sports Engine. CNBC Media and Sports reporter Alex Sherman is here with me. Alex, you have more details on this new name. Very exciting. This is corporate history that we are living right now, Kelly. Yeah, versent, like converseant.

The name was chosen. I'm told to speak to the versatility of the brands in the company. You just listed them a lot of different brands covering different things, news, sports, digital assets, Rotten Tomatoes, and Fandango are part of this company. Also, I'm told that originally, through the process of this, choosing the name, I spoke with the CEO of the company Mark Lazarus, a thousand names, or even more than a thousand. names were thought of, were vetted. The legal process ticked that list down from 1,000 to 43. That's how difficult it is to actually find a name. It's not taken. That's not already taken.

There's, you know, trademark things that people have to go through, both nationally and globally. That list was then called down to about 12. There were presentations made on the 12, and eventually Versant was chosen. Also, the word Vercent itself is an actual word. something that I learned through the process of this. It means the slope of land. Mark Lazarus joke with me that perhaps he could see that as a sloping upward, like a line that was moving up for a stock symbol. So perhaps, you know, the pathway has been set for this company of assets

once it trades publicly, which will be later this year to be a riser. Is there any more clarity on when that date might be sometime in the next few months? Well, we don't know exactly still. I just know it's toward the end of the year, which has always been the plan. I can say that none of the market, you know, a volatility of the past month or so has moved Comcast off the date of spinning off the assets later this year.

Versant. Alex. Get used to it. Yes, indeed. Try it on for size. Sounds good. Alex Sherman. As we had to break,

check out oil higher today after closing at its lowest level in four years yesterday. Energy stocks down 9% in three months. After the break, let back out to Milken to hear from a real power player on this very issue. We'll be right back. All right, welcome back. Your next guest is a power player in the energy space. And it has to do with a deal between Calpine and Constellation Energy Group.

It's a company called Energy Capital Partners. The founder and executive chairman is a guy named Doug Kimmelman. Doug Kimmelman is sitting right here. And if you don't know this name, you will because this may be the greatest and most profitable energy deal. Once it's done, it's not closed yet of all time. Very happy to be joined by Doug of Energy Capital Partners. You got a pretty big act to follow up on on that. I saw that, but thank you, Brian.

But, you know, it's all connected because what Alex and Thomas Tull just talked about, a lot of tease there, is one reason that Calpine is selling itself or will be closed soon to Constellation Energy, which is, by the way, the number one stock in the S&P 500 right now of about 12%. Because the demand for power to do the kind of stuff they're just talking about, is driving that deal. You saw it years ago. What did you see? Well, believe or not, I've been decades in electricity. Demand has been flat. It's very boring for a long time, by the very boring. Now I'm an exciting guy. What can I say? But demand has been flat for decades because any growth in population GDP, all of that has been offset by energy efficiency.

But we've kind of changed all the light bulbs and made homes more efficient and all of that. But all of a sudden, demand has come in. It's not just AI, though. Think about electric vehicles. Think about onshoreing manufacturing. All of this LNG export, we need electricity, right, to liquefy and freeze that natural gas. So we've got multiple sources, but even bigger than all of this demand growth, we've got aged infrastructure, aged power generation. 40% of our electricity today comes from coal and nuclear. Coal, a lot of environmental challenges, at least many people view it that way, but nuclear's old. It was built in the 60s, 70s, 80s, So we're losing supply at the worst possible time that we've got all of this demand.

Where do we go wrong? Because somebody somewhere screwed up. Well. A lot of somebody's. A lot of somebody's slewed up because I like to say I'm in the business, not electricity. I'm in the business of reliability. We've got to keep the lights on. We've got to keep the factories going.

We've got to keep the cities working. Spain and Portugal just learned that the hard way. One of the biggest blackouts in European history. And it's not just a human issue. You've got people in the hospital, babies being born. It's scary. It's also a massive economic cost when your entire economy shuts down because somebody screwed up the power grid. Yeah, well, exactly right.

But we talk about the economy. We talk about economic growth. And we had Treasury Secretary Besson here yesterday. He talks about 3% GDP growth in the near term. Really critical to grow us out of this huge budget deficit and $37 trillion of debt. So what do we have right now? We've got, I call it the triangle of AI, power generation, and the natural gas production that is needed to fuel. fuel that power generation. We're talking about several trillion dollars of investment into the economy

now. Power, I said all of the nuclear and coal that's going to be shut down and this demand growth coming, you know, 300,000, 400,000 megawatts. That's a trillion dollars of investment. Well, lower oil prices that we can throw oil up guys, you know, mid-50s if it keeps falling, is that going to hinder natural gas growth in America? Because natural gas is largely a byproduct of oil drilling. To some degree, an associated natural gas comes off of that. But, you know, We are so fortunate that we have such a resource of natural gas and, you know, this can happen in this country, this economic growth. This AI boom needs to stay in the United States and it needs power to make it happen. I normally don't do this, but you're a friend.

