Companies that leverage cloud to deliver their products. But it's an interesting, it's an interesting chart to see where enterprise AI adoption is going. This is from Ara Kharazian over at Ramp. He says Ramp AIs. Ramp's AI index shows AI adoption held flat at 45% of businesses in November, driven by slight declines in finance and technology sectors. By the model, OpenAI went down negative 1%, Anthropic up 0.8% and Google went up 0.7%. And so you can see that this chart in 2025, OpenAI had a big, big jump and then just has been flat for the last couple of months. A little bit down. Anthropic's been on a much smoother. It seems like they didn't exist in 2023, basically, or maybe they didn't have a paid plan that really appealed to businesses. Then 2024, you see some slight growth here. And then 2025, a much smoother, like, okay, uptick, growing more, growing more, growing more. Kind of all keeps up with this. You know, they've been 10xing every year like an Instagram hustle account. But OpenAI had a massive spike at the start of 2025 and then has been a little bit flatter. But it is interesting. I wonder how much you read into this. So Ara has a. As a take here we should go through. Before we do, let me tell you about Fin AI, the number one AI agent for customer service, automate the most complex customer service queries on every channel. So he says adoption is flat. Is the bubble popping? And he says, I'm not calling it yet. The slowdown comes at the end of a rapid run up in adoption rates in 2025, which coincided with a significant step change in the capabilities of these models. Now the effect of the latest advancements has faded. If we want to see another run up in adoption we've seen, we would have to see at least one or two step changes. Technological gains, the models get even better, spurring faster adoption or implementation gains. Early adopters figure out the best use cases for AI and the rest of the market follows, driving incremental adoption. Both are likely the latter even more so as adoption actually rose in several industries with relatively low adoption rates like retail, construction and manufacturing. Is Google's AI good now? He says it's still underrated. Adoption of Google's Gemini rose point, its second highest monthly increase on record. And then he says, what's going on with OpenAI? He says, he cautions an overreaction to the latest results. Which show OpenAI adoption dropping while Anthropic and Google adding new users and what that means for OpenAI's business. I've learned a lot covering how companies buy and use AI. Mainly this market changes rapidly in the typical rise and fall timelines of companies. Does not apply here. OpenAI enjoyed rapid adoption of growth in 2025 in many ways emerging as default spend category for businesses. And it's not unreasonable to think that rate would normalize as competitors find their fit in the market. Also there's just a lot of overall the interesting thing is every like it seems to be pretty obvious at this point that the model is just getting smarter, is not going to by default make adoption like accelerate massively again. They need to become more useful. They need to be able to. Duart Keshe has talked about this. They need to be able to learn on the job basically in the way that a human does. I think until we have that there's some the forward deployed engineer movement of having people out embedded in companies trying to unlock the potential of the models. That's great but I think the models actually just need to become generally more useful. Yeah, more useful. I don't know. Yeah, it's tough. I mean when I look at those, when I look at the number like the headline number, it's 45% have adopted have paid for AI chat apps. Basically. You know, he's tracking OpenAI, Anthropic, Google and Xai on there, a few others and you flip that around and you're like 45% of businesses don't pay pay for AI, don't use AI. And the question is like are they just not using AI at all or are they just good with free? Because if you told me like 55% of companies are in sort of like a private equity scarcity wartime mode where they're watching every dollar that goes out of the door. Even a $20 subscription is something that will be scrutinized because you know what, like the son we've been running this business for a generation. We're not gonna spend frivolously every dime that go business is. It is precious. And so we're not just gonna go sign up for whatever. No, you don't just spend, spend, spend. You need to really understand if this is valuable. And you know what if the particular business we have, we have a specific POS system we have, you know, we don't have that many unstructured questions, right? Like realistically, it's like we're not doing a lot of random email A lot of random research for our business, it's more like, you know, customer comes in, take their order, put in the POS system, charge them, go do the thing. Which AI is probably not replacing. So it's interesting, but it is. One of the cool factors about this R. Karazian report is that it comes from ramp customers who I don't think of as. I think of them as a little bit more forward thinking than the general population. If I were to say that there's bias in the data set, I would say it's biased towards like being open to AI. For sure. For sure, right? Yeah. If. I wonder if a company like Amex would look at their data of like AI adoption just for cohorts that signed up for Amex over a decade ago. Yeah. You would expect it to be lower. Yeah. Or, or even, or even putting Amex aside, like a, like a community bank. Like, because if I go to my local Laundromat and I ask like, what, what corporate card do you use? They're probably like, we don't do corporate cards here at all. In fact, we have a relationship with the local bank down the street and they put up the mortgage for the building and we write checks or we do cash. So they're even using a debit card. Yeah, exactly. Or if they do have a credit card, it's probably from the local bank branch. It's not like they haven't built like their whole system. So. Yeah, it's fascinating to read into. I do think there is something to be said to. I do think there is a bit of a slowdown here. Like it's not accelerating and it is real. That, that there is. It's just getting harder and harder to justify. Okay. It's no brainer to bring it in. It's a no brainer to pay for it. Part of that's the competitive dynamic of, you know, 50% of the value from AI. You just get. Also what are you getting out of, out of some of these. If you're a business that's not doing Cogen, you're not generating a bunch of assets, et cetera. And somebody says, hey, for $20 a month or $100 a month, you can get somebody that's PhD level in math, but extremely low agency. You have to like tap, you have to tap the person on the shoulder. I like the Laundromat. The Laundromat's like, actually, I don't need somebody who's PhD. You know what, sir, this is a Laundromat. We've been running this laundromat in this family for 45 years, and I've never run into an IMO gold medal level problem ever in the past. I bet there are. I bet there actually are. I bet, I bet, I bet, I bet, I bet if Mark Chen wanted to create a laundromat. Laundry allocation of laundry. You need to efficiently. I mean, if you have a big load. Yes. Laundry. Yes. How do you split it? How do you distribute it? Okay, yeah, yeah, no, that makes sense. But, but yeah, so, so you have, like, you have a robot on your computer that is incredibly, incredibly smart, incredibly low agency. It can't learn on the job. You have to, like, tap it on the shoulder and say, hey, do this. Okay, now do that. Now do this, do that. And it's like, well, a lot of these businesses don't have, again, these, these problems that need that level of intelligence, and they need people that are, that are even have some level of agency, not even maxed out on intelligence, but can just kind of like chug away and learn on the job and get. Better, or they need SaaS, products that have AI functionality built in.

