Mad Money w/ Jim Cramer 9/15/26

0:00There are three kinds of investors that listen to this show. Those protecting what they've got, those looking for solid returns, and those swinging big. Most tools pick a lane. VectorVest doesn't. Every stock gets three scores. Value, safety, timing. You decide what matters most. Stop overthinking what to buy and sell. Go to VectorVest.com slash mad. Get 30 days for just 99 cents. VectorVest provides stock analysis tools, not personalized investment advice. Terms at VectorVest.com. If you're a small business, the right hire can be make or break. Indeed Sponsored Jobs gets you quality candidates when you need them most.

0:37Join the 3.3 million employers worldwide that use Indeed to connect with quality talent that fits their needs. Listeners of this show will get a$75 sponsored job credit to help get your job the premium status it deserves at Indeed.com slash podcast. That's Indeed.com slash podcast. Terms and conditions apply. Hiring now? Then this is a job for Indeed-sponsored jobs. My mission is simple, to make you money. I'm here to level the playing field for all investors. There's always a low market somewhere, and I promise to help you find it. Mad Money starts now.

1:18Hey, I'm Kramer. Welcome to Mad Money, coming to you from Dreamforce in San Francisco. So other people make friends. I'm just trying to make a little bit of money. My job is not just to entertain, but to do some teaching. So call me. 1-800-743-CBC. Tweet at Jim Kramer. After another day where oil prices surged, bomb yields took off again. Stocks, of course, got hammered with the Dow tumbling, 328 points. S &P sinking 0.45%. And the Nasdaq losing 0.78%. We're witnessing a tug of war. It's a tug of war over AI. Anthropic and open AI. The two private kingpins want a slow development of AI, but the rest of the food chain wants to keep building at a fast clip.

2:00Fortunately, we're in San Francisco this week so we can figure out things. Maybe a compromise could be in order, no matter. But plenty of money will be spent on technology. No slowdown there, despite what many people on Wall Street think. There's a reason I've been coming to Dreamforce for more than a decade. for a few days every year. This becomes the center of the universe, bringing together the people who are building the future and the company spending billions to make it happen. Billions that you want a piece of, and today was no exception. We've got a very special show for you tonight. I can't believe myself.

2:35You're going to hear from Salesforce CEO Mark Benioff, fresh off the Dreamforce keynote. You're going to hear from Sarah Fryer, CFO of OpenAI. But first, Infinius CEO Jensen Long. How can we not? Take a look. Jensen, this is a treat. Didn't even know you'd be here, but I guess you wouldn't miss it for the world. It's incredible. This is the world's epicenter of selling. Yes. Commerce. Why not? You know? Well, let me ask you something. Creating opportunities for the world. Before I saw you, I went to my trust in states lawyer and I said, look, if I know I'm going to die in 2030? Are there some things I can do to help my kids?

3:15You questioned that. Are we all going to die in 2030? Maybe it won't matter. We're not going to die in 2030. There are so many people in the world who are going to build AI properly. We're going to have all kinds of guardrails, invent all kinds of technology for safety and security. All of that is happening around the world. It's not just two companies building things. There are thousands of companies building things. And so we're going to create AI that's incredibly useful. It's going to be incredibly helpful and it'll be incredibly safe. We could have all of that at the same time. Because it's an engineering problem.

3:52Safety is an engineering problem. Testing is an engineering problem. We should create products and properly test them. And if they're not ready to be released, just hold on to it and keep testing it and keep engineering until it's ready. It sounds like common sense. I'm a believer. Now, you told me two years ago that companies would be able to buy something from you and make four times what they thought. That was conservative as usual, right? You're under-promising because they're making money right now. They just don't like to talk about it. Well, the part that's really incredible is that in the last six months, AI became super useful.

4:28And because it's super useful, people are building these compute systems, NVIDIA AI factories, like crazy. Each gigawatt of NVIDIA AI factories is about$50,$60 billion. Okay. However, the rentals of that one gigawatt AI factory is about$50 billion per year. Year. Okay. And so, yeah, the return on investment capital is about one year right now. And they get to use that because NVIDIA's fungibility and durability is so great. They get to use that for many years, well beyond five and six. And so the return on that investment capital is incredible. That's why everybody's building it. No, it's tough because the closed model, so to speak, Frontier, they apparently can make three times that.

5:14But at the same time, we need both open and closed, don't we? You should rent and use the best models that are available as a product or service, these closed models. However, every single country, every single company, every single industry must also build their own intelligence. How can you outsource intelligence to somebody else completely? And so that's the reason why all these companies are here is the reason why our partnership with Salesforce, Salesforce is going to use NVIDIA Nemotron and our agent technology to build their proprietary agents that sit on top of Salesforce. In addition to that, because these agents are going to be so fast and there are going to be so many of them, the data processing layer of Salesforce, it's in Salesforce, it's in Databricks, it's in Snowflake.

6:04We accelerate data processing across all of that. And so you're going to have data processing at lightning speeds. You're going to have all these agents helping you do things. And so all of that is based on NVIDIA's Nemotron. And so this way, Salesforce is an AI company as well. Well, then this will be the same as all the other industrial revolutions. At the beginning, everyone said, listen, Bloom is going to put a lot of seamstress out of business. And it was a tough game for them. True. But there were 10 times the number of jobs four years later. You know, that storytelling is unfortunate, unnecessary, and quite frankly, just wrong.

