SpaceX Surges, Anthropic Stumbles, and Bezos Bets $12B

Dave: How did SpaceX first week in public markets go?

Aman: I think it went really well. This is a little bit of the consequences of how Dario has been talking about AI. So far, he's launched a technology that he himself says could be a cyber weapon that will overrun cybersecurity and traditional protocols on the Internet.

Dave: I think we're going to see pressure on, um, both cost structure and sort of productivity. We're definitely going to see limits on, um, how much energy can be produced, acquired or used to build out, you know, these data center platforms. Happy Father's Day.

Aman: And happy Father's Day to you too, Dave. It reminds me, I got to call my dad today. Shoot. Hold on, let me take that note.

Dave: There we go. Yeah.

Aman: All right. Happy Father's Day. You have any special plans?

Dave: Uh, my daughter is kidnapping me to go down to Santa Cruz. I'm not sure what's happening, but it's also the summer solstice, longest day of the year, so enjoy the sunshine while you can. All right, here's what we're going to cover on the show today. We're going to talk a little bit about the global macro story. What's happening in Iran, uh, what's happening with oil. We'll cover SpaceX first week as a public company and how that. Yeah, yeah. Some big news with Anthropic. Um, actually, several news, uh, stories, I guess. Government is laying on some export controls over concerns about cybersecurity issues. And Mythos and Fable. And we'll talk about Jeff Bezos's new startup, Prometheus. This startup raised at a 40 billion valuation recently. That's kind of crazy. $12 billion series B at a 41 billion valuation. Uh, European startup, French company Mistral, uh, just recently raised at a $20 billion euro or $23 billion US valuation. Salesforce is buying Fin, formerly Intercom. Actually, that's a 500 startups company for 3.6 billion. Uh, great day for Owen McCabe and his co founders. And then I guess talking our own book a little bit. Saranic was, uh, in the news for picking up some American soldiers who had been downed by Iran. Illustrated Hormuz. So Saronic Products actually working in the field. Well, let's, uh, start at the top here with the global macro picture and in particular the Iran war. There was a truce and deal that was announced, but then some fighting in Lebanon between Israel and Hezbollah kind of took that off track, I guess. The latest news in the last 12 to 24 hours is that the Straits of Hormuz are closed again. Amman they're open. No, wait, closed. No wait, they're open. Uh, but at the moment they appear to be closed. And J.D. vance is going to talk with some folks in Switzerland, I guess, and see if we can get that all straightened out. But.

Aman: Okay, well, that's, that's promising. I guess the good news is the price of oil is, had gone from between 100 and 110, you know, over the last, uh, three to four weeks. It's down in the mid-70s and it's holding around 76 in the futures, uh, market. So again, I guess the market is anticipating, despite some starts and stops and, you know, and fits and starts here, that the, uh, the level of traffic in the streets of Hormuz will be normalizing. So I guess that's, I guess that is probably good news. Although I'm sure there'll be by the time this drops on Tuesday. I'm sure. The noise. The noise.

Dave: Well, if we look at Kalshi for some indications on when traffic in the Straits of Hormuz will return to normal, uh, the market is pricing in sometime in September or October, I guess so, so probably still a few months, uh, to get traffic back to normal, even if things proceed in the right direction, might take a little while to turn on all the oil infrastructure, clear up everything in the straits and allow traffic to return to normal. I guess it's kind of important whether that happens before or after November, right, Aman?

Aman: Yeah, I think so. Well, in November we have the midterms in the U.S. chances are, let's say 95% or so that the Democrats take the House. In which case now, uh, the ability for the government to strike deals becomes less. You kind of need Congress to support anything around the War Powers Resolutions and War Powers Act. So I think if this gets done by November, you know, that's uh, that's obviously Trump's incentive and that's what they're, that's what they're shooting for. But Iran, you know, I'm sure the Iranian IRGC knows our election calendar as well as you and I do, so I'm sure they're also, they're also, they have an incentive to drag it out until after November and then they're dealing with, uh, the term I use here in the US Is a lame duck. I don't know that Trump will ever be a lame duck the way that we think about it, but I think that's certainly the incentive that the IRGC has.

Dave: Feels like he already is a lame duck given his position in the polls and position of the Republican Party even. But I think, you know, what we're really getting at here is oil prices and inflation and interest rates. So what do you, what do you think, do you think that interest rates are going up or down the rest of this year and uh, that a result of Iran or other things going on?

Aman: I guess that is the $64,000 question. I think what we saw from the Fed that did, they did meet last week. Kevin Warsh gave a very good, very thoughtful overview of his, his thinking on how he wants to manage the Fed. I think what he's going to aim for is a less interventionist Fed. He's definitely kind of a sound money guy and I know he was talking about interest rate cuts in advance of being nominated by President Trump and they're being confirmed by the Senate. But if you listen to his tone last week it was a wait and see approach. I think he's going to, he's going to see inflation running hot. I guess in the positive scenario if oil prices stabilize and prices come down a little bit in the second half of the year, we're still talking about inflation being above 2% which the Fed's target. So I think there's a chance that he actually might, I think he's going to hold rates where they are through the end of the year. But if you look at the betting markets, there's actually a chance that the increase rates between now and the end of the year based on those hotter inflation reads and I think a very low and vanishing possibility that there's going to be a rate cut this year.