So your wife was a teacher at inner city, L.A. for a long time. Yes. I'm a child of a, you know, long time ago, but downtown L.A. You guys are spending your own money to help build this awesome playground and park for the underprivileged use of L.A. I normally don't do this. is not part of our deal. I'm doing it because it's a great cause. I don't want to say thank you and your team for doing this and also for coming on the show. I'm honoring my late wife. We're building the largest academic and athletic community project in Los Angeles,

partnering with Tiger Woods and the United States Tennis Association. We're very proud. We're going to change lives of the kids that most needed here in L.A. Lulu's Place. That was your nickname. You didn't even know I was going to do that. There's no deal, but it's a great cause, great kids. Thank you, man. Appreciate it. Doug Kibbleman, Energy Capital. partners. Thank you very much. Kelly. I think that's very cool to know about, Brian. Thanks very much. Up next, forming a new bond with investors. Treasuries and corporate credit have been in focus, but our next guest has a new idea. Market Navigator tackles that next. Welcome back to Power Lunch. Just keeping a quick eye on

markets. We're down near session low's top of the hour, but we're coming back somewhat, does down 300. S&P and NASDAQ down about half a percent right now. And that, Domchew, brings us to Market Navigator. All right. So, Kelly, what we are looking at right now is equity markets, It's still in the state of flux. Bonds have come back in favor somewhat. Treasuries and corporates have always been a story. It may now be the time, though, to shine for another type of bond, and one might be a little bit less taxing on your portfolio and peace of mind. So joining us now is Patrick Schaefer of UBS. He's talking about a new power asset in the market, and we're talking about an old one. It's municipal bonds. He was just involved with two major orders from family offices for millions of dollars. And Patrick, let's talk about why the high net worth, ultra-high net worth clients that you service as a private banker and wealth manager are now looking at muni bonds.

It's an old asset class, but why are they in vogue again now? The short answer is muny bonds have sold off considerably relative to other bond asset classes. And we're seeing incredibly attractive tax equivalent yields right now in high-tax jurisdictions such as California and New York for high-income earners. what that means in dollars and sentences, we're seeing yield in the 4.4, 4.3% range and from a taxable equivalent basis, that's approaching 9%. And with some issuers, we're actually seeing yield in the neighborhood of 4.5, 4.7%, getting close to 10% on a taxic level. This is crazy, Kelly, because what we're talking about is a bond that in many cases has a triple tax advantage. It's exempt from state and local and federal tax, and it's yielding on an after

tax basis in that four to four and a half percent range, which means that it would have to be yielding on a taxable basis around eight to nine for some people. Yeah. And in fact, even above nine for people in the highest marginal tax brackets in California and in New York. And so this is attracting significant flow over the last two weeks. We've seen record flows and the year-to-date underperformance of municipal bonds has shrunk by about 20 percent over just the last five trading days. That's because these yields got so attractive relative to treasuries.

Because it's weird they're disproportionately hit lately every time there's like market disruption. But I mean, again, if you're looking to get in or you're looking to hang on for a while, a couple, sometimes you look at the stories, you go to more to make sure all the financial situations would be okay. Like in Boston, this is a great journal story today where they said, you know, their revenues are down so much from a vacant office building so they're having to turn and increase taxes on homeowners who are furious and they're pushing back. And so, you know, are there any areas of the Muni bond sort of world where you'd have a little bit more concern about their long-term ability to service the debt or no? Not right now. I don't think we necessarily see major credit concerns in the muny bond market. There are certainly some policy concerns that have focus around the tax-exempt status of municipal bonds remaining intact.

That's been a big story in municipal bonds this year and a big reason for the relative underperformance to treasuries. But for the most part, we think that's a relatively low probability event. that has stayed intact since 1913, despite multiple Congresses taking a look at it. There are areas of municipal bonds that may be more at risk, such as private activity bonds. That's obviously gotten a lot of focus because of some potential policy changes there from the administration. But in general, we do not have credit concerns and municipal bonds. So let's talk about those high tax areas, places like California, places like New York, New Jersey, Connecticut, Massachusetts and others. when you are looking at trying to buy bonds on behalf of your clients,

if your clients come and say, hey, we need to place X million dollars of an order and we want to be in a tax advantage situation, how do you identify what kinds of bonds? You mentioned private activity, but we talk about general obligation bonds, we talk about revenue-based bonds, we talk about certain jurisdictions, are they local school district level, stadium fixtures, anything like that? What exactly is attractive to you right now? longer dated bonds from large issuers that are a very high quality.