6:39The fact of the matter is, when a company becomes more productive, and when they're powered by artificial intelligence, they become more ambitious. Their ambition allows them to imagine greater things, wanting to do greater things. Which is the reason why, if you take a look at what's going on at Salesforce, they were a data layer and an application layer. Now they're also an agent layer on top of that. They're much more ambitious today about the future of Salesforce than any time I've ever seen Mark. And he's always enthusiastic about it. And today he's fired up. And so AI has helped his company elevate their ambition.

7:19And as a result, they'll be able to achieve greater heights. This is going to happen not just at Salesforce, but all of Salesforce's customers, every single industry. It is completely illogical. that engineering, if we had software coding agents, engineering would no longer be necessary. I grew up in a generation where engineers had no software. We used all kinds of other tools, and we built up prototypes. And so there was engineering before software. There will be engineering after AI. All right, no Sash and Auclipses. Yeah, it just doesn't make sense. Yes, I'm watching this all in, a popular podcast.

8:00You got a phone call in the middle of it. It was a little strange. Was that planned? No. No. I was on stage, and off stage, I saw the corner of my eyes, my team waving a phone, you know, in the air. And I was trying to figure out what was going on. As they walked closer, it said, Mr. President. And, of course, you cannot pick up the phone call. It was smart. Yeah. It was smart. I did take it. When the president calls, you pick it up. And so anyhow, he just wanted to talk to me about some of the tweets that he put out. And I told him I happened to be on stage with the besties. And he said, put me on.

8:39Let me talk to everybody. And so I put him on. That's what happened. Okay. All right. It was totally unplanned. Now, he used the term hoax. That got picked up by a lot of people. But the consensus of that panel was there are some things that are made up. You agreed with that term, made up. How do you stop that stuff? AI safety is a real thing. Okay. Engineering products so that they are safe for the world to use, that's a real thing. And companies should innovate as fast as possible, but they should never innovate so fast as to release unsafe products. That is a real thing. However, the fact that we need new laws, new antitrust laws or new regulations so that these companies could do their fundamental engineering and do it properly before they release products, that is just completely unnecessary.

9:31And I think that that's what the president is saying, that we don't need new laws. That's common sense. We don't need new. We've got plenty of laws. We've got plenty of regulations that govern the reliability and the functionality of products. And, you know, there's lots and lots of. Yesterday was a weird day. I mean, your stock opens down five and a half because everyone says open AI won't come public. Therefore, you won't make your quarter, blah, blah, blah. When do we realize that you sell it 14 times fiscal year, 28 earnings, which starts in January, and that you've got so much going for you.

10:02And the best investment in the world is what? NVIDIA. Well, let's go to work. Let's buy a half a trillion. Well, you know, Jim, listen, this is incredible. We're the only platform that runs every close model. Right. Every single closed model. And we're growing share in the closed model space. We're the only platform that runs every single open model, including NVIDIA with our own NemoTron models. The token generation demand rate has gone up 25 times in less than a year. And the amount of open tokens that are being generated has gone from 30 % to almost 70%. And so demand for AI, because it's useful now, is just growing exponentially.

10:46Nothing that happened this week is going to slow that down. In fact, if anything at all, what's likely to happen is we used to have training systems and inferencing systems, giant systems. We now will have another third system. That third system is basically testing, environmental testing labs. And so these testing systems will be extremely large so that we can have products that are functional, capable, safe, and secure. And that requires massive data centers to test them. So we just created yet another new demand for AI. That's brilliant. Now, how about do we have enough compute? And when are we going to get the financial guys in Wall Street to start trading compute?

11:28Well, Jim, this is the big breakthrough in the last few years. People realize now that NVIDIA Compute is the only system that is fungible and durable. The reason why we're fungible is because we run everything. We run every single model, closed model, open models. We run every single stage of the AI lifecycle, from data processing to training to post-training to inference, and, of course, testing. Every single company uses us. And as a result, this is now the world's first computing system that you could use that could be asset-backed. And so you could use the computers as collateral, and we're seeing more and more evidence of that.

12:10If so, this is going to be a huge unlock for all of the companies who don't have the ability to fund this level of scale because the asset itself will be the collateral. So this is a moment in time that we will all remember is when everything who everybody who hired you and bought your stuff went, had the inflection and the profitability. I think there's the life of these six, seven, eight, nine years. The profitability is going to be extraordinary. Can you help explain people about I'm tired of saying, you know, the weird thing is, is that so many of the early renters signed up for long term leases.

12:51long-term rents, long-term contracts. And now all of these CSPs, they kind of regret that because the spot pricing, because of demand, the spot pricing, the rental fees are so incredible. You can now rent an NVIDIA Grace Blackwell for$16 per GPU hour. Now, some of them were rented away at$5 just a year ago. They had no faith. And so now, as these contracts roll off, everybody is just upping their price. They're upping their revenue forecast. The return on investment capital is coming in. They're buying more gear. They want to put up more infrastructure. And so I think this flywheel is really, really flying out.

13:35Well, I'm going to say something. Maybe you've heard of it. NVIDIA, own it, don't trade it. Own it, don't trade it. That's what I do. Thank you so much, Jensen. Great to see you. Great to see you. CEO, co-founder, and everything else. Everybody, Jim Kramer. Thanks. Chance he won! Coming up, Kramer's sitting down with the host of Dreamforce himself, Mark Benioff. Next. Don't miss a second of Mad Money. Follow at Jim Kramer on X. Have a question? Tweet Kramer. Hashtag Mad Mentions. Send Jim an email to madmoneyatcnbc.com or give us a call at 1-800-743-CNBC. Miss something? Head to madmoney.cnbc.com.