Dave: Right. So probably flat or slightly up by the end of the year. I guess that's not the end of the world for the IPO market, but not super great. If we kind of want to take a look at uh, our private mag seven maybe top 10 private companies, there's uh, been a few changes.

Aman: Right.

Dave: I guess because SpaceX is now public, they're no longer at the top of our private nag. 7 They're a public company.

Aman: Yeah, they graduated, they've been called up to the majors, Dave.

Dave: They're no longer the show. Right up with the show and cursor. As a result of getting acquired by SpaceX or in the near future they're probably also out of that top 10 list. So I guess if we, if we're looking at companies uh, here Anthropic and OpenAI definitely, you know, approaching a trillion dollar valuations, both of them probably likely to go public in the next 612 months. Uh, I think Anthropic probably before the end of the year. OpenAI, maybe ByteDance, which is a sort of Chinese company, uh, at least some parts of it, um, on there at half a trillion.

Aman: Right.

Dave: And then Stripe and Databricks definitely valued north of 100 billion, I believe. Databricks kind of negotiating a new round that would put them a little bit above Stripe and Revolut. Also discussing, uh, a new round that might be more of a secondary than a primary round that we be above 100 billion. Um, so I guess those six companies are in the, you know, north of $100 billion valuation range. Um, maybe debatable. Who's number seven? I think Andrell probably has been trading in the uh, $100 billion range. Uh, if you look at notice, um, co, um, Andre raised at a $60 billion valuation, uh, a month ago, two months ago, but already trading north of 100 billion or at least close to 100 billion. So, um, arguably in the Centicorn list or valuations, if we uh, sort of wanted to guess who's next on that list? Deepseek maybe. Uh, although a Chinese company valued at 50 billion ramp in Canva in the 40s. Anybody else on your horizon? Aman?

Aman: Well, let's see. So he said ByteDance, they're said they're kind of a Chinese company, I guess they are nominally headquartered in Beijing, but 60% of them are owned by global institutional investors. That's not the case for Deep Seek. So I guess Deep Seek is maybe questionable, but if I had to pick another one, I don't know. So Canvas had 4 billion in ARR and counting. They're growing at 40% year over year. I think we've seen a bit of a SaaS rebound in the past month or two. We're certainly seeing these top growing AI enabled SaaS companies, you know, the, the cloud flares and the, the Palo Alto networks and companies like that trading at 15 times revenues or 15 times ARR. I don't think Canva is 100 billion, but they could be climbing north of 60. That might be a candidate. I think that would be the one that I would say might be on the edge of that top seven list.

Dave: Now if we look at some of the companies on, uh, notices, uh, top 20, who in here would be kind of on your list? That's north of 50 billion, maybe trying to crack the $100 billion list that we haven't talked about. I don't know. Shine, which is also a Chinese company, makes that list. Maybe Neuralink, maybe. Maybe ripple.

Aman: Well, Prometheus is Brand new to the list. I guess that's still a little, uh, They've gotten great investors. Neuralink's been around for a while, has great investors, but maybe maybe a little bit more, uh, bit more of m. An emerging company. You know, the one that's not on this list but could also be interesting is, uh, Perplexity. I think given the demand we're seeing for AI companies and the fact that SpaceX, uh, is largely an AI story, I guess wait and see how anthropic and OpenAI goes. No, we'll talk about Anthropic in a second. Perplexity could be the next one to jump onto the list and then jump right into the public markets. If I were them, I'd be running for the public markets as fast as my little legs could carry me. But even if they can pull off, if they can pull off a capital raise, uh, excellent company, fantastic CEO, very interesting technology. That might be. They might be in an interesting position to kind of come from behind and sneak onto the list here. Somebody's awfully quiet back there.

Dave: I'm not going to call him dad.

Aman: Brennan, um, you're 39 years old. I would not expect you to call him dad. Well, I'm not going to ever.

Dave: Moving on to SpaceX, which is in the news every week, but now as a public company, not as a private company. Uh, aman, how did SpaceX first week in public markets go? And uh, are you bullish or bearish on SpaceX at 2 trillion, 2 1/2 trillion.

Aman: I think it went really well. I thought after day one, remember I gave them an A minus for their execution without doing any price discovery in the traditional sense. They had priced the IPO at $135 a share. That was 1.77 trillion in market cap. Their first tick was at $150 a share. They closed me to 160, so they were north of 2 trillion, around 21 and then. And the only reason I gave them an A minus is because this being a little bit of a tough grader and I thought maybe the first day pop of uh, at 19% was maybe on the low side of what they wanted, but in the second day they were actually up over 200 and they closed at ah, you know, 185 at the end of the week. So I got to give them an A. I think Morgan Stanley has done a great job. They're the dip in the lead banker on the retail side. Goldman Sachs has been the lead investment banker. On the institutional side. They've signed up some great long only players. The price of the stock seems to have gone up and that, that's the perfect pop like to get through 20, 25, 30% pop. Give your initial holders a bit of a head start, bring in the retail investors. I think it's been really positive and I, I don't think I would uh, I don't think I would short them right now because if you, if you're following like the, the progress of the company. They just allowed options trading on Thursday and the on Friday was Juneteenth, it was the US market holiday so there wasn't a lot of um, no activity in the stock market. But now that they allow options trading they may get caught in a position something called a gamma squeeze. A gamma squeeze is when you have a rapid jump in the stock price because of uh, the way that options are trading. It usually happens when call buying drives call prices higher. The stock price goes up, which is what's been happening. And usually when you buy a call option there's a market maker on the other side and if the price goes up then they have to hedge their exposure and they end up becoming buyers as well. So this could be another pretty good week for them overall. I think they're managing the liquidity really well.