That's where our focus is right now. Like what? Major general obligation bonds, city revenue bonds, things of that nature, out in the 15 and 20 year part of the curve. Last week we were buying 10-year AA-minus California municipal bonds with yields in excess of 10-year treasuries. And so we got over 100% of the treasury curve last year on a 25-year bond that had a 10-year call. That is the exact sort of structure that we're looking to buy in this market. I don't know. Maybe you just go with Palantir, you know.

Or, yeah, maybe it's not as exciting as the Mag 7, but there's a lot to do. I think for a lot of people, especially when those ages approaching, you know, 529 type of events or, you know, retirement. I know you have to be a little bit more careful that, you know, that capital is there. Patrick, thanks. We appreciate it today. Oh, that's great. Thanks so much. as well. Patrick Schaefer, Domchew for Market Navigator. And up next, Invidia's Jensen Wong,

joining us live. The stock is still about 26% off its recent highs, but it just came off its first back-to-back positive weeks since February. So what's next for the chip giant? Power Lunch. I'll be right back. Welcome back to Power Lunch. Quick glance at the markets. The 10-year yield around 431 right now. A little bit of relief there after the bond auction top of last hour. Believe it will check in with Rick Santelli and get his thoughts on that now. What can you say? Well, there's a lot of moving parts to what the market's paying attention to today.

Let's start chronologically. 8.30 Eastern this morning, we had a record-setting trade deficit. Look at the chart going back to the beginning, 1992. Boy, this is what it looks like when you're pulling forward import activity due to uncertainty. It's just unreal.

Keep in mind, pre-COVID, the biggest trade deficit we had was about $64 billion. We're more than double that. Now, look at the intraday of tens. You see that big slide on the right side of that 24-hour chart? That was a 10-year auction. It was fantabulous.

However, there's something to pay attention to. Yesterday, the Fed bought $20 billion of the $58 billion, three-year notes. Today, they bought $14.8 billion of the $42 billion of 10-year notes. If you do the math, each case, it's about 35%. So we're going to see that Fed balance sheet, to grow again. And remember, every Thursday afternoon we get to see the size will pay particularly close attention this Thursday. And finally, the last chart is one week of Fed Fund Futures, the December contract. It has been moving down. When it moves down, it implies less easing. You notice how

it's moving sideways at this point? That's important, especially in front of tomorrow's Fed meeting. Kelly, back to you. All right, Rick, thank you very much, Rick Santelli. We're now a few years into the AI boom and some of the biggest thinkers are still grappling with how to get the best out of AI while avoiding its worst potential. Paul Tudor Jones on Squawk Box this morning talking about some of the dire warnings that are coming from the people most involved in developing AI. We just have to realize, to their credit, all these folks in AI are telling us we're creating something that's really dangerous, it's going to be really great too, but we're helpless. us to do anything about it. That's, to their credit, what they're telling us. And yet we're doing

nothing right now. And it's really disturbing. The AI advancement continuing a pace we just heard from Alex Carp of Palantir about his company's continuing efforts. And today's Service Now announcing a redesigned AI platform utilizing agents for customers. Service now also saying they expect to reach a billion dollars an annual contract value for the AI business by the end of next year. That would be a quadrupling from current levels. Let's get out to Las Vegas, where a judge, John Ford is at the Service Now Knowledge 2020 conference with some special guest, John. My experience in firsthand San Jose. Kelly, yeah, I'm in mid-conversation with Bill McDermott and Jensen Wong, Bill, CEO of ServiceNow.

Jensen, of course, CEO of NVIDIA. Thanks for having me back here at Knowledge 25. It seems bigger than it was last year than the year before. Let's jump right in, right? This is a room where it's happening right now with AI. You have this billion dollar now assist driven target by the end of next year. Tell me how much of that is ServiceNow working directly with customers' data? How much is Control Tower and your agents able to orchestrate even other competitors at?

Sure, sure. John, first and foremost, I want to thank Jensen and the great Nvidia company, because ServiceNow would not be the company that it is without Jensen and without Nvidia. And so it's an honor to be with him today. Thank you, Jensen. Thank you very much. And I would never miss a knowledge conference. Who doesn't love knowledge?