Sam sa. To shape up. You came to this world. To reach the stars. To shape a future. We came to feel the new. The death match. Saund's back there. You're watching TVPN. Today is Thursday, December 11, 2025. We are live from the TVPN Ultradome, the Temple of technology, the fortress of finance, the capital of capital. Christmas is around the corner. That's why you're here. You want an update on Christmas 14 television. Two weeks from today. Presents under the Christmas tree. Ramp. Under the Christmas tree. Ramp.com under the Christmas tree. Time is money save. Both easy to use. Corporate cards, bill payments, accounting, and a whole lot more all in one place. Also, SpaceX IPO under the Christmas tree. 1.5 trillion for the big man. For Elon Musk, he's. He's gonna raise $30 billion, something like that. That's an incredibly small amount of dilution. If he actually goes out at 1.5 trillion. Should be interesting. I thought he hated just 1.5 million. Public company. Just 1.5. Just 1.5. Just one point five trillion. He said the biggest number. Because you know that there was that whole, like, leak, I think it was in Reuters. OpenAI going out at 1 trillion. You got to go a little bit higher. What's just a little bit higher? He should have done it. He should have. 1.1. I'm going to IPO. 1.1 trillion. Really stick it to Sam. But he didn't. He says he's going out at 1.5. He's got to leave some room for the. For the first day. Pump. Yes. Right. Get it up to 2. Get it up to 3. Yep. But I mean, in terms of companies that should benefit from the space economy, like, is this not the Nvidia of space? What other company is there? It's the space company. It's Space x. And so 30 billion coming in. Company was doing great. Launch, launch, launch. Product is fantastic. The rockets go up, they come down. Putting sonic booms over Montecito, apparently. But I think the actual launch market for anything but satellites, but for anything but data centers has just been a little small. And so Starlink has been the big unlock, of course, for SpaceX. Massive market there, putting the screws to Verizon, AT&T and the other telcos. The other telcos, of course, have been noodling on working with a rival, putting up smaller satellites in a sparser constellation with more bandwidth. Per. Somebody was talking earlier that to get the same amount of compute as a 1 gigawatt data center, you would need 10,000 satellites. That's not that many though. Starlink has over 10,000 up there already, I believe. Sure. I know that's actually not crazy. That seems like completely. That's actually like extremely bullish. Well, there is no. There is a pathway to doing it with far fewer which would just be like ultra, ultra long. I don't think they want to do big. I think the whole thing is Constellation. That was the beauty of Starlink was that you could go up there because it g gave them the ability to have all this residual capability, as Gwynne Shotwell puts it, which is like if the CIA shows up and they want to put a spy satellite on there or some company shows up and says like, hey, we want to put something really serious in space, but we only need 60% of the fairing, or we only need 40% of the fairing, or we only need 80% of the fairing, it's like, okay, great, that's. That will fill the rest of the space with two starlinks, five starlinks, 20 starlinks. Well, we have the founder of K2 Space coming on the show later. Raised a quarter of a billion dollars at a $3 billion valuation from T. Rowe Price. Fantastic head of Sophia Altimeter Lightspeed and some others. So we will talk more about that then. And we will also talk about fall generative media platform for developers develop and fine tune models with serverless GPUs and on demand clusters. So in terms of the space, I think, I mean we keep going back and forth on this. Is it just a pump? Is it just a new narrative? Is it a pivot to AI? I don't really care. I think it all comes down to this just idea of like Elon math. You know, the Elon math is always crazy. He's always putting up sort of wild predictions that then may or may not come true. A lot of them are way off. But at least he's telling a story that's like optimistic about the future. And at least it feels like that Trey Stevens good quest thing is like get to Mars is good. If he misses it by five years, at least he's still the first. He's the only person that cares about it, which is great. I went through, I tried to pull in true hit piece fashion. I pulled up all the different times he's made a prediction and then not landed it or not delivered on it time. And it's crazy. The guy loves to rip predictions. He said he was Gonna put early SpaceX concept when they started SpaceX was to put a small greenhouse with plants on Mars as a PR demo before they would even start the company. That of course did not happen before starting the company. Yeah, this was like the original thesis was like he went to some Elon went to some like space kind of space nerd meetup and was like we need to put a plant, a physical plant with a webcam on it on Mars. And if we that there and we prove that life can live on Mars, we bring life to Mars. It will inspire everyone and marshal all the resources, all the capital and all the excitement to actually go put boots on the ground. Which also he predicted. He predicted humans on Mars by 2024. Of course, hasn't happened that we know of. It's possible. Snuck one up there. They keep saying that the star. They keep saying the starships blow up. But maybe, maybe one of them didn't. Maybe one of them sneaky, sneaky went off and put a person on Mars. Yeah, you don't know the shell of the real ship. I like the inverse, the inverse conspiracy theory, you know, like, like oh yeah, we never went to Mars or we never went to the moon. No, the inverse conspiracy theory. We're going all the time, secretly. But you're not cool enough to get in on it. You don't have a ticket. Everyone else does. Except for you. Because I was just on the moon yesterday and I'm not telling you about it first crewed Mars landing as early as 2029. Little too early to say, but that one feels aggressive. We're just four years out. You gotta put a whole crew down there. He also said tourist trips were gonna be happening around the moon by 2018. Didn't hit that. Yep, self sustaining city on Mars of 1 million people. 