Dave: Do you know when um, SpaceX joins the NASDAQ 100 and the FTSE, is that already happening? Are those buys starting?

Aman: So there are a couple of important dates. Um, you should write these down if you're a trader in SpaceX or buying or selling shares or an Options Trader. So June 18 and 19, I guess 18th was before Juneteenth. They joined something called the S and P Total Market Index which is a, uh, it's an FT SC Russell benchmark which exited. They're buying on the 18th. June 26th is 10 days after the IPO. And on that day they will join the um, MSCI tracking benchmark. So the Russell 1000, the Russell 3000, I think they joined both of those indices on that day. There'll be some force buying then. And then I think the big one you're talking about is going to be 15 days post IPO. If my math is right, around the July 4th holiday. Happy 250th anniversary by the way America. It's around July 6th and on July 6th that'll be 15 trading days after the IPO. That's the NASDAQ 100 which officially includes it. And there are a number of popular ETFs like the Invesco QQQ and its Little baby brother, that's like $600 billion in money, um, that's tied into that index. So they become a forced buyer around July 6th. So those are the next three dates where I think you'll see some forced buying. And the potential, now that the options or are, uh, trading, you get the potential for a short squeeze or a gamma squeeze as well. Then they DO earnings on August 6th. And after August 6th, they will release the first 20% of their lockup, potentially 30% of the shares in lockup from select investors. So those are the major dates to be following if you're a SpaceX shareholder.

Dave: So here's a visualization. I think this is from, uh, Seeking Alpha. But, um, when are shares coming out of lockup? When are they eligible for transfer or purchase? And we sort of see over the next 180 days is where most of that's happening. But as you mentioned, in the next, uh, 30 days or so, um, probably 20 to 30% are available after Q2 earnings. And then I guess whether there's uh, that extra 10%, uh, depending on if the stock is at least 30% above the IPO price for five of those 10 traded years.

Aman: There's the five of the 10 days before the earnings call. So if the earnings call is August 6th, they'll set the, uh, the timeline of the. Oh, is it ending on the, on the QQ earnings announcement date.

Dave: Right, I see.

Aman: Yep.

Dave: Got it. Count of three. Name your favorite dinosaur. Don't even think about it, just name it. Ready? 1, 2, 3.

Aman: Velociraptor.

Dave: If you were a chick, who's the

Aman: one guy you would sleep with? John Samos. What, did we just become best friends? Yep.

Dave: Our next story, that's a big story, actually, several stories is going to be Anthropic. Um, right. Amun. Uh, a little bit of a bump in the road for Anthropic. Uh, what happened recently.

Aman: Here's the bottom line. So I think the basic short, the short story here is that there has been a confrontation between Anthropic and their CEO, Dario Amade, and the US Government. And the US Government has issued an export restriction that has essentially shut down their latest model and has taken their newest model, which is, uh, it's called Fable, off the market. This is potentially a real problem for Anthropic, but there's a broader timeline or story here we should talk about. Anthropic launched a model. The US Government flagged security risks within that model and told them to cease and desist using an export order. This is an ongoing confrontation between Dario on the one side. Uh, previously Pete Hexseff, the Secretary of War and the administration, and now Howard Lutnick, the Secretary of Commerce, who is issuing the cease and desist. But let me just go back a few weeks and just remind you that the company Anthropic, uh, actually launched Opus 4.6 in February. And what happened when they announced Opus 4.6? Well, it caused a potential a trillion dollar trade off in the, uh, software markets, in particular vertical SaaS companies. And one of the areas that got hit really hard was the cyber security space. And the reason was that what Opus 4.6 does or introduces is uh, just a dramatic leap in autonomous agentic planning. The ability to break complex multi day engineering goals into independent subtasks and then spin up parallel sub agents and debug massive code bases. And this is a nightmare for cybersecurity companies. So the market thought that this could be very damaging to the cybersecurity infrastructure that protects the Internet. Okay, if I, if I then go to what Anthropic said when they, when they launched this, this model, they declared and they said this is a, basically a cyber weapon. They've built a cyber weapon of unprecedented proportions that the industry took seriously and that you can see the stock market took seriously.

Dave: They were proactively like not releasing some of those features.

Aman: That's right.

Dave: Market. Right. They had a selective release to I guess a, uh, limited number of financial and CyberSecurity, uh, all US based. That's right before they were releasing it to the rest of the population at that point.

Aman: Right.

Dave: Anthropic and the US Government were kind of on the same page, I think, I think so.

Aman: I think so. Because Dario had said so. They said, hey, we're now launching Mythos. And hey, Mythos is so powerful that it could potentially damage cyber networks. And so we're going to watch out.

Dave: Don't try this at home, folks.