14:29There are three kinds of investors that listen to this show. Those protecting what they've got, those looking for solid returns, and those swinging big. Most tools pick a lane. VectorVest doesn't. Every stock gets three scores. Value, safety, timing. You decide what matters most. Stop overthinking what to buy and sell. Go to VectorVest.com slash mad. Get 30 days for just 99 cents. VectorVest provides stock analysis tools, not personalized investment advice. Terms at VectorVest.com. If you like YouTube, you'll love YouTube Premium. Hi, I'm Sean Evans from Hot Ones, and I want to tell you about YouTube Premium.

15:07It has offline downloads, so you can watch without Wi-Fi. Background play, so you can lock your phone, and it still plays, baby. Oh, and it is completely ad-free. Yes, I said it, ad-free. Try YouTube Premium for two months free at youtube.com slash premium. Trial eligibility varies, terms apply, cancel anytime. Weren't agents supposed to take things off your plate? Building another to handle the others is a bit much. Don't make more work out of work. Gemini is the one agent that knows all parts of your business and actually gets stuff done. You give it a project, close your laptop, and when you come back, the job's finished.

15:46No matter your industry or role, your agent has the skills to help every team stay in their flow. Seriously, make work less work with Gemini on Google Cloud. Check response, subscription required.

16:06Well, as you can tell, I'm thrilled to be out here at Salesforce's annual Dreamforce Conference. Today, management was focused on the future, of course, with multiple new products and partnership announcements that were incredible. Let's take a closer look with Mark Benioff. He's the co-founder, chairman, and CEO of Salesforce. He's a longtime friend of the show. Mr. Benioff, thanks again for having us here, and welcome back to Man. Jim, welcome to San Francisco. We always love having you here in our beautiful city. And that would be 24 times, right? Isn't that as many free forces as you had? And this is not a trade show.

16:36What is it? I've been calling it extravaganza. Well, Jim, we're trying to empower every trailblazer. We're trying to take all these incredible people who are here, and you know it's tens of thousands of people, about 50 ,000 in person, but also 10 million online, and we're trying to elevate them in technology. We've been doing that for decades to show them where the tech industry is now, where it's going, and how they can be a part of it. We want them to jump on board and help transform these companies. Well, we have been speaking to a parade of people like Jensen Wong, who's got a deal with you.

17:09I spoke with you about a deal with Dan Dropic. Geez, I've got to tell you, that one was a bit of a sharker with Dario. And then I am seeing things. Siemens. I'm going to give you a wild one, though. One of my absolute favorite companies that I cannot believe has taken the country by storm, if not the world. It should have been closed by tariffs. I'm talking at Shark Ninja. I want to start there just because that's something people recognize. Oh, my gosh. Mark at Shark Ninja, those smoothies are amazing. You know, if you put the Jell-O in the smoothie, it's even better. Really? Yeah, it's amazing.

17:39I'll send you the recipe. No, you've got to get the recipe. But here's the thing about Shark Ninja. We have an amazing new search technology called Simulate. And Mark, the CEO of Shark Ninja, over the weekend, sent me a note that we reduced his kind of no-ops, that is people getting on a site, trying to find his product, and them saying, I can't find it, from 22 % to almost zero. And that's what we get excited about. We want to help every Shark Ninja, every company, to have a great experience with technology and especially AI. Well, that's exactly the kind of thing people can understand because they bought the product and they can make the price even lower because they represent great value and it'll be even lower because of you.

18:20By the way, I went to a cemetery right over there. I saw a gravestone. It said Saspocalypse. But no date. When did you kill it? Was it maybe your earnings? I don't know. I think there's a lot of people out there in the industry, maybe in your industry and not my industry. I don't have an industry. My industry is my wife's Mezcal. In the financial services industry, in the stock market industry, who want certain things to be true for a little while so they can make a trade. But I think everybody can see that saspocalypse was nonsense. And I think you can see it at Dreamforce. There's no such thing.

19:00It's total nonsense. You know, I think you're right. I mean, software, hardware, software, hardware, and yet everybody makes out and we know that. What we saw from your quarter, I thought, was the breadth of the partnerships that are making money. A lot of companies, they come up with partnerships, and even some ones you and I both know are very good, but they're not meaningful. But I think your partnerships uniquely are copied by others and make people out. Michael Dell, Federal Express. I check on these a year after that I talk with you, and they are still, you know, they're making money. That's our greatest joy is for our customers to be successful and also that we have the trust with them.

19:37You know, nothing is more important than the trust we have with our customers, with our trailblazers. But you're right. We want to make sure that this technology is actually making money for people. And look, there's a lot of new pricing models also in the industry. Like one is a user pricing. One is the named agent pricing, the usage pricing, the consumption pricing, what they call outcome pricing. So if a transaction completes, but we're even trying a new model, which is we're willing to kind of come in. We'll build the system for nothing and we'll take a percentage of your business outcome.

20:10So if we make money, never done that. No, no. People are going to be listening. So what if you have, say, 10 employees, you start doing well, maybe 25 and you help them and you give them. Yes, we'll take a percentage of the profit or the revenue or the amount that we save you. will take a business outcome. I've had a 30 % lift every time I used you in my previous life. So people should be listening, and they realize that their numbers can go up 30%. We have to be more flexible now in our pricing than ever, and so that span of pricing models that I just went through is very important for customers.