2050 feels a little early, but I. Don'T know, 2050, anytime you're getting more than a couple decades, couple decades, anything's possible. I agree, I agree. But then of course there's a bunch that he succeeded in. He said that he was going to deliver reusable orbital rockets. He did it. Yeah. And what else matters? At least you know, you, you know. He's also predicted fully reusable starship with rapid turnaround. It's a little too soon to tell. He's behind some of the early dates. But you know, in general that project seems like it's real, it's going to happen. It's taken a little bit time. He has, he has a bunch of other funny ones. But it is interesting how the SpaceX narrative has shifted from first we're going to Mars. Then there was the whole. Do you remember SpaceX? Point to point. Are you familiar with that? So there was a pitch for a. While, like a commercial. Is this like, this is a replacement for an airliner? You'll just go up and then come down and land. Exactly, exactly. So it was like New York to Tokyo in like 30 minutes. Because I would go so hard. Insane, right? Insane. So the idea was like you're in Netherlands. Imagine being able to commute to Tokyo for some obscure job. It'd be amazing. It would actually be remarkable. And there's a whole bunch of economic research that shows that when you increase transportation, reduce transportation times, that you really do get an economic boom. Because people can do business in areas where they couldn't before. Even just the reducing the time from San Francisco to la, from a six hour drive to a one hour flight, all of a sudden people can go up and do business, come back the same day. They just get there more often. Yeah, I had a hot take for a while that if I was the President, I would recommission the SR71 Blackbird, which of course goes, I think Mach 3. So you can actually get from DC to London in like two hours because it's a supersonic jet. And so imagine that, imagine the, the. Presidential SR71 would go also very remarkably hard. Right? And imagine the aura of, you know, something happens in. And I remember I thought of this because the Queen had just passed away very sadly and of course the President was planning on visiting the Queen, paying respects to all the people of Great Britain. But imagine if, you know, the news breaks and it's like, this is important. I'll be there in two hours and I'm there on the ground like same day the news breaks. Not, oh, we got to charter the 747 and bring the whole crew and land and I'll see you in two days and it'll be a whole thing. It's like, it's like jumping in the car. Yep, yep. Or imagine these trade deals. These trade deals. Oh, the H2 hundreds are going to China or something's happening in Taiwan. Hey, Xi Jinping. So what happened with the Point to Point program? Has there been any. Oh, well, it's all just predicated. I mean, so the space, the starships are still blowing up, so no one wants to get on those. But in theory, if starship. But there was nothing that they never had any comms. It was like, we're scrapping this. No, no, no, no, no, no, no. Not at all, not at all. It was just like an interesting, it was an interesting, like back of the envelope, like, bull case for this, for like the business basically just saying like, okay, yes, starship is going to be really good at getting to the moon, getting to Mars, getting to, getting a bunch of Starlink satellites in orbit. Right. It's like a good, it's a good business. But Elon is always the king of opening up new markets and just kind of saying like, well, what if, what if he was also replacing commercial air travel? Yeah. And people are like, wait, that's a huge, huge market if you can do that. And of course, if you could make it more economical, who would ever pay for a ticket on a 747 when you could ride a rocket and get there literally 90% faster? Like, it's crazy the amount of time that you would save. Now you have to go from Manhattan on a boat out to a launch pad. You can't take off from jfk. Like there's a whole bunch of wrinkles. And that's why this is all practically probably like 20 years in the future. But it was one of those famous examples of an Elon project where if it doesn't violate the laws of physics, it eventually will happen. And that's a lot of what's happening with the data centers in space. We need to inference things. We need to inference things like cognition and the team, because they are the team behind the AI software engineer. Devin, crush your backlog with your personal AI engineering team. Right now, Devin's inferenced on Earth. Who knows, maybe in the future it's inferenced space. I think that the space data center thing, people are still having the debate on, like, is it possible? It's like, obviously it's possible. The question is over under a gigawatt by 2027, which is when the other big clusters come online. Will it be competitive in the short term, in the near term? And then on the flip side, like, if you're a bear, are you saying that it's not going to happen in 20 years? And then, I mean, the big question is there's one more Elon gambit that he could run with. One more like, oh, you wanted a fifth act? Just one. I mean, there's tons, but there's one that's really wild, which would be if he straight up said, we're building the Dyson sphere, like we are going to launch. Tyler's nodding. He's pumped. He's pumped. Isn't he the obvious. Isn't he like the only, only real candidate? Well, he's the only candidate for everything in space because he has the leading space company by like it's two orders of magnitude. I think someone at Anthropic has said that the Dyson Sphere is in the, in the plans of Anthropic of Anthropic. They weren't trolling. I mean, I don't know. They're. They're going to need to acquire a space company if they want to do that. Because how do you get the. Isn't the Dyson Sphere just a bunch of solar panels around the sun, directly capturing 100% of the energy that's put off by the sun? Yeah, I don't think it has to fully. Like, it's not like there can't be any light coming out. It's just like strips of. Oh, I thought Kardashev 1 is you're capturing 100% of the energy that's hitting your planet. Kardashev 2 is you're capturing 100% of THE energy that is given off by your star. Yeah, actually. So I think maybe what I'm talking. About is there's a different term. Maybe they're do. Maybe there are two different things. Maybe Dyson Sphere could just be, okay, you're only capturing 50% or 10% and then Kardashev2 is you're capturing 100% of the energy coming off your sun. Tyler, the chat wants to know, did we ever find out if this kid, I'm assuming they're talking about you, is related to Elon? Why would he be related to Elon? I mean, he's got a lot of sons. He's got a lot of sons. Oh, okay. Yeah, could be anyone, maybe. Should we read through this? So, Eric Berger over at Ars Technica. You should also read through the daily newsletter, Tech analysis and news daily from tvpn. We're doing ad reads for our own newsletter now. We love that. Get our daily op ed, top headlines and best posts from the timeline. Every. I can't read the rest of this. Tyler, you got every day. Tvpn.com tvpn.com Go subscribe. Thank you. So Eric Berger is writing a piece called after years of resisting it, Space X now plans to go public. Why? And Elon actually replied to Eric and said, as usual, Eric is accurate. So I thought it was worth reading through this. Yeah. Eric says SpaceX is planning to race tens of billions of dollars through an initial public offering next year. Multiple outlets have reported and ours can confirm this represents a major change in thinking from the world's leading space company and its founder, Elon Musk. Wall Street Journal and the Information first reported about a possible IPO last Friday and Bloomberg followed that up on Tuesday evening with a report suggesting a $1.5 trillion target valuation. This is an enormous amount of funding, the largest IPO in history incurred in 2019 when the state owned Saudi Arabian oil company began began publicly trading as aramco and raised 29 billion in terms of revenue. Aramco is a top five company in the world. Now. SpaceX is poised to potentially match or exceed this value. That SpaceX would be attractive to public investors is not a surprise. It's the world's dominant space company and launch space based communications and much more. For investors seeking unlimited growth, space is the final frontier. But why would Musk take SpaceX public now at a time when the company's revenues are surging thanks to the growth of the Starlink Internet constellation? The decision is surprising because Musk has for so long resisted going public with SpaceX. He has not enjoyed the public scrutiny of test Tesla and feared that shareholder desires for financial return were not consistent with his ultimate