Aman: We're experts, uh, we're the experts, which actually I appreciate. And they announced a thing called Project Glass Glasswing. I don't know who comes up with these names, but they have Mythos, which is the new model. Project glasswing is basically their defensive cybersecurity initiative. It's just, it's an elite coalition between Anthropic, major tech companies, cybersecurity firms. And they were like, we're going to go real slow and release it to our partners. And that, you know, it sounds like the responsible thing to do. And in their press release they said that they're they're worried about the implications, and so they want to make this, uh, as responsible as possible. So I think at that point, there was no issue with the government. Okay, so now what happened? Well, let me just go back to what David Sacks tweeted about a week ago, and here's. Here's his distillation of what happened, and I think it's a pretty clear summary of what happened. So, first of all, they rebranded the name again. I don't know why they do that, but they call it Fable, not Mythos. But Fable is Mythos with guardrails. And they launched it into this project glasswing group of working partners, a highly credible partner of both Anthropic and the US Government. We don't actually know who, but it's reported in Semaphore that, uh, it was Amazon.

Dave: So I think it was Amazon. Yeah, that's pretty. Pretty likely that it was Amazon.

Aman: It was reported in Semaphore, so let's assume that's correct. So Andy Jassy says, hey, we've found a problem with Fable. Uh, it's basically they came forward and said, there's been a jailbreak of those guardrails. I'm sure Andy Jassy then went to Dario and said, hey, buddy, we found a problem. We're part of Project Last Wing. And Amazon, by the way, is the biggest shareholder for the own 18% of anthropic is their single biggest shareholder. They're the biggest singular, single cloud provider in the world. And so they have a vested interest in making Anthropic succeed. But at any rate, he went back and he said to Dario, hey, I think there's a problem here. Dario said, thanks for your feedback. Um, we're going to continue to roll it out anyway. At which point, I think Andy Jassy goes to the government like a whistleblower and says, there's a problem with this technology. We think it could actually jeopardize cybersecurity infrastructure. And, um, Dario's not taking it seriously. And so now the government went to Dario and said, hey, can you take this off the market until we resolve the issue? And at that point, I think Dario said, thanks again, but no thanks. And, uh. Which is kind of a weird response, right? But anyway, then the government issues the order. And so I feel like the problem here is just a communications problem. Like, I don't know why this became a government issue and why Anthropic, given their history of just being AI safe, wouldn't have taken it off the market voluntarily. But I think that's kind of what happened. And, uh, I would just. My only warning to Anthropic would be you've launched a great product, you've gone around saying it's a cyber weapon, that it can actually jeopardize the U.S. infrastructure. Um, when your top partner Amazon, warns you and the government asks politely, like, you should take that stuff seriously. And I don't think that they really did and that caused the order from the government. So that's a bit of a comms problem more than anything else, I think.

Dave: And I think maybe they created a monster they couldn't quite control. Because the response finally was the government saying, yeah, maybe you're right, maybe we should be a little bit more careful, cautious about this. And as a result, we might want to limit AI capabilities, use outside top US uh, enabled providers. Now, it's interesting that Amazon, Andy Jassy would have been the ones calling, you know, blowing the whistle on this because it actually would have a negative impact on their ability to offer their services to non US companies, I guess. Right. Kind of a weird and interesting sort of little development there, you know, certainly on, um, the overall impact of that U.S. export ban, you know, that, that would maybe create some concerns for non U.S. companies, um, you know, Canada, Europe and you know, other governments even about, hey, can we get access to the top AI, you know, tools and services, whether that's anthropic or even anybody else. And then if we even want to get a little further into this, you know, there's been recent conversations from, you know, Treasury Secretary Scott Besant and, and Commerce Secretary Howard Lutnick about nationalizing AI. I think there is some legislation proposed by Bernie Sanders that 50% of AI revenue be held back for, uh, nationalizing the revenues. And I don't know whether that's the right number or not, but the fact that sort of seeing leftist sort of politicians and right wing politicians agree that's the worst, some level of AI, nationalization, privatization, ownership. What do you think about all that, Oman?

Aman: This is, this is a little bit of the consequences of how Dario has been talking about AI so far. So let's just go back to. He's launched a technology that he himself says could be a cyber weapon that'll overrun cybersecurity and traditional protocols on the Internet. Number one, he's also been the one you said, um, a year ago, right? He said there's going to be, he was predicting 50% job loss of white collar workers in the next one to five years. I don't believe that for One second, by the way. But these are his words, not mine.

Dave: He's talking, well, I don't know about 50%, but I think we are going to see job loss as a result of AI. Whether that's happening yet or whether that's just a news cover story for companies to make cutbacks for productivity, I don't know. But I do think we're going to see some long term employment impact as a result of AI.

Aman: Yeah, employment impact. I totally agree. The reduction, uh, of jobs as opposed to the creation of jobs. That's where we disagree. But if you're the CEO of the company and you're leading the charge and this technology, you've got the market share leader and you're talking about job losses, you're talking about wiping out security protocols you've put, you know, it's obviously the regulators now, left wing and right wing are going to be concerned. Let me and you hit it spot on. This is the way he's been talking for a couple of years and warning the market. Let me take you back to something else he wrote called Machines of Loving Grace. It's a long blog post he did back in October of 2024.

Dave: Yeah.

Aman: And uh, again, his word.

Dave: Bill Gurley was talking about this, right? Bill Gurley was talking about this on the all in podcast.