I love it. I love it. And so, John, we are the control tower for transformation in these enterprises. And what's happened is the 20th century software industrial complex is now consolidating onto the Service Now platform because it goes east to west across all the functions and north to south up and down the tech stack. And so we integrate with all the LLMs, the large language models. We obviously have our own built into our platform thanks to NVIDIA.

And so we welcome all data from all data sources in the enterprise so we can be that control tower. The big announcement with NVIDIA today was the reasoning model that we've announced, and we want to be the workflow data fabric that allows all that data to come through service now and of course everything in service. Service Now is powered by NVIDIA. Jensen, I'm glad that Bill mentioned this reasoning model in the partnership, because you've been saying for a while,

we're not just about the chips, you know, this is platforms, software is included. I want to talk to a lot about software. Tell me about the role of companies like ServiceNow, who you've been working with for longer than the kind of open AI driven hype around this, in building the sort of value and productivity necessary to realize the vision you've been talking about. First of all, AI has reinvented the entire computing stack.

And agentic AI, these agents, are the most complex of software that the world's ever built. Remember, this is software that can understand almost any content, any PDF file you give it. It can watch video. It can look at websites. Go into your database. Look at SQL. Well, these agents understand data of all kinds, raw data, unstructured data, structured data.

It can reason about solving problems. It can think. And so this software stack is insanely complicated. The reason why we do it is so that we can help the world build new software on top of this new computing platform called GPUs. And so that's the reason why we have to do it. Working with ServiceNow has been incredible. Six years ago, we shared a vision.

that ServiceNow can very much be the enterprise AI operating system. And on this operating system, we would have all kinds of incredibly smart AIs that help us do work. And it's accessing data from all these different sources, just as Bill was saying earlier. Today we announced this brand new reasoning AI model. It's called April Nemotron. It's fast. It's smart. And it's integrated deeply into the Service Now platform. So really exciting day.

So your communication about the AI vision has been consistent over years, but the facts on the ground keep changing. The goalposts keep moving. You just had to have a huge write-off over these issues of what chips exactly you can sell into China, for example. How much does it affect the ecosystem and Nvidia's influence over how AI pans out when the rules are changing? Well, you know, the world's dynamic today. You've got to stay agile. We took a huge ride off, about $5.5 billion.

China is a very large market. It's probably going to be a $50 billion AI market in a couple of two, three years. It would be a tremendous loss not to be able to address it as an American company. It's going to bring back revenues. It's going to bring back taxes. It's going to create lots of jobs here in the United States. And of course, of course, you know, we just have to stay agile. Whatever the policies are of the government, whatever is in the best interest of our country, we'll support and we'll stay agile and keep moving on.

The market for AI is just gigantic. Remember, Enterprise AI that we're going after is brand spanking new. The billion dollars that Bill is going after for next year, that's the first billion dollars of Enterprise AI. There's a trillion dollar opportunity out there to revolutionize the way companies are. are built and the products that companies make, not just the way they run, but the products that they make. Every part of it is gonna be revolutionized by agenetic AI.

Bill, you told me something two weeks ago, it's part of earnings, and I wanna revisit. Sure. Because in Morgan Brennan's conversation with Alex Karp yesterday around their earnings, it came back up again, CEO of Palantir. And it was around the monetization of AI and how right now, sure,

Sure, the model creators and whatnot have to pay a lot to get those things going. That's maybe why they're not profitable, but you're able to be really efficient and how you actually provide AI value to customers. What's the delta? How do you think strategically about how much more value through partnerships like this, through your internal work, you can get out of the hardware resources and how that translates into profitability? John, you've got to think of it this way. Between now and 2030, there will be a $22 trillion

business impact from AI in the global economy. Four trillion of OPEX will come out of these enterprises. So the prize is gigantic. We are also putting a different spin on AI than, let me sell you an AI for a silo somewhere in a company. We're talking about autonomous, agentic AI that goes across the entire corporation. So what's different about us is many companies want the predictability of the number of users that will be using their software so they can budget and plan for it. That's fine. But also with AI, we give all of the AI to the customer on our platform and thousands and

thousands of use cases for free. And at some point, when they've used up their tokens, yes, a consumption model kicks in, but that's only after they've either saved or made. hundreds of millions of dollars. So John, the prize here is bigger than anything we've ever chased before. And that's why I love Nvidia and Jensen's vision so much, because this is Mr. Exponential here. And Nvidia is the exponential company. And to think that we had that vision, Jensen had a great conversation with me. He said, Bill, we have to build a company that if everyone called in sick on the same day, the company would run just fine. And so there's a lot in there.

and that's what we're doing we're building the autonomous company and you know smart building autonomous vehicles even a smart lawnmower the best products will be AI driven they'll be smart products Johnson I want you to peel back for me another announcement that you guys had this week around parakeet so this software that you guys have built of course on top of your stack on top of your platform that I'm hearing is able to translate kind of audio voice to text, an hours worth in one second, with a 6% word error rate. So this is competitive.