goal of settling Mars. Next section is called Data Centers. R spoke with multiple people familiar with Musk and is thinking to understand why he would want to take SpaceX public. A significant shift in recent years has been the rise of AI which Musk has been involved in since 2015 when he co founded OpenAI. He later had a falling out with his co founders and started his own company Xai in 2023. At Tesla he's been pushing smart driving technology forward and more recently focused on robotics. Musk sees the convergence of these technologies in the near future, which he believes will profoundly change civilization. Raising large amounts of money in the next 18 months would allow Musk to have significant capital to deploy at SpaceX as he influences and partakes in this convergence of technology. How can SpaceX play in space? In the near term, the company plans to develop a modified version of the Star like satellite to serve as a foundation for building data centers in space. There you go. John Musk said as much on X in late October. SpaceX will be doing this, but using a next generation Starlink satellite manufactured on Earth is just the beginning of this vision. The level beyond that is constructing satellite factories on the moon and using a mass driver electromagnetic railgun to accelerate explain to us to accelerate AI satellites to lunar escape velocity without the need for rockets, musk said last weekend on X. That scales to 100 terawatts a year of AI and enables non trivial progress towards becoming a Kardashev2 civilization. Based on some projected analysis, SpaceX is expected to have in the neighborhood of 22 to 24 billion in revenue next year. 65 times revenue is what you'd be paying if SpaceX goes out at 1.5 trillion. For reference, 65 for 65x revenue. For SpaceX, Nvidia is a 24x revenue to market cap, market cap to price to sales. So like twice more, twice as expensive as Nvidia, but also probably even wider lead to the rest of the industry, newer industry. I don't know about the margins, but potentially higher margins. And again, mass markets articles talking about space being more vertically integrated. Anyways, that is a lot of money. It's on par with NASA's annual budget. And SpaceX can deploy its capital far, far more efficiently than the government can. So the company will be able to accomplish a lot. But with a large fusion of cash, SpaceX will be able to go much faster. This is the thing that I'm not certain on. Like is SpaceX has SpaceX been capital constrained or are they capability constrained? Like I don't. It hasn't been clear to me. You know, clearly there's been like infinite demand for SpaceX equity for a very long time. So much so that people are, you know, investing in triple layered SPVs just to get, you know, and enduring massive fees just to get a taste. So yeah, the question is like how does this massive capital infusion actually accelerate the business that much? Or is it just the right moment in time for a number of reasons. Well, you know that SpaceX, whenever they launch, they stream their launches. They gotta be on Restream one livestream, 30 plus destinations. If you want to multi stream, go. To restream.com Abhi Tripathy, a longtime SpaceX employee who is now Director of Mission Operations at the UC Berkeley Space Sciences Laboratory, believes that once Musk realized Starlink satellites could be architected into a distributed network of data centers, the writing was on the wall. That is the moment an IPO suddenly came into play after being unlikely for so long. If you have followed Elon's tactics, you know that once he commits to something, he leans fully into it. Much of the AI race comes down to amassing and deploying assets that work quicker than your competition. A large war chest resulting from an IPO will greatly help his cause and disadvantage all others. Yeah, what's interesting is so I totally get that SpaceX was never crazy, crazy capital constrained. Even though like building a massive rocket that feels like so it feels so expensive. But at the same time they've just, they've just been able to raise the money and it just hasn't been a problem. And of course the, like SpaceX as a business generates a lot of revenue. Like some of the, some of the, some of the contracts with the government are in the billions. And then Starlink is just throwing off, it's telecom, so it's just throwing off tons and tons of cash. I mean it's incredibly expensive to get the Constellation up there, but you do have cash flow from it and it's effectively, you know, like I'm, I've been subscribed to Starlink for like a year or two. I never use it. I'm basically 100% margin for them. Right? Yeah. And I use it just as like a backup in case the Internet goes down. I have it. Yeah. And you just look at, you look at the growth opportunity in effectively telecom you have. T Mobile is a $217 billion business that is threatened by Starlink. AT&T is $172 billion business and Verizon is $170 billion business. But I think the reason you need 30 billion, like sort of paradoxically or counterintuitively might not be to build more rockets, it might be to actually just buy the GPUs, because like in theory, if you're putting a GPU in space, like Elon, SpaceX doesn't make those GPUs, they don't just pull them out of thin air, they got to buy them from somewhere. And if they're buying them from Nvidia, those are expensive and you got to buy a lot of them. And to make a dent in like, let's assume all the physics works, let's assume that the launch costs are cheap, all the math works out well, if you're putting a gigawatt of compute in space, you probably have like a billion dollars of hard cost just on the chips or something like that. I don't know the exact number, but you could imagine that you raised 30 billion. The business is humming, but you still have to outlay a ton just for the GPUs. And so yeah, you do have a new, a new consumption for cash. Anyway, let me tell you about public.com, investing. For those who take it seriously, they got multi asset investing that's trusted by millions. Did you want to, do you think. That XAI gets rolled into SpaceX or Tesla? That's the hard part. I thought it was going to be Tesla for sure because Tesla has The lineage. They have their own chip, they have massive data centers to train AI models for full self driving. It felt like such a logical place for XAI to land. But now maybe it lands at SpaceX, I don't know. Yeah, if SpaceX becomes the data centers and space company and they're effectively like reselling also. I mean it feels like there's some sort of arcane truth about consolidation where it might be less of a headache than people think to have different entities. Like if you think about, I mean certainly like Stargate is a separate entity from. I'm not even thinking about it as a headache. It's just more so like XAI needs to get quite a lot more traction I think in enterprise and consumer. Yeah, yeah. Effectively somewhere needs to become like the dominant lab in at least one domain. Otherwise I think it will just make sense to roll into one of these. Other yeah, well we should read through some of the history about Founders Funds investment in Space X because it might be one of the greatest. It might be. It might become the greatest venture capital investment of all time. It certainly feels like it's in the running. This from Nico Wittenborn says founders fund investing 20 million out of a 20 out of a $220 million fund two which also invested in Palantir and Spotify early, is just nasty. And Dan Pramack says flashback to chatting with Founders Fund Luke Nosek when he led the first investment in SpaceX from an old Peweek Wire newsletter in 2008. I can't believe Peweek is. I read that