Aman: He did. And they made a, they made the same point, which was, uh, this has been an ongoing dialogue, whether it's fundraising, whether it's marketing, whether it's, I don't know what. But this is, this is what, this is what Dario says in his blog post, again, his words, not mine. He speculates about what the world needs and the role of AI. And he says the world could need universal basic income from everyone. Okay, um, bad idea. But that's, you know, that's just the start of it. He talks about a capitalist economy of AI systems that basically redistributes resources to humans based on what the AI thinks we need. Like this is his vision of AI. And when you talk about massive job losses, when you talk about cybersecurity run amok. When you talk about, um, you know, redistributing wealth among people based on what computers tell us to do, like there some kind of a benevolent entity. I think he should put regulators in a position where now they are reacting to everything that he said. And when Andy Jassy calls up and says this technology has a jailbreak, and then Dario says, oh, but it wasn't serious or we're on it, I think what he's done by talking to regulators this way is he's causing them to, uh, overreact on that Bill Gurley point, by the way. Bill on that pod also said he had been baffled by Dario. And why is it that a technology leader is being so negative about the technology?

Dave: I think his comment was, I've never heard anybody who's, you know, the leading voice of a particular type of new technology wave also being the biggest critic of that new technology wave. Uh, and I kind of agree with him. I can't think of any time in the past where someone who's been leading the charge has also been critical. But, you know, Bill also put together a really interesting talk on regulatory, uh, capture at, uh, last year's, uh, All In Conference.

Aman: Uh, highly recommend that to everybody. But let me give you an example of what has happened. So let me educate you and Bill from, uh, a lesson that's actually. It's in my upcoming book, A, uh, Brief History of Financial Bubbles. And I go through some of what's happened. There is a precedent, and it's Thomas Edison, who was the founder of General Electric, and he was in a battle with George Westinghouse and Nikola Tesla, and we now call it the War of the Current. So it's basically the direct current group under Edison and ge, and the alternating current group, which was Tesla. And Tesla was a great inventor, but not a great businessman. And he ended up having to merge with Westinghouse. And this is the rivalry that played out in 1890 and 92. Edison did what Dario is doing. He basically scared the public. He lobbied the government for a limit on voltage, which would benefit his business because direct current only operates at 10 to 20 volts and alternate current needs 1,000 to 3,000 volts. He actually, uh, launched a safety pamphlet, talking just like Dario did about the potential for electricity to, uh, kill people. And what do we do to business and what do we do to safety? Um, there was a whole lobby around the electric chair. The electric chair was invented in, I think, 1890. And once people realized that electricity could kill you, he used that as a propaganda tool. He actually hired a guy named Harold Pitney Brown to electrocute animals. And he did that as a way to prove that electricity is dangerous and therefore should be regulated. And, uh, he figured that regulation would help him and hurt his biggest competitor, which was Westinghouse. Well, if you had invested in Thomas Edison and ge, I'll tell you how it played out. You know who won the War of the Currents? It was Westinghouse. It was Tesla. It essentially eliminated ge And Edison from the War of the Currents. Edison ends up getting written out of, um, the story of ge. He ends up going on to, you know, in the movie business and stuff like that. Ge, 100 years later becomes the most valuable company and a Fortune 5 company, I think. But they lose the War of the Currents and the regulation that they, that they engendered did not actually help them. So it is a dangerous game that they're playing that Anthropic is playing. I think the danger of it is, um, you maybe get the regulators on your side for a while, but once you invite them into this process, now the talk is about taxes and nationalization and robust security and all the stuff you're talking about. And uh, that's just a function of regulators being scared and being reactive. And that's probably not a great thing for the AI company.

Dave: Do you think, do you think he's doing this because he's playing some Machiavellian game, or do you think he's doing it because he believes it? I, I actually think he believes it. It's not some, um, game he's trying to play with, you know, regulatory capture, although that may be the outcome. I, I think for Sam Altman that is a game that Sav was really playing. Sam, Sam is probably more Machiavellian, I think, uh, in his, um.

Aman: I don't know what that is.

Dave: Maybe the same is that now we're starting to encourage government to get more involved in these companies and again that's happening on the, both the right and the left. And I think I probably share your concerns that when both, there's bipartisan agreement on legislation that makes it more likely that it will happen. How do you think this affects, uh, Anthropic's potential for an ipo? And you know, does this create headwinds for the ipo? Does it create headwinds for their revenue growth?

Aman: It's a great question. I, I would have said before last week the chances of Anthropic going public was 90% between now and the end of the year. And the SpaceX has gone well, so should have gone higher than 90%. I think you got to calibrate that back to like 70%. And, and the reason is just because they have now misstepped at, at a minimum, they've misstepped on communications. And if, if my description of what happened is correct, then I think they've just had a back and forth with the government and their regulators and their top partner, but just shows a certain, you know, lack of maturity. The Worst case is they really have scared the government, the regulators and they've got their top model banned by the government and effectively taken down until further notice. And that's got to have a dent in their revenues. How much? I don't know. And those are all things that are probably negatives now on their uh, on the prospects for revenue predictability. If um, the government goes further and starts scrutinizing their models more heavily. Those are all things that you have to worry about now on the IPO roadshow. So I think it's probably taken the probability down from around 90% to around 70% and I'm really um, looking for them just to step up their game of like communicate. If you're, if you're going to talk about things that politicians care about, you better get good at politics and you better do that before you go public.