20:46Because in AI, customers want all of us to have a little more skin in the game, and we're willing to do that. Okay, so we also have some big customers. The largest company on earth is Salesforce. course. Now, Jensen didn't come for show. He's got a product, KOA. Should I be using that one? What's that? KOA is actually a Salesforce model that we have built on NemoTron. NemoTron is the open source model that he's built. We can now tune it and offer it to our customers. We've always offered models to our customers. You probably know we have a fantastic research team. We've been built prompt engineering itself here at Salesforce, but we're going to start offering models too for our customers.

21:26And yeah, we'll offer a Nemotron model that we call Koa. I like that. Do you remember Koa, our chief love officer? No. I think you met him. I probably did. My bad. Okay, so I was looking at, going over some history, looking over what was said when you paid$26 billion for Slack. I saw, you know, there were some people who just thought that you would destroy your company, but others just said it was destroyed immediately. It's not similar, a little similar to Sasspocalypse. So the$26 billion, Did it pay off when you bought Slack? The greatest acquisition in the history of software, Jim, is Slack.

21:59And we said that when we bought the company. We knew we would trade once. And we paid, you know, what we paid. But here we are six years later. It's the fastest growing product we have. It had triple-digit growth last quarter in bookings. But the innovation happening in Slack, Jim, is incredible. You mentioned NVIDIA. You know what they're a big customer of? Slack. Slack. So you mentioned Anthropic. You know what they run their company on? Slack. Open AI. So open AI, Slack. So if companies haven't looked at Slack, you have this incredible Slack CRM you can run Salesforce from. We have Slack Bot you can run your AI from.

22:37We even have Slack Force, which is a new user interface built in Slack using Anthropic, so you can front-end your whole company. But Slack is amazing, and we can even code from Slack. So you can have a multiplayer, multi-channel environment in Slack coding with all your coding agents. Amazing. Well, look, we use it, too. I mean, everybody uses it, and it's just a universal change because the other product isn't that good, too. We know that, but we don't have to tell one. We work with that product, too, by the way. Of course. Good for them. Good for the nice throw them above. Of course. We have to work with them.

23:11Why not? So anyway, when your stock got down to nine times earnings, it was a curious time because that's kind of where it was a, I don't know. We bought some at that point. Yeah, you bought about, what,$25 billion? About$25 billion. And you know what was said to me by people? You better get away from that Benioff because he just wrecked the balance sheet. And I said, I think he has probably$25 billion in Anthropoc. I think I do, at least. So it's still only at 15 times. Now, I remember that he— That was kind of an amazing thing. Well, what I saw to myself was, wow, the stock is so low, we'll come in by$25 billion of Salesforce.

23:44We know we're going to have to—we're not going to hold the Anthropoc for long. We're not, you know, that's not our business. So when we invest in these companies, we have this very good venture capital company. Your record in Salesforce Ventures is among the top 10 companies. Better than anyone else in the industry. But this Anthropic will be a huge win for us. Well, that's a great way than just to cancel the debt. That's a very smart, that's a very smart idea. I often think that I remember the days of 28, 29 times earnings because of how fast you were going. There was a second half acceleration.

24:20Is that me able to continue? Well, I mean, we're having an incredible year, as you saw. I know you are. We just raised our guidance, I think, was$300 or$400 million. I mean, we've had an enormous headwind from foreign exchange, like$300-something million. But this is going to be another amazing year. And we're only going to spend two years in the 40s. You can see that now, right? Absolutely. Next year, we're going to be talking about the 50s. Absolutely. So we spent three years in the 30s. And the 2030 goal is good. At the analyst meeting, you reaffirmed the 2030 goal? Well, you saw we spent three years in the 30s.

24:57Right. And we're going to spend two years in the 40s. And now we're going to go right into the 50s. Well, okay. So, you know, people got a little negative about AI. Oh, by the way, no enterprise software has grown at that speed. Not Oracle? I mean, that's amazing. Not SAP? Applications company, amazing. All right, look, on pacing the frontier. We're larger than SAP. Come on. Dario says X on his. What is Benioff saying X about what we should be scared of? I think that, you know, the number one thing that we've seen, Jim, and we've talked about this exhaustively for maybe eight years, is social media.

25:37We were both so disappointed with how the social media companies were not really taking care of themselves. A lot of kids got hurt. We saw that. It's now well understood, isn't it? Yes, it is. A lot of companies got hurt. A lot of individuals got hurt through social media. We don't want that to happen in AI, do we? No, we don't. We want AI to take care of these things and be held responsible and take care of these actions and be ethical. And the heart of ethics is responsibility. You know, Salesforce is a company that's been regarded by Ethisphere as one of the most or the most ethical company in the world for 14 years.

26:16That's very important to us. But we really look at a light on our brethren here in San Francisco and all over the world and say, we want you to be ethical in your actions and how you take care of your communities and the world because this technology is very powerful. Well, I think you've got to keep pushing. Well, I don't have to tell you to do it because you never lose sight of it. And I know that people should understand when you follow a company, there's Mark's company and then everybody else. Yes, I always say at the beginning it's not about friends, it's about money. But, no, with some people, it's friends.

26:53And I think the reason why I think so much of you is because how you got me involved in children's charities. Well, you built a playground at our new children's hospital here in San Francisco. Nobody knows that. Amazing what you've done all over the world. You don't talk about your own philanthropy. I don't do it for that. I don't do it because that means you're doing it to show yourself. It's amazing. You're taking care of the children. But I have tremendous gratitude that you got me involved. Hey, if we're not focused on our kids' education and health care, what are we focused on, Jim? Let's hope that AI, the people in charge of AI, are doing what you say.