newsletter back not in 2008, but like back then. Wow. Primax been in the game for a long time. So says. I spent some time on the phone earlier this week with Luke Nosek of the Founders Fund. The Founders Fund, which I used to have. They had the. The. They dropped it. And then there's also an apostrophe here which I think has been lost at some point. I don't know. To discuss his firm's $20 million investment in private space launch services provider SpaceX. The company is run by Elon Musk, who co founded PayPal along the founders Fund Partners. A few questions and answers, Dan. Founders Fund usually invests no more than a few million dollars in a company. Why invest 20 million when your fund is only 220 million? Luke we obviously have internal caps that we follow. What we do is look at how much a company needs and invest that much. For example, we only invested $500,000 in Facebook because that's how much it Needed at the time, a lot of firms were offering more. Dan says most of the companies you back have some sort of Internet angle, which makes sense given your PayPal background. But what do you know about rocket science? Oh, there's no Internet angle at SpaceX product. No one predicted the Internet angle. But it is remarkable that a bunch of Internet guys fund a rocket company and then it just becomes an Internet company and it's an ISP and now it's a data center company and everything collapses down to the Internet. And it's the Joe Weisenthal take. AI will get every resource that it needs like the Internet needs. The Internet is the business model for everything. Doesn't matter. If you invest in a space company, eventually you're gonna be an Internet business. And so Luke answering about the rocket science question, he says the most important aspect to us is the team. Remember, the PayPal founders didn't have banking experience. Dan Elon has put over $100 million of his own money into this company. That is crazy that they'd already burned 100 mil. And then founders fund comes in with 20 and developed both the Kestrel and Merlin engines, which are the first new rocket engines developed in the US in a very long time. When we did due diligence, we wanted to speak with other investors of successful new rockets, but they were all dead. I love that. Yeah. So we have to. We're going to be the first here. Dan, your announcement of the funding just came days after SpaceX had yet another launch failure. Any worries? Dan, Any worries? Luke says. Dan? Luke says no. There are obviously going to be some technical kinks like an exploding rocket, but they get amplified because rocket science is more binary than would be a technical kink with a website either gets into orbit or it doesn't. The team has some very smart people. And they'll make Anybody that ships just regular old SAS should have a huge amount of sympathy for anyone in hard tech. Because launching any type of app and not having a single bug, even a single critical bug is tough. At least early days. And of course critical bug in aerospace is quite explosive. Anyways, I did want to pull up just the estimates. FF is estimated to own 10.4%. We can just keep it at 10. That means that this initial again, they've invested in multiple rounds since then. Historically, the greatest venture investment ever in my view was MASA's 20 million investment in Alibaba. That's true. Turned it into 70 billion at the time of the IPO. Hard to. I wonder if we can figure out how much. FF owns about 10% of SpaceX, but. I don't know how many, how many they've put in. Oh yeah, yeah. To maintain that stake they put. So again, it's still like going to be a 150ish billion dollar position. Just the carry check alone will be nasty. As Nico says, delightful. Wow, you hit the gong for the old FF crew. My former boss. Yeah, let's hit the gong. John's former boss. My former boss. How does he do it? How does he do it? Yeah, really quickly. Let me tell you about Profound and there's big news today from Profound, the company that helps brands show up more often and more accurately in AI platforms like ChatGPT. They've just announced their new workflows product which lets marketers automate jobs like research, reporting and content generation to drive even greater discoverability. Wherever people are searching, you can check out workflows and marketing that runs itself. I was texting the CEO of Profound James last night and he was giving me an update on the business and he shared just like a clearly like copy and pasted it like the new customers that they had signed. And I don't think we, I don't think we can share all of them yet, but it basically looked like every company in every company, every big company, it looked like they had them. They already are working with MongoDB, indeed. Mercury, DocuSign, RAMP, Zapier8Sleep, US Bank, Figma, SIKI, LG. So many of these different companies are choosing Profound and we feel lucky to be partnered with them. Here's a good post from just another pod guy. We could sprinkle this in before we move on to Oracle. Just thought this was an interesting point of view. Elon is not really a person. He is the human embodiment of network effects. When you've been operating at that level and breadth that he has been, you aren't really dealing with an individual. You're dealing with a deep bench of talent and capital with a comms guy who is addicted to Twitter. The same warnings about underestimating trump the person vs Trump's cabinet apply to Musk, but 100x and with far less irrationality injected and with a much higher caliber of people. Does he over promise over hype, over hyperbolize? Of course. But I would guess the median IQ and depth of expertise of the bench around Musk is higher than that of any nation state. I thought this was obvious to most people. So yeah, it's just crazy how like yes, you just have to accept the over promising the overhyping and stuff, but recognize that it's like no one else is trying, really. There's very, very few people. Oh, they're trying. Yeah. But in terms of like the real, the real people making a run at a lot of these crazy ideas, like, they're just, it's incredibly thin. It's incredibly thin. There aren't that many people. There are a lot of people that. But yeah, just don't take it seriously. But anyway, let's move over to one of Elon Musk's best friends. I believe they're boys. They're boys. I think they're boys. Elon can count on Larry for a billion or two. Over text so Larry Ellison is addressing, he's sitting on stage here in this Wall Street Journal article on the news that Oracle shares have tumbled as AI spending outruns returns. So Oracle is facing mounting anxiety from investors how much it's spending to build out data centers for the artificial intelligence industry. The cloud computing company's revenue and operating income for the most recent financial quarters fell slightly short of analysts expectations, while the company raised its spending forecast, adding fuel to concerns over the timeline for turning the AI industry's ravenous demand for computing capacity into profits. Before we continue this article, let me tell you about graphite.dev, code review for the age of AI. Graphite helps teams on GitHub ship higher quality software faster. So I mean, this is all part of this. I think investors didn't realize that, like when you do one of these crazy AI deals, you have to go build the data center and then you have to wait and then you have to get power and you have to buy the chips and chips have to ship, turn it on, rack of chips and you got to turn it on, then you got to test it and then you can generate some tokens and hopefully sell them all. Well, enterprise AI adoption might be stagnant. See, there was some crazy data out of Ramp today, our Karazi and