Incredible. With the most efficient proprietary models out there that I think are in the 2 to 3% range, but this is open source. You guys are just putting this out there. I imagine it's going to be built into games. It's going to be built into enterprise software. It's going to be built into a lot of consumer apps.

What's the role of that kind of work? that you're doing in seeding the kind of innovation that you're talking about that's going to grow the AI use case and demand for in VINVITES technology. We have one of the world's leading industrial laboratories incredible research team. We do fundamental research so that we could see the future of algorithms, the future of technology so we could build the necessary computing platform for it. We also do these fundamental AI research so that we We could partner with companies like bills so that our researchers and his researchers could

build this revolutionary reasoning model. Now the future of our AI agents are going to be multimodal in the sense that they'll understand information of any form, including voice and vision. Of course all of your digital employees will be able to understand information of almost any form. You'll interact it with it in almost any way and practically everywhere. And so we want to have this AI research team explore the future and build the best possible AI models for that reason, both for computing reasons as well as so that we could contribute

it to the industry and use it to partner with great companies like ServiceNow. Now one of the things that you were leading up to just now, and Service Now is unique in this way in Enterprise AI today, they are a full-stack company. They build their own infrastructure, host their own AI supercomputers, deliver the service directly to the enterprise companies. They work with the cloud. They work on-prem. As a result, they could deliver excellent quality models,

really smart models, really efficiently, cost-efficiently, and that interact really fast. And you've got these models across different industries. So talk to me about who's accelerating now. We talk a lot about customer service and sales as areas that got going early in using AI and using agents. What's the next area? area, the next industry that you see really powering for?

Well, we started out, I think the first big one we did together years ago was telco. And it was all about the network and the customer service and all the predictive analytics that have to go in to keeping the customer happy so you retain them. Now we're into financial services, we're into manufacturing, we're into pharmaceutical, we're into retail. There literally isn't an industry that we're not touching with agentic AI right now. What's happening in logistics?

and everything that's happening with Tarras? Great question, John. That's a great question. If you think about the systems that these companies are tied down with today, they were 60-year-old systems that are running the companies we know today. I'm not even getting into the 1959 systems

that are running public sector. I'm just talking about the main companies that you know and recognize, they've been around six decades, and they were built around rigid org charts. So they serve a particular function within a company very well. The problem is when Steve Jobs brought the iPhone to the market,

everybody started with mobile business and they work in teams. And all the decisions are made across functions. So that's why having an east-west, north-south platform powered by NVIDIA is such a critical success factor. So what I want to put out there is today, when we announce this reasoning model, we showed people that last year, Jensen and I were talking about avatars, we were talking about voice and animation, and really demonstrating the live agent in an enterprise.

One year later, we're into reasoning models that is solving complex corporate problems across every function of an enterprise. So my way of thinking of it is, second mover advantage is a really bad idea. Everybody has to get started right now. Jensen, the policy forces very, various different developers to effectively build on Huawei.

Right? Building that ecosystem, what happens? The United States has to recognize that we're not the only country in that race, that we have competitors. And so, look, we're confident people, we're a confident country, we have confident companies. We have, we're not afraid of a race. We look forward to a race.

Just let us go race. And so I think that now is the time when the United States needs to realize, that we need to put the pedal to the metal. We have to fly that plane the way Maverick flies that plane. And we've just got to go for it. Waiting around, talking about it, trying to hold other people back, you know, it's not necessarily the best move. The best move is let Americans do American.

Let us go after it and win it. And so I think that the world is right now hungry, anxious to engage AI. They're hungry for AI. Let us get the American AI out in front of everybody right now. Speaking of competition, there's some other chip companies out there that would like to try to muscle their way into the data center where Nvidia has been so strong. One of them's reporting today. How fast is the pace development wise? I know you're on this yearly cadence of new platforms that you need to keep on in order to feel like you're keeping that distance, that Usain Bolt distance ahead of the other runners.