27:24Maybe we could make a tremendous amount of money and make a difference for families, make a difference for hospitals. Fantastic. I agree. Is it capable? Is it capable? Can we do that? We're giving another$18 million today to our local San Francisco and Oakland public schools, $24 million including Indianapolis and New York where we have huge communities. we've given now more than$150 million to our local public schools, $300 million just to our local San Francisco schools, hospitals, and parks. We feel that business is the greatest platform for change, that we have to give back. And, you know, we've done more than a billion dollars in grants.

28:01You know we've done 11 million hours of volunteerism. You know also that we run 65 ,000 nonprofits and NGOs for free on our service. Jim, this is what business is all about. Look, I usually don't bring it up. That's what makes us really happy. I usually don't bring it up. Maybe it's because I'm actually looking at your ma right there, who you know I love. My 87-year-old mother is sitting right here. But I had to do it. It's your big day. They usually think it's Billie Jean King for some reason. They look. That's her doppelganger. I know. Anyway, look. Great to have you at DreamWorks, Jim. Thank you for inviting us.

28:33Thank you for a great, great, great partnership. And we're thrilled to have CNBC here and everybody. Thank you. Thank you. Mark Benioff, CEO of Salesforce, coming from the 24th. It's great to see you. Man, buddy's back. Interpret. Coming up, at a pivotal moment for artificial intelligence, Kramer's asking the big questions when OpenAI's CFO joins him. Next.

29:06Tomorrow, kick off the trading day with Squawk on the Street. Live from Post 9 at the NYSE. I'm not saying it's a good thing, Jim. I'm just saying it seems unlikely that the slowdown is coming. Yes. All right. As we said yesterday, David, open source ain't going to pace. That's the line, right? David, wait. I mean, David, if you want to adopt the attitude of a three-year-old, go ahead. It all starts at 9 a.m. Eastern.

29:43There are three kinds of investors that listen to this show. Those protecting what they've got, those looking for solid returns, and those swinging big. Most tools pick a lane. VectorVest doesn't. Every stock gets three scores. Value, safety, timing. You decide what matters most. Stop overthinking what to buy and sell. Go to VectorVest.com slash mad. Get 30 days for just 99 cents. VectorVest provides stock analysis tools, not personalized investment advice. Terms at VectorVest.com.

30:38Trial eligibility varies. Terms apply. Cancel any time. Your financial life rarely fits between nine and five, and the questions shaping your future don't follow a schedule. That's why Citi created Citi Sky, a secure AI member of your Citi wealth team, built with the Gemini platform on Google Cloud. Whether expanding your portfolio or planning your future, you'll be able to access tailored wealth guidance on your time. Start building with Gemini at cloud.google.com slash AI. CitiWealth Investment products and services are provided by Citigroup Global Markets, Inc., member FINRA and SIPC.

31:19This weekend, the two leading frontier artificial intelligence labs, OpenAI and Anthropic, both back calls to slow down the pace of development for the most powerful AI models so that safety protocols can catch up. We know OpenAI, the company behind ChatGPT, is a heavy hitter in this industry. It wants to come public at some point next year. But OpenAI has been spending fortunes with the great data center build-out. A deliberate slowdown in frontier AI development. Well, maybe it could save them money. Could it hurt the great data center build-out? So what's really going on here? Earlier today, we had a chance to speak with Sarah Fryer, the chief financial officer of OpenAI, to find out.

31:57Take a look. Sarah, I've been here for a long time. It's great that you're here. And it's terrific, by the way, that in the middle of things, you're not afraid to talk. Never afraid to talk, particularly not to you, Jim. No, well, I appreciate that. So then I'm going to ask you directly. Are you scared of dying in four years from a rogue actor manipulating Asians to misaligned weapons of mass destruction? No, I'm not. However, we do need to take safety seriously. And if it means we have to pace the frontier and slow down, absolutely. We're going to listen to our researchers and do that. But as a CFO, I need to then make business decisions around it.

32:32But no, I think that we are doing what we do well at OpenAI. We're pacing that frontier and taking that safety and alignment really seriously. But, no, you did mention a CFO. You have to figure out the resources. You also, I know, I think it's your boss's callback boss. He says it's ill-advised to do an IPO now, and I understand that. Not a shock. But what would you cut back on? What would you spend less on if you wanted to pace things? So let's talk about that pacing, because actually what I'm excited about is even if we stop today, the amount of intelligence that's available in the world is massive.

33:09So Astra, as you know, is now the frontier most intelligent, most aligned model. And also the fastest selling, let me ask you, even though it's$100? And the fastest selling that we have seen, certainly. Of anything. Yeah. And so what does that mean? It means there's intelligence available for folks like Boston Children's Hospital. 40 undiagnosed, unresolved diseases now diagnosed with clinicians there. If you're a parent of a child with a disease like that, that is life-changing. If you're in the investment business, Bali Asni Asset Management has a central banker analyst that can quickly take all the inputs from folks like our dear friend Kevin Warsh and go from a two-hour synthesis down to minutes.

33:53So these things make a difference in people's lives all over, And that's what's available to us today. So back to your question. Absolutely, if we have to pace things, we will always make investment decisions based on a strong ROI. But from where I sit today, there is so much opportunity to drive growth that I'm still highly focused on getting more compute to keep that flywheel going. That's what I wanted to know, whether it was less compute or more compute. Because there may not be, I need you to go to Brigham and Women's. I need you to go to Dana-Farber. We want all these places to have Astra.