the Economist over there. At least some businesses might be saying, we got enough AI, we like it, but we have enough. Apparently 55% of businesses are just like good without paying for AI. I mean the paying thing, at least directly, at least the paying thing here is so and so. The example is like a lot of businesses don't pay for cloud. Yes. But they effectively pay for cloud because they pay companies that leverage cloud to deliver their products. But it's an interesting, it's an interesting chart to see where enterprise AI adoption is going. This is from Ara Kharazian over At Ramp, he says Ramp AIs. Ramp's AI index shows AI adoption held flat at 45% of businesses in November, driven by slight declines in finance and technology sectors by the model. OpenAI went down negative 1%, Anthropic up 0.8% and Google went up 0.7%. And so you can see that this chart in 2025, OpenAI had a big, big jump and then, and then just has been flat for the last couple months. A little bit down. Anthropic's been on a much smoother. It seems like they didn't exist in 2023 basically. Or maybe they didn't have a paid plan that really appealed to businesses. Then 2024, you see some slight growth here. And then 2025, a much smoother like, okay, uptick, growing more, growing more, growing more. Kind of all keeps up with this. You know, they've been 10xing every year like an Instagram hustle account. But OpenAI had a massive spike at the start of 2025 and then has been a little bit flatter. But it is interesting. I wonder how much you read into this. So Ara has a as a take here we should go through. Before we do, let me tell you about Fin AI, the number one AI agent for customer service Automate, the most complex customer service queries on every channel. So he says adoption is flat. Is the bubble popping? And he says, I'm not calling it yet. The slowdown comes at the end of a rapid run up in adoption rates in 2025, which coincided with a significant step change in the capabilities of these models. Now the effect of the latest advancements has faded. If we want to see another run up in adoption, we would have to see at least one or two step changes. Technological gains. The models get even better, spurring faster adoption or implementation gains. Early adopters figure out the best use cases for AI and the rest of the market follows driving incremental adoption. Both are likely the latter, even more so as adoption actually rose in several industries with relatively low adoption rates like retail, construction and manufacturing. Is Google, is Google's AI good now? He says it's still underrated. Adoption of Google's Gemini rose 0.7%, its second highest monthly increase on record. And then he says what's going on with OpenAI? He says he cautions an overreaction to the latest results which show OpenAI adoption dropping while Anthropic and Google adding new users and what that means for OpenAI's business. I've learned a lot covering how companies Buy and use AI mainly. This market changes rapidly in the typical rise and fall timelines of companies does not apply here. OpenAI enjoyed rapid adoption of growth in 2025 in many ways emerging as default spend category for businesses. And it's not unreasonable to think that rate would normalize as competitors find their fit in the market. Also there's just a lot of. Overall, the interesting thing is every like it seems to be pretty obvious at this point that the model is just getting smarter, is not going to by default make adoption like accelerate massively again. They need to become more useful. They need to be able to, you know, Dukesh has talked about this. Like they need to be able to learn on, learn on the job basically in the way that a human does. And I think until we have that there's some things, you know, the forward deployed engineer movement of like having people out embedded in companies, like trying to unlock the potential of the models. That's great, but I think the models actually just need to become generally more useful. Yeah, more useful. I don't know. Yeah, it's tough. I mean when I look at those, when I look at the number, like the headline number, it's 45% have adopted, have paid for AI chat apps basically, you know, he's tracking OpenAI, Anthropic, Google and XAI on there, a few others and you flip that around and you're like 45% of businesses don't pay for AI, don't use AI. And the question is like, are they just not using AI at all or are they just good with free? Because if you told me like 55% of companies are in sort of like a private equity scarcity, wartime mode where they're watching every dollar that goes out of the door. Even a $20 subscription is something that will be scrutinized because you know what, like son, we've been running this business for a generation. We're not going to spend frivolously. Every dime that goes through our business is precious. And so we're not just going to go sign up for whatever. No, you don't just spend, spend, spend. You need to really understand if this is valuable. And you know what if the particular business we have, we have a specific POS system we have, you know, we don't have that many unstructured questions, right? Like realistically, it's like we're not doing a lot of random email, a lot of random research for our business. It's more like, you know, customer comes in, take their order, put in the POS system Yeah, charge them. Go do the thing. Which AI is probably not, you know, replacing. So it's interesting. But it is one of the, one of the cool factors about this R. Karazian report is that it comes from ramp customers who I don't think of as. I think of them as a little bit more forward thinking than the general population. Like, if I were to say that there's bias in the data set, I would say it's biased towards like being open to AI. For sure. For sure, right? Yeah. I wonder if a company like Amex would look at their data of like AI adoption just for cohorts that signed up for Amex over a decade ago. You would expect it to be lower. Yeah. Or even putting Amex aside like a community bank. Because if I go to my local laundromat and I ask like, what corporate card do you use? They're probably like, we don't do corporate cards here at all. In fact, we have a relationship with the local bank down the street and they put up the mortgage for the building and we write checks or we do cash. They're even using a debit card. Yeah, exactly. Or if they do have a credit card, it's probably from the local bank branch. It's not like they haven't built like their whole system. So. Yeah, it's fascinating to read into. I do think there is something to be said to. I do think there is a bit of a slowdown here. Like it's not accelerating and it is real that there is. It's just getting harder and harder to justify. Okay. It's no brainer to bring it in. It's a no brainer to pay for it. Part of that's the competitive dynamic of, you know, 50% of the value from AI, you just get. Also, what are you getting out of that? Out of some of these. If you're a business that's not doing code gen, you're not generating a bunch of assets, etc. And somebody says, hey, for $20 a month or $100 a month, you can get somebody that's PhD level in math but extremely low agency. You have to like tap. You have to tap the person on the shoulder. I like the Laundromat. The Laundromat's like, actually, I don't need somebody who's PhD. You know what, sir? This is a laundry mat. We've been running this laundromat in this family for 45 years and I've never run into an IMO gold medal level problem ever. In the. I bet There are. I bet there actually are. I bet, I bet, I bet, I bet. I bet if Mark Chen wanted to create a laundromat laundry. Allocation of laundry, you need to efficiently. I mean, if you have a big