Dave: I'm more thinking about like what kind of impact does this have on revenue? And you know, for non US companies is that a concern? I think that know could play into Mistral, uh, having a higher valuation could play into company cohere which we're invested in having a higher valuation. I think also, you know, maybe not quite as widely reported is, you know, there's been a lot of talk about companies that are seeing really high bills for token uh, costs and usage and whether they're seeing enough productivity gains to offset the rise in their token costs. And I wonder if we're going to see some pullback in the growth of revenue for anthropic in this past three months and maybe the next three months that might have material impact on their ipo. I don't think all this government stuff is going to stop them from going public. I think that'll still happen whether it has an impact on pricing or valuation. Maybe, but definitely if anthropic revenue growth slows down, that'll have an impact on valuation.

Aman: I think having your top model taken down by the government until further notice has to have some impact on revenue. To your point, they may be strong enough anyway and it may be fine, but that would be a concerning headwind. I don't see any slowdown in demand for tokens. I think so far what we're seeing is uh, the value that people are getting out of it. There seems to be a very strong demand from it from the hyperscalers and other companies and I'm sure that'll begin to moderate over time.

Dave: But don't you think there's also risk coming from Chinese uh, open source models being a lot less expensive? And their capabilities approaching, you know, the top models capabilities. I, I do wonder whether you're going to face, you know, pricing pressure. Both, both from a desire for companies to reduce their, you know, rising costs for token usage. Unless there's, unless you're offsetting, you know, productivity gains or revenue gains, but also just, you know, 80, 90% as good alternatives emerging at 1 10th or less the cost might also put pressure on revenue.

Aman: Oh yeah, I'm, I'm sure that'll happen over time. I think there's a lot of value accruing to the frontier. And as long as you're a frontier model, I still think those companies will be fine. So I haven't uh, we haven't yet seen or I haven't yet seen, you know, evidence to the contrary, but intuitively you're right. Over the next, you know, six months to a year there's got to be, there's got to be some price pressure and I think there'll be demand for different models for different things and hopefully as a cfo, people will put pressure on token prices. But you know, that hasn't happened yet.

Dave: I think we're going to see pressure on um, both cost structure and sort of productivity. I think we're definitely going to see limits on how much energy can be acquired or used to build out these data center platforms. So if projections for data center growth are limited, there's going to be more attention on productivity and efficiency and I think that plays into less expensive open source frontier models being more attractive for companies to use, at least for basic tasks. Maybe not for know, the top end set of tasks.

Aman: Absolutely, absolutely. And I don't think we disagree or saying different things. I just think there's also value to be had at the edge of the frontier and that'll continue to drive the, the valuations of uh, anthropic and that's their, that's their IPO story. So I think that that part of it still feels to me to be pretty good. Mr. Huff, Mr. Jeaner is ready for your interview.

Dave: Actually we'll be interviewing as a team.

Aman: We're here to it up.

Dave: Well, there's been a new entrant in the latest uh, set of decacorns that have been on the radar. And one of those is a pretty big uh, new company but from uh, a familiar face. Prometheus is Jeff Bezos's AI startup and they had previously raised $6 billion at a pretty high valuation. Now they just raised another $12 billion at a $41 billion valuation. That's pretty crazy for a company that's maybe only a year old. Aman, what do you think about this new company and what do they, what did they actually do? I guess physical AI is part of their storyline, right?

Aman: Yeah. Is that like the, is that the biggest Series A of all time? Is that a Series A or a Series B?

Dave: Is that uh, that's a Series B but that's still a pretty big Series B.

Aman: It's gotta be the biggest ever. So they're a startup but they're 40 billion. I mean Jeff Bezos is a pretty amazing entrepreneur and once in a lifetime entrepreneur or maybe two in a lifetime we're benefiting from Bezos and Elon Musk. Um, the other co CEO is uh, Vic Bajaj who uh, you can't really trust him because he's a Canadian. Canadian. American. So let's. Aside from that, aside from that negative. I think they're in the physical AI infrastructure world. So I think they are now they're competing in hard manufacturing. These are things that Jeff Bezos knows. So it's about logistics and hard manufacturing and the world of atoms but using that to enable AI heavily, heavily software integrated with the hardware processes. But I think they're going to be a legitimate competitor in the space of things like hard manufacturing, maybe drugs or other R and D science where this stuff requires a physical manifestation. It's a really, I don't frankly know what exactly physical AI specifically means beyond that. I think that part of the business plan is still a little bit loose but I'm sure that they're going to be successful because of the people behind it, the money they've raised.

Dave: Well I think there's more attention being given to hardware based companies that are uh, including AI whether that's robotics companies, factory automation, uh, other physical products that are incorporating AI the same way that software is incorporating AI. The uh, sound bite of some kind of artificial general engineer or the sound bite of physical AI is maybe a little fuzzy and not well understood but I think we're talking about physical production and atoms having the same productivity gains because of AI the way that software companies are having productivity gains the way of AI. And there's a pretty substantial interview with both Jeff Bezos and you know this new co CEO Vic Bajaj on uh CNBC recently where they talk at more, more length about what they're trying to accomplish. Aman, if we look at like Vic Bahraj background uh, he's been at Google for quite a bit of a time. He's been involved in a Lot of life sciences related projects. Uh, I guess he's an adjunct professor at Stanford and has a degree in physical chemistry but also experience in life sciences. Our next story is going to be about a French AI company, Mistral, which is raising at a new valuation of 20 billion euros, maybe $23 billion US. Uh, Aman, any thoughts about Mistral? This, this company hasn't been in the news as much as OpenAI or anthropic or even XAI, but they're uh, substantially increasing revenue and valuation.