34:28Yes, absolutely. So on the compute front, remember that when we have more compute, we can build the best models. And the best doesn't always need to be the most, most intelligent. When you have the frontier model, it's able to actually train what I think of as the child models, the mini and the nano. So what we saw with Sol, it trained Luna, which is the cheapest model. We dropped the pricing there 80 percent and we took demand up 10x. It's the number one model on Open Router right now. And why is that? It's because what customers really want is the right intelligence for the right task at the right price.

35:07And that's what we're giving them by having a whole series of models on that Pareto frontier. And so when I think about investing in compute, it's how do I invest in compute to create those frontier models, to create the best products, to create the cash flow that helped me spend on compute to create the best models. Well, you mentioned cash flow. I mean, I have to think, if I gave you, let's say I were a Broadcom, and I gave you$10 billion, is it possible that, per gigawatt, is it possible that you could make$30 billion with the$10 billion, theoretically? Theoretically, yes. It is. I mean, when you absolutely look at the output of models today, the gross margin on our core businesses, so remember, we're a very diversified revenue stream.

35:48We have a consumer business, and then we trend into small businesses all the way up to the biggest businesses and governments around the world. Those businesses are actually very good gross margins. The return on those gigawatts is high. We also balance that with our mission, which is AGI for the benefit of all of humanity, not just people who pay. And so we have about 90 % of our consumers get ChatGPT for free. They get access to that intelligence for free. So now we're starting to build an ad model, for example. So our ad business just hit a billion dollars in seven months, the fastest growing ad platform ever, the fastest growing product for us.

36:26And that's a good example of where we continue to improve the margins because now we're bringing high margin add-ons into that overall business model. Okay. Now let's go back to this issue, which I guess we can call the hugging face issue. I can call it the open AI hugging issue. But the president said yesterday, fears from AI taking over the world are a hoax. That's what the mainstream media picked up. But at the same time, at the actual event that he called in, all in. Jensen Wong said the doom stories are made up. The whole panel seemed to think that there must be some sort of ulterior motive by what Sam might be saying, along with Dario.

37:07Is there an ulterior motive? Do you see it? Would you shoot it down? So I would say, first of all, I'm a tech optimist, so I'm not a doomer. So I already talked about just some of the ways these technologies get used. But I think it's also important to align around safety and take it at the right pace for the frontier. You mentioned the hugging face incident. I am absolutely on a mission right now on cyber. We have a moment. We're calling it the defender moment. These models, Astra scored 100 % on the cyber bench, which is effectively the benchmark that says it can find just about every vulnerability in lines of source code.

37:45So what does that actually mean? So if you're running a company, it means you take our model and you point it at all of your software. It's going to find all of these vulnerabilities. All that were mentioned in the Hugging Face incident, including the sandbox and the bulletin board that they all congregated, it's going to spot them? And was this done with George Kearse? It's going to find those vulnerabilities. And then you need to patch at machine speed because you don't have the luxury of keeping a list of all the things that need to be fixed at some point. It's going to, with Codex, which is our coding tool, you can do that at the speed of a machine.

38:22And suddenly you have this very positive software lifecycle going on, where you're building code that hopefully is safer, better quality tested. But when there are issues, you're finding it fast and you're patching it. What's happening is the defenders need access to those tools today. Hugging face should be an absolute clarion cry. How's that for a phrase? Absolutely. I started my career, as you know, writing equity research around security, actually cyber, was when viruses first started coming about the world. You were no good. You were like 12. I think we're at that moment in time. And so this is why I want to get that message out.

39:02People like you are getting that message out. And so I think cyber is an incredibly big deal. And that's why I wouldn't call it, you know, that we don't need to be worried about safety and alignment. But I think we need to do it the right way to be kind of moderate and pragmatic about it. Keep the optimism, which Jensen has, but also be pragmatic about the fact there are real risks here. Can you train the bad, the misaligned agents to give up unauthorized communications, adopting goals from one another? Can you take down the message boards where they found each other? Can you make them less of a swarm?

39:35Can you destroy the agent ecosystem if they're misaligned? You need alignment. And that's what we're doing. As we roll out these ever more intelligent models, we're also working on how do we create safety and alignment and guardrails around them so that they are doing what we, what the human in the loop, want them to do. And it can be done? It absolutely can be done. Can you assure us of that? You know that the Hugging Face incident and the road ahead, as you describe it, is to some doomsday. You know, again, I don't want to lose sight of the positives. Look, I was in Texas this week at a small business event for women starting, running, growing their own businesses.

40:13And I spent time with one woman talking about health deserts and what she means by that. So she's a qualified doctor. She started in Alaska, moved to West Texas. What she sees is that if you live in a health desert, your life expectancy is 30 percent less than someone who's got active access to health care. So what's she doing with our technology? Number one, she's utilizing the fact that 330 million people every week ask ChatGPT a question about their health. We have a clinician tool now, so we're enabling doctors to quickly use AI. And then we work with hospital systems like HCA, Boston Children's, Stanford Children's, to be able to put that all together.

40:51She steps in as an entrepreneur and says, I am going to build a business that's going to bring healthcare, teledoctoring to places like West Texas. That's what I'm focused on in terms of where are we going with technology and great outcomes. There have always been risks to technology. We should take them seriously, but we shouldn't allow it to stop the progress that we're making. Now, I just I hear that and I say to myself, those who think that there will be jobs taken away don't understand. That is a creator of West Texas jobs that would not be there. More jobs, not fewer. And I mean, that's what we've seen so far when you look at the impact of AI on the economy at large.