load. Laundry. Yes. How do you split it? How do you distribute it? Okay, yeah, yeah, no, that makes sense. But yeah, so you have like, you have a robot on your computer that is incredibly, incredibly smart, incredibly low aging. It can't learn on the job. You have to like, tap it on the shoulder and say, hey, do this. Okay, now do that. Now do this, do that. And it's like, well, a lot of these businesses don't have, again, these, these problems that need that level of intelligence and they need people that are, that are even have some level of agency, not even maxed out on intelligence, but can just kind of like chug away and learn on the job and get better. Or they need SaaS products that have AI functionality built in. Right. And so to them it's just, oh, yeah, my, my POS system just flagged a couple more fraudulent orders and took care of that. Or like ramp. You know, it's like, yeah, my receipts get classified better, but I'm not like a direct user of AI. Maybe they need to be on numeral. Numeral.com compliance handled numeral worries about sales tax and VAT compliance so you can focus on growth. Disney is investing $1 billion in OpenAI. Before that. Before that. I do think it's notable. So Oracle's remaining performance obligations, which we know refers to contracted revenue not yet recognized, is $523 billion. Currently their stock is trading at 500. Worthy of a gong hit, sure. But currently the stock is trading at only $568 billion. And so I just look at this as like, Oracle continues immediately. When they announced this backlog, they got a bunch of credit. It was almost immediately. And it has basically just been downhill ever since. As the market has basically said, we're giving you less and less and less and less credit here. And to the point where in many ways they're getting like, negative, negative credit for the partnership. That was the Financial Times take. The deal is worth negative 7 billion. Yeah. And there I think that the way that they framed that title was not entirely edgy. He's definitely edgy. It's a blog. It's Alphaville. They're having fun. But. And then I. So I have one more. I have one more post here that's relevant. So suspended cap says, I gotta feel like that at some point. Oracle, Larry was misled Was it OpenAI showing him something in the lab? Was it Jensen lying about token prices? Or was it a partnership that was supposed to happen, that didn't? It's always been clear to me that this was a race to the bottom. Why would you want to take these economics that someone as sophisticated as Microsoft was happy to pass off? Remember, I've said this a bunch on the show. Satya has more information. He owns OpenAI's IP, right? He's been serving OpenAI models. He's been, you know, running an AI. You know it's one of the most scaled platforms, right? He has more information than anyone in the room. Momentarily, he looks. He didn't look at when Oracle announced this backlog and the stock jumped however many hundreds of billions, everyone was kind of like, wait, did Satya miss this? And I think the market is maybe realizing some of his wisdom. Yeah, Suspended cap says Jensen's whole thing is that the rack costs 4 million and you make 20 million, you net 16 million. No one can make you 16 million. Got it. But a lot of assumptions baked into that. Someone has to pay you 16 million for however many trillions of tokens you can spit off. If a token is a commodity and someone can serve it at 50% lower than now, it's net plus 6 million on the 4 million of costs. What if the dollars paid for the infra results in massive losses for whoever is purchasing them? Is the assumption that the capital will just keep funding it? Maybe they will. I actually think they will. But I'm just confused at what Oracle thought they saw here, so. Well, story is not over. Let's move on to TurboPuffer, serverless vector and full text search. Belgram first principles on object storage. Fast, 10x cheaper and extremely scalable. So, Joanna Stern has a little bit of a screenshot take joking around about Disney, which is investing $1 billion in OpenAI and is going to license their characters for use in Sora. She says, look at this hype. Isn't it neat? Wouldn't you think My bubble's complete. I think that's from some Disney song. I actually don't know, but just the cadence. Do you have any idea what that was? Santa Claus is coming to town. No, definitely not. Santa Claus is coming to town. That's not a Disney property, is it? Oh, I. Oh, yeah, look at. But the tune you were singing sounded like that. This is because I'm bad at singing, I suppose. Okay, let's see if Grok can help grok it. Let's see. Okay. Anyway, there Is news. Oh, is this Frozen? Maybe. Look at this. Isn't it neat? Do, do, do, do. Okay, accompanying. I don't know. Someone will have to tell us. Anyway, so Disney's making a $1 billion investment in OpenAI and will allow the AI platform to use its characters and properties to generate short user prompted social videos. He says it's from the Little Mermaid. The Little Mermaid. What is the original line from the Little Mermaid? Look at this. Okay, Anyway, Disney's three year licensing deal will let users generate videos using Sora, OpenAI's short form AI video platform of more than 200 Disney, Marvel, Star wars and Pixar characters. A curated selection of these short videos will be available to stream on Disney. Whoa. That's the bombshell there. I didn't realize that because when I think about like going to Sora, it's very fun to make a mashup video of Mickey Mouse fighting Iron man or something like that. And that's something that it doesn't make sense to underwrite within Disney. They're not going to spend the time. The CGI is too expensive. It's just not going to happen. But for a kid who loves a Pixar character that wants to see Wall E join the Avengers and go fight Thanos, that could be a delightful, delightful experience for a young child who enjoys Disney's ip. And yet it could make no sense for Disney to actually create. That makes a ton of sense that they would allow this to happen in Sora in sort of a. In a way where the economic value is accounted for appropriately. Now you can is this kind of deal slop? But I just think the idea that this is going to Disney is crazy. That's the crazy thing. I mean, we covered this when we were talking about Sora generally. Just this idea that is it a creation tool or is it a consumption tool? And we were saying I think pretty quickly that this doesn't feel like an app that we're gonna hang out in. Everyone in the studio was having a lot of fun. We checked their screen times. A few days later, no one was scrolling Sora or more meta vibes for that. But we continue to see outputs from Sora going viral on other platforms. Totally. Even within our own group chats. What's been the most impactful Sora video that you've probably seen in the last couple weeks? Me making a Sora video of myself promoting John Fio's energy drink company in a suit. And I just sent it to him directly in a group chat. And it didn't exist and it wasn't funny in any broader context. Now the weird thing is that is there real demand for SORA level quality content in the Disney app? Like, when I open the Disney app, I feel like it's a place that has extremely high polish and I feel. Like ultimately to me what's exciting about this is personalized entertainment. And which is why I think that Disney would be interested in doing this if Tyler figured it out with Henry's help. The line earlier was, look at this stuff. Isn't it neat? Wouldn't you think? My collection's complete from the Little Mermaid part of your world. But if somebody loves the Little Mermaid and then you can prompt a really high fidelity personalized video for like your kid