Aman: You know we covered them in a valuation corner a couple of months ago when I was on vacation in France and decided to do a bit of a deep dive in the French ecosystem. At the time they were, you know, a, a uh, second tier AI LLM player but had the benefit of being European. And the reason, the reason why that's important is because you're, if you're European, you're not Chinese and you're not American and in a world where uh, the US is putting export, export controls on anthropic and um, you don't necessarily know what you're getting with some of these Chinese open source models in terms of the security and the involvement of the ccp. Like just having an independent, a sovereign independent AI model could be interesting. Right. And that was a little bit of our thesis with cohere, the Canadian AI player as well.

Dave: And Mistral is effectively the Dr. Pepper of AI, uh, frontier platforms.

Aman: And uh, in the case of Mistral, you know, French and technology don't necessarily go well together in terms of a sentence, but they have been able to develop a, uh.

Dave: There are nerds in Paris.

Aman: There are you. That's. And Mr. Al Pro proved it. Their annualized revenue run rate has gone from. It was just 20 million, I guess 20 million euros a year. A year ago. They're 400 million euros right now. Their CEO Arthur Mensch was just at Davos and said they're on track for 1 billion euros in ARR by the end of 2026. So you know, at a 20 billion euro valuation on 400 billion of ARR, I guess that's a 50 times forward multiple, which sounds even in a normal SaaS market that sounds insane, but if they do get to 1 to $1.2 billion in ARR, we're talking 20 times revenues. And in recent months this is concern in Europe, I guess has intensified that the US foreign policy is leading. It uh, could maybe force a tech decoupling or at least just value and independence. I think Mistral is finding itself In a sweet spot. Did you know, by the way, the UA that the EU today relies on, um, overseas providers like mostly Americans for 80% of their digital services infrastructure. So if you're in Europe, you've got to be thinking, how do I get some independence at least another, you know, another um, another partner. Not an American, not a Chinese player. It's Cohere, it's Mistral. That might, that might be the shortlist right now of these sovereign independent AI platforms. And the company's done a great job raising money. They were Lightspeed partners who's been their investor right from the get go. They were true believers right from the seed round In June of 2023, E16Z led the A general capitalist and Index were in the subsequent round. And now they're talking about the semiconductor, um, giant asml. They're a Dutch monopoly. That, that is a uh, huge semiconductor company. They're jumping in now as strategic. Uh, and I think we'll, I think we'll be the largest shareholder for Mistral after the Series c. They're over $1 billion, 1 billion euros in investment, so about a 10% or 11% stake. So I think they've done a really good job of executing and are in an interesting place right now.

Dave: Hey, I never asked you. Yeah. Do you like guacamole? Uh, well, one of our last stories today, fairly uh, significant acquisition and this is a company that's slightly familiar. It's Salesforce, uh, acquiring AI customer service platform Fin for $3.6 billion. Fin, uh, used to be a company called Intercom and actually I was an investor in Intercom, uh, pretty early. Uh, we invested when I was still running 500 startups. I guess they raised a million dollar round back in 2012. Owen McCabe was the founder CEO of that company and they actually were operating out of 500 startups offices for a little bit of time way back when there, um, so pretty nice acquisitions. I guess it's been a long time, almost 15 years now. Pretty big win for us at 500 and certainly a big win for Owen and his team.

Aman: This is also a big win for Ireland. This is Ireland's biggest exit ever. 3. I can't imagine Ireland has a lot more $3 billion acquisitions. So that's uh, that's good.

Dave: Well, I guess it depends on whether you think of Stripe as an Irish company. But uh, Stripe, uh, also some Irish

Aman: founders there, some Irish heritage, some Irish. The Collins are uh, nominally from Ireland. Yeah, I guess, I guess that counts. But you know, I think Owen has done A great job of putting, uh, we follow this company very carefully, obviously as a 500 portfolio company. They a couple years ago they were at 300 million in ARR and were growing at like the mid single digits, 6 5, 6 7% year over year. And they went from 300 to 400 million and accelerated growth to 25%, um, revenue growth pretty amazingly.

Dave: Owen coming back to the company, I guess he had left the company, um, but came back in and they rebuilt a new product. Fin really was an internal AI driven product. You don't often see a company that is a non AI company disrupt itself by building a new AI, uh product. But really they were pretty successful in doing that. Fin became so successful they rebranded the company from Intercom to finish.

Aman: Right, well that's the playbook for, you know, pre AI SaaS companies. I think Owen's laid it out. Now whether you can execute on it, I don't know. It's obviously awfully hard to do. But this, this to me is like any SaaS company that predates AI. Any board meeting right now has got to be like, how do we take that playbook?

Dave: How do we get AI? How do we get AI into our products?

Aman: Yeah, yeah, exactly.

Dave: Well, our last uh, story for today is again, uh, talking our book on a story. Sironic, uh, had a drone boat that rescued some pilots off the Strait of Hormuz. Uh, unfortunately for those guys, they were shot down by Iranian drones. Uh, but uh, they were picked up by American drones.