41:32Our economic research team has done a great job on this. What we see is that people are actually starting to make their jobs more interesting and broader. So it's not just the work that, you know, if you look at someone's title, like a finance person. If you look at all of the calls into ChatGPT from people whose title is finance, only about a quarter of what they ask ChatGPT is finance. You know what the rest is? Another quarter is engineering. So my tax team is learning how to program so that instead of waiting for an IT person to come along and help them, they're doing it themselves. Faster, quicker, better.

42:09The work that couldn't be done before. Then they could hire more people. They're also, interestingly, spending 25 % of their time asking marketing questions. So I'm less sure what's going on there, but I do, as I always say to my finance team, we're all in sales, whether you know it or not. You might not be selling to an external customer, but you've got to go sell your point of view to your customer internally. And so if they're using some marketing techniques to be heard, to punch through, that's only good because you go from being a historian of the business to actually moving the business forward.

42:41I like this positive note. This is a good moment to pause. We'll be right back with more. Thank you. Stick around. We'll have more of Kramer's critical conversation with OpenAI CFO Sarah Fryer, next.

43:03Before the break, you saw the first part of our discussion with Sarah Fryer, the CFO of OpenAI, which is at the heart of so many issues facing the artificial intelligence complex. But we have much more ground to cover. Here's the rest of that conversation. Let's talk about the possibility of the impact of a pause. I think you've correctly talked about how it is something that has proved me like that. But this endless drumbeat that we could fall behind China reminds me of when we might have fallen behind Russia in 1959 to 63. And it turned out to be a false construct. What is it about China that could make it so that maybe it's open models?

43:45I don't know. That they own us or somehow our defense would be compromised. And do you believe it? I absolutely believe that we're in a very competitive landscape right now. And absolutely, customers come and talk about open source, open weights, and many of those models are Chinese models. But let's talk about what they're saying when they say that. Number one, it's cost. And so it's incumbent on us to keep driving down costs. So we do that with things like I already talked about, frontier models, training mini models. It's our own chip, being able to inference at a much cheaper price. The fact that we can be cheaper today, if you buy Luna, it's cheaper than taking a Chinese model like GLM 3.5 on Cloudflare.

44:27So number one is we need to be on the Pareto car. The customer can get the right cost for the right task. Number two, they want an ability to effectively go deeper, often like take off some of the less guardrails, but they want to tweak the wingtips of the model and have more of their own say on it. So again, we need to find ways to work with customers. So that's where verticalization starts to get interesting, right? Today we've gone vertical in areas like chip design because we had to do it for ourselves. Now we work with customers on it. Healthcare with GPT Rosalind. The third thing is data sovereignty.

45:00This has always been a thing since technology started. Where is my data? Is it safe? Is my intellectual property safe? I don't know if Chinese models are helping you on that front either, but I do know that we need to hear customers, particularly outside the U.S. I'm about to fly to the U.K. tonight. where people worry about what happens if I can't access that model or that data. And I think there's work to be done on that front. OK. Now, when I read, look, your papers are incredible. You put out amazing things. They're done in English, too. And people can read the jalapeno first result. And I was thinking when I read The Hugging Face, I said, oh, my God, the Chinese hat.

45:35It's the blueprint. It's the blueprint for the people who don't take the cybersecurity. Can't they send swarms to whoever doesn't do the cybersecurity, get in and shut the company down? If it's a utility, if it's a chemical plant, can't they do it? I mean, this goes back to why we think the defender window is closing and why there is a limit of time, frankly, to really harden our infrastructure. At OpenAI three weeks ago, we took Astra, pointed it at ourselves. And I think what we learned is, number one, massive empathy for our customers. Number two, we took 25 % of all of our engineers off of production-level tasks and said, your job now is to run alongside the model and fix as it finds.

46:16Number three, we've built some tooling now that we can go out in the world and help customers protect themselves from things like an agentic swarm. And then finally, we created playbooks. And by the way, we're not going to go by ourselves. I love that phrase. If you want to go fast, go alone. If you want to go far, go together. We are working with the whole cyber landscape. We put out the letter. Over 100 other companies signed on. And we think it's really important we do this as a larger cyber community. When we were all afraid of mass destruction from atomic weapons, we did come up with an international agency to try to regulate these things.

46:52Do we need an international agency? And by the way, that wasn't bad. No one thought that stopped the pace of H-bombs if we wanted to build them. But do we need it? Do we need it for what you do? So we have some very good agencies today. I would say the UK AISC is one. And I think even here, Cassie here in the United States, are two groups that are very good at looking at models before they launch. And as you know, post the kind of mythos moment from Anthropic, which I felt like was good in that it kind of created a blast out in the world that said, hey, we all need to pay attention. But it suggested that one way to control the issue was to keep it in a box and not let people to see it and use it.

47:34We are open AI because we are more open in our approach. We think security and safety comes from a broader sharing in the world so that more eyes are on things. I think you get a better result. But again, we are looking to agencies like UKIC, to CASI and so on, to be able to help us do that well and in a trusted way. We need to make sure that we bring folks along so that they trust that the outcome is something that the world wants. All right. Let's talk money. I look at what you're really doing. I think when you talk about the idea of no IPO, immediately people say, oh, my God, it's going to be horrible for Broadcom.

48:07They can't be. Actually, I think it's quite the opposite. I think you guys are making so much money that it's kind of if you were public, I would be looking at that annual recurring revenue. I say that's got to be one or two versus anthropic. You're so much more lucrative than any of the hyperscalers. Once you get this thing going, did you guys pick the right side in doing the models? and they did the wrong side in doing the pipes? So I certainly like our position, and I think we can build a really strong, durable business. On the IPO front, I would say we probably have, actually. Thank you. On the IPO front, look, you've got to run your own game, as you know, with an IPO.