We've got to keep reinventing ourselves. You know, when Bill and I first met, Envideo was really a processor company, a GPU computing company, and then we became a data center company. Today, we're an AI infrastructure company. The projects we are talking about, people are talking to us about,

are $50 billion, $100 billion large. $100 billion. Now, put that in perspective, that's Boeing. Not the plane, the whole company. And so these projects are gigantic. is no longer a data center play. AI is now a factory play, industrial play. It's a new industry. And every country realize that this is now part of your social infrastructure. Every country needs to have it. Just as we had the information infrastructure called the internet, we now have an infrastructure on top of that called the intelligence infrastructure, AI.

Every country wants to engage it. So now is the time for us to go after it. Our company has really evolved today. now becoming an AI infrastructure company. On that industry, I'm going to talk in a little bit to the CEO of Aptiv, one of your customers, in the auto business, which is certainly in the eye of the storm here, with the tariff stuff that we've been talking about. What can service now do for them? What are the unique sort of challenges that you see

in an industry like that that has, and I'm not talking about Apted specifically, but the broader auto industry, the labor cost issues, the geographic issues, the interconnected supply, chain issues. Yeah. So think of it this way. Right now, if things were status quo and Jensen and I weren't here and we weren't innovating,

there's a $10 trillion legacy tax on the companies run in America, that's 7% of global GDP. So doing what you always did and getting what you always got is insufficient, changes in the air. So with regard to active as an example, as policy changes, geopolitics, geopolitical dislocations hit, you have to have a real-time company. The platform has to be able to change in real-time. So if you now need to manage around the tariffs and you need to think about tier two, tier three suppliers, signing on new ones, a car has 30,000 parts. If you don't manage the tariff scenario properly, you have to pass on a $10,000 extra price to the consumer.

They may or may not be able to handle it. So what do the companies have? have to do. They have to say, okay, I have to figure my margin, perhaps with less revenue, and still have a viable company. So they have to use the power of Nvidia and service now to manage their OPEX hard and have different scenarios that AI will give them on how to do that, headcount, cost, etc. But at the same time, they have to think about, what am I going to do to grow? I can't roll over and play dead. I have to grow. So we're reimagining businesses and business processes. So let me give you an example. You could sell a commodity piece of hardware and have a certain margin on it. Let's say you want to go direct a consumer to a new market.

You're not going to beat the entrenched competitor, but with AI, I can arbitrage the labor and the service model. AI can figure out how I can sell a service contract and give you four-hour response time that you'll pay 30% more for. Now I've got a new business model. So all of this is happening in real time on the ServiceNow platform. In the old world, you would have to go to business logic, testing, deployment. A year later, you might be able to make the change. We can make the change on the fly with Nvidia and ServiceNow. Jensen, there are some companies that are rightfully concerned about OPEX,

and then there are some companies like META that are upping their CAPEX range, you know, by the billions, frankly. We know a lot of that money is going to be spent on your technology. What is, and you told me around earnings time, you're hearing about these build out plans years in advance, but sometimes it can be lumpy going forward. What is the role that you see of the largest customers, these hyperscalers, in driving the pace of innovation

and frankly the competition to have the power that these early movers need to really build out AI? The early hyperscalers realized a very profound insight. site, that they are now intelligence factories. They're intelligence generators. It's not a data center that stores information. It's a factory that produces intelligence. And these intelligence tokens could be reformulated into music, images, words, avatars,

recommendations of music, movies, or, you know, supply chain optimization techniques. And so it could be reformulated into intelligence. They realize that they built factories now. And these factories are gigantic in scale. The old data center industry is still going to get reformulated into AI. It's a trillion dollar industry. It's going to be reinvented. But on top of that, it's going to be a multi-trillion dollar AI factory industry.

It's going to require energy, but it's going to generate growth. This is the new insight. Some of the old thinking is, you know, how much of the old data center, CAPEX, you know, how much of the old data center, cap, is going to be used for AI. That's interesting, but the big idea really is that there's a new industry in town that's called AI factories,

and they're producing intelligence at scale. Got to leave it there for now. Bill, thank you. Jensen, thank you. Bill, I know you're coming back. You're going to join me in overtime. I'll be back. But again, I want to thank you, Jensen.

It's an honor to be with you as always, my friend. I just want to say, I want my service now. You got it, Jensen. We're going to give it to you. Thank you, Jensen. Kelly, back to you. The best. John Ford, Jensen Wong, Bill McDermott.

Our thanks, and that's it for Power Lunch today. Closing bell starts now.