saying like it's time to do the dishes and singing a little song, right? That's the kind of, kind of moment that I imagine Disney would be excited about. Of course this will be immediately abused in a thousand millions of different ways, but I'm sure they'll put a lot of guardrails in. One thing that's notable, this went out this morning, Disney accuses Google of using AI to engage in copyright infringement on massive scale Disney sent a cease and desist letter to Google demanding it to stop the infringement. So yeah, basically the article says, as Disney has gone into business with OpenAI, the Mouse House is accusing Google of copyright infringement on massive scale, using AI models and services to commercially exploit and distribute infringing images and videos. On Wednesday evening, attorneys for Disney sent a cease and desist letter to Google demanding that Google stop the alleged infringement. Quote Google's infringing Disney's copyright on a massive scale by copying a large corpus of Disney's copyrighted works without authorization to train and develop gen AI models and services and by using AI models and services to commercially exploit and distribute copies of its protected works to consumers in violation of Disney's copyrights. So, the letter continued, Google operates as a virtual vending machine capable of reproducing, rendering and distributing copies of Disney's valuable library of copyrighted characters and other works on massive scale. And compounding Google's blatant infringement, many of the infringing images generated by Google's AI services are branded with a Google's Gemini logo, falsely implying that Google's exploitation of Disney's IP is authorized and endorsed by Disney. So the question so Disney's not just allowing OpenAI to generate these assets, but they've actually invested. And so the immediate question is, will Disney do any of these licensing? IP licensing with other platforms it doesn't this kind of lot. Maybe it feels like we're in the. Press release economy all over again. Yeah. I don't know. It does feel like Disney's sort of like picking a winner here in AI video and they're saying like we like OpenAI and we don't like Google, which is sort of a, sort of a bold, bold move. You would think that they would be a little bit more platform agnostic. I'm fascinated by this. Then again, Thrive Capital Bet. Bet the fund in some ways on OpenAI. Who's co owner of Thrive. Bob Iger. No way. Remember, he bought. He was part of a small cohort that bought a few points of Thrive. Get out the red string. It's all connected, of course. Get out the board. Get out the board. You were right, by the way, on the structure. So Disney's putting a $1 billion investment in, but they're also getting warrants to buy more stock in OpenAI at its current 500 billion dollar valuation. So there's a little bit of opportunity in the future if the valuation goes up, that they'll get a good deal there. I really can't get over this Disney plus news. I really. Okay. It does say a curated selection, so I think that I would be very upset about the idea that I would. Open just a random selection of whatever. You know, where I'm going with. I think that Wall E is A is a film that is like. I think it's art. I think it's art and I think it's okay. You don't just watch for my children to watch. Wall E. You studied wall. Erm. But I do, I do. You took notes. I do think that a lot of the Pixar films and a lot of the Disney films, they have an opinion, they have craft, they have, they have something that elevates the human experience. The human. It is not purely brain rotting, it's not slop. And if you were in a funnel where you watch something that's a great, interesting film with an opinion, with something to say, and then it's like, do you want to continue watching 5, 4, 3, 2, 1. And then it's just blasting you with the craziest sora mashups possible, you're going to make a lot of parents very upset. So this better be in another tab. But I do like that they said it's, it's curated because as long as there's some sort of human in the loop to say okay, they clearly jailbroke on that one because, you know, people are gonna get crazy with this. Yeah. So are people doing things. Every Disney character doing the little yachty. Walkout, you know, okay, that's actually okay for the kids. I fully support that. Coffin is a banger and I'm good with that. But no, I mean, YouTube kids famously went through a really, really dark period with Spider man and Elsa mashups and that were very suggestive and bizarre. And it's been a game of Whack a Mole for a long time. And I think people don't expect when they open up Disney, they don't expect it to be a game of Whack a Mole. They expect it to be curated. And so Disney really needs to continue to instill this idea that they are curated. Yeah. So Disney has sent cease and desist letters to Meta Character AI as well as against Midjourney and some other companies. I wonder, have they ever sent a cease and desist to who? OpenAI. Oh, I don't know, but they should have because early as soon as ChatGPT launched, I was having it create stories of Spider Man. I was particularly a fan of the deliberately IP infringing. So I'd be like, write a bedtime story for my son where Spider man teams up with Superman. Which of course cannot happen because those are rival pieces of intellectual property, some owned by Warner Brothers, some owned by Disney. And so you cannot ever have those cross into the same multiverse. According to Google Gemini, which Disney has now sent this cease and desist to, Disney has never taken any type of legal action against OpenAI. Well, they picked, they picked their, they picked, they picked a side. Well, fortunately, we have someone who can add a lot more context here. We have Dylan Byers from Puck News joining us in person in just a few minutes, whenever he's ready. In the meantime, let me tell you about Gemini 3 Pro, our sponsor. Google's most intelligent model yet. State of the art reasoning, next level and deep multimodal understanding. Let's pull up this clip. Bob Iger and Sam, we're on CNBC this morning. We can pull it up. This is 12 minute video. Can we play some of this and see if there are key moments there? And then we will bring in Dylan. Yeah. So the current deal is a three year license with exclusivity for the first year. So again, it's very possible that a year from today, Google ends up with the same ability to leverage Disney ip. Disney will set and evolve the guardrails for how its 200 characters will be used. One year exclusivity doesn't seem like that big of a deal to me because VO3 and Nanobanana are barely rolled out into YouTube. Like, you can put them in shorts, but it's kind of a creative tool. Like YouTube. Does not seem like we got to win the AI vertical video game this year. This feels more detrimental to Facebook and meta than OpenAI or than YouTube, honestly. Anyway, let's table this video. We can come back to the clips. I'm sure it will be clipped in more precise detail. Instead, let me tell you about adeo, the AI native CRM. Adeo builds scales and grows your company to the next level. And let's bring in Dylan Byers from Puck News. Dylan, thank you so much for taking the time to come down to the Ultradome. Good to see you in person. Welcome, welcome. Please have a seat. Thank you. And if you could kick us off with a little bit of an introduction on yourself, however you describe yourself these days, whatever's quickest. I'm the senior correspondent for Puck, where I cover the media industry. Yeah, media, business, media, gossip. How is covering the media different than covering the oil and gas industry or tech or anything else? More navel gazing, more ancestral. Because you're.

NEWSMAX