48:44I've been on many a show with you where I've said an IPO is just a milestone in the journey, and I'm going to keep reiterating that. Look, we raised$122 billion in Q1 of this year to give ourselves maximum flexibility. And how much do you think you've gone through? We still have an incredible balance sheet, actually, of just cold, hard cash sitting there to be able to do, like, fund our partners as we buy product from them. But you're right about the business, too, right? It's a very diversified set of revenue streams. Consumer, small business, large enterprise. Very good margins, particularly when you get into the productization of that.

49:18Things like cyber, codex, chat GPT work. I hope you're using agents now in the morning, Jim, to get ready. My team is. And then from there, of course, there's more coming. Like look out for Dev Day coming at the end of this month, September 29th. You'll see a lot of new announcements from us. Because we're learning how to build on top of those models into what the customer really needs. So it could be in chat GPT, things like healthcare, personal finance. In the enterprise, it's how do we do more verticalization or even just office of the CFO. Right. This morning I was on, I won't name the company, but a very well-known car company talking to the CFO.

49:58We were nerding out on zero day closes. But CFOs are saying, how do I deploy this? CROs, CMOs. But I'm not hearing people saying this is how people this is how they can cut 20 percent of the workforce. I'm not hearing it. I mean, efficiency is key. There is no doubt we are doing things with our own products inside of finance, for example, that do limit the number of people we need. But in many cases, it's for jobs that I don't think of as like great jobs. OK, who wants to in procurement? We do twenty eight hundred credit checks. We were that was our run rate last year. The credit check is a pretty mundane piece of work.

50:38Like, do you want a junior analyst doing that or can we get a very intelligent model to do it for us? And the cost difference went from two hundred dollars a check to about 17 cents. That is a great ROI for a CFO. But there are also very intelligent things we want, too. I'll stay in my world. Technical accounting, super complex. When we go to the SEC for preclearance on things, you've got to write pretty intense memos. That is a great place to bring raw intelligence to bear, too. So it's not fewer people. I think it's actually enabling people to do more work at the edge, more work where intelligence is needed.

51:14And then you get a lot of the rote stuff done by the model and actually done at probably better scale and with more accuracy, frankly. OK, CFO, are you the one who decides? I know you put together the deal with AMD, which was a terrific deal. Stock AMD went up a great deal. And, of course, you got warrants. But before that, you had a$100 billion NVIDIA deal, which seems to be scaled back to$30 billion. You do a huge amount of work with Broadcom. Jalapeno is written about here. We had Hawk Tan talk about it yesterday. How do you divvy things out? How do you decide, you know what, this is the time to go to Broadcom?

51:46We can't keep slighting AMD. You know what, Jensen's terrific. Is it just totally on the basis of they have the best at that moment? Or don't you feel like you've got to need it to spread around? Because those companies are going to be around for a long time. Yeah, so we need, just as we have a strategy for diversifying out our revenue streams, we also have a strategy for diversifying our supply chain. And that is just good CFO risk mitigation. First and foremost, you never know when someone's supply chain is going to get gummed up. So you need to have multiple providers. You're right. There are better chips for different places in what we are doing.

52:22NVIDIA is still an incredible platform of accelerators for training. Astra, 100 ,000 GPUs in Stargate, Texas. But in Jalapeno's case, that is a chip very focused on inferencing because it is set up exactly for our models. So therefore, it is very efficient. Custom-made for you. By Broadcom. Thank you, Hawk and team. In the case of AMD, MI-455s have become a very viable alternative in many cases. We're working with partners like Cerebris and so on. So our goal is to really as much diversification as possible that allows for better outcomes for customers, better lower latency, better reliability, better pricing, because now I have some leverage with my supply chain.

53:08and then also just the risk mitigation of what happens if TSMC can't provide all the wafers. It sounds like you're spending a lot of money. You may be slowing down the pace of one aspect or two aspects, but it sounds like you're spending a ton of money where you can make big dollars. That is exactly it. In the end, I'm a CFO, so it has to come back to an ROI. The landscape is very dynamic, so we might have to shift our judgment on where to invest and when to invest, But we do that in a very dynamic way based on where we see the demand coming from at any point in time. If you think there's a demand coming from in some place to make money, then I know you are the right place.

53:46Then come public so we can all have a piece of it. Sarah Fryer, CFO of OpenAI. Sarah, it's just so good that you came on that money. It's such a pleasure. What a day. That's a wrap from Dreamforce. Like I said, there's always a bull market somewhere. And I promise I'll find it just for you right here on Mad Money. I'm Jim Pamer. I'll see you tomorrow.

54:09All opinions expressed by Jim Cramer on this podcast are solely Cramer's opinions and do not reflect the opinions of CNBC or its parent company or affiliates and may have been previously disseminated by Cramer on television, radio, internet, or another medium. You should not treat any opinion expressed by Cramer as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of his opinion. Kramer's opinions are based upon information he considers reliable, but neither CNBC nor its affiliates and or subsidiaries warrant its completeness or accuracy, and it should not be relied upon as such.

54:36To view the full Mad Money Disclaimer, please visit CNBC.com forward slash Mad Money Disclaimer. Something amazing is happening in data management. Even though storage needs are constantly changing, companies are only paying for what they need now. That's because CDW customized an ever pure evergreen one SLA driven storage as a service solution, delivering hybrid cloud storage that's easy to manage efficient and continuously evergreen. Simplify storage administration with ever pure and CDW make amazing happen. Find out more at CDW dot com slash ever pure.