Aman: It was an AH64 Apache helicopter, went down in the Gulf of Oman. So right near the Straits of Hormuz. We don't really know the circumstances under which it was shot down. Let's just say it was, uh, what do they say? Contested circumstances is the term. But uh, yeah, they ended up in the water. And the U.S. navy dispatched a Saronic Corsair, which is the autonomous service vessel. It's a 20, uh, four foot unmanned surface boat. And they located the two downed pilots and pulled them, pulled them out of the water. And uh, I think this is, you know, drone warfare has reshaped the region in terms of the warfare, but I think now drones are reshaping how we think about search and rescue. Lucky for those, uh, those two downed soldiers and you know, great win for Saran has gone from military contracts to operating in the Gulf within about 18 months and building this technology. And that's a nice story for Sinonic and it's a good story for uh, the US as well.

Dave: Aman, um, what are we covering this week?

Aman: Well, this week we're going to do another installment of our retail venture capital fund series. We're going to talk about Power Law, which just listed on the NASDAQ on May 27th. This is a fund, they basically buy interests in late stage private companies, but they are publicly traded. So it's a way for retail public investors to get in on the action that you and I are a part of. This one is backed by Acadian, Mike Dinsdale is the CEO and Ben Black is the chief, uh, Investment Officer. So since their ipo, they've actually traded down a fair, a fair amount. So if you want to buy them on the nasdaq, you can, the ticker is pwrl and uh, as a retail investor you can go and buy them. So I'll give you just a reminder on how we've thought about these retail VCs in the past. We've covered three before. We covered Robinhood Ventures, which trades under RVI. We covered fundrise, which is VCX, and then the Destiny Tech 100, uh, which is also a publicly traded vehicle. And what you said before was you got to look at really a couple of things. One, what does the portfolio own, what are the underlying assets and how do we think about the marks? And then they all trade at a price to nav, typically at a premium. So it's a north of, north of 1x price to NAV. That's important because whatever they own has an underlying net asset value based on mark to markets. But what you get when you buy in the public markets, that's necessarily tied to the, to the book value or the NAV value. Sometimes it can be paying up to five times what the NAV value suggests. So in the case of fundrise, for instance, this is just an updated price to NAV ratio for all these different funds. Uh, fundrise is trading right now at five times what the NAV is. So you don't really want to be buying something that's 5x priced unless you are really sure that the underlying investments have a valuation that's severely undermarked. In the case of the Destiny Tech 100, we said there's a lot of cash on the book and so paying a premium to that doesn't really make sense. Fundrise, they have a couple of great assets, notably Anthropic, which I think is significantly undermarked because they seem to be now in or closing around. That's way above their last mark. But 5x is a steep premium to pay. And then robinhood was almost 2x. Well, here's how those three have traded. So we covered them in uh, May and you can see all of them are kind of down or flat. In fact the VCX is down quite a bit. There's been, there was a lot of excitement in the market after they went public, but they're down about 30 to 40% from their peak. And the others are um, slightly down or you know, modestly down. VCX is basically the big loser but all of them are treading water even though there's a lot of public market excitement. So similarly, let's go into this one. Power Law. Dave. This might be the one that I love the most. Scarecrow. Okay, the category, but here's why. So Acadian is backing this. Um, didn't Stale in Black have put together a portfolio that you can see is heavily in two big names? SpaceX about 20% of the portfolio. OpenAI about 8% of the portfolio. So 28% of the portfolio roughly is these two really big names. You could argue those are fully priced. I wouldn't say overvalued, but I think there's some risks given where OpenAI, ah, maybe that gets marked down. SpaceX I think is fine, but it's going to be volatile and they just went public so you're not getting a deal on SpaceX. Arguably even that one's a little bit overvalued. However the other portfolio companies are Calci, Deal Stripe. There's companies in there. They've got um, a very um, uh, solid long tail of companies which includes uh, Canva and Merkor, um, Databricks includes Perplexity and Waymo and the whole thing trades about 1.1x price to nav. Kind of reasonable. And SpaceX was in the portfolio and traded up on the IPO. And even uh, even though SpaceX went up, power Law went down. So there was actually uh, a narrowing of the price to NAV in the portfolio. So I guess the question really is do you like the portfolio? Well one, do you like the team and their management track record? Two, do you like the portfolio? And then I think on the pricing I think it's not a bad price. So the pros I would say are they've put together stakes in some high profile private companies including um, some with pervasive AI exposure as AI continues to roll out through the markets. I think that positions Power Law to benefit from AI driven growth. Um, and I kind of like the fact you're paying a premium nav of only 1.1x. On the downside you would say well I don't like paying any premium to nav. Um, I don't like the valuation risk because SpaceX and OpenAI are maybe a little uh, you know, questionable in my mind. They don't have anthropic exposure and maybe that's the one that everyone wants to get in on. And then the last thing just to remember with all of these things is sometimes there's a post IPO decline. So if you buy these hot companies right before their IPO and uh, and are holding um, you know very often you get a uh, an underperformance. This chart goes back to you know 19, uh, 80. So this is a, a 40, uh, five year chart and you can see what happens to IPOs relative to the market in uh, in these trading periods they have underperformed the market by about 20%. Um so if, if all you are doing is buying on the IPO when you were holding you often end up with a, a uh, an underperformance. And the risk there is that you're buying companies at the top of the market, they go public and then they, in the next six months or so they begin to trade down. And um, and that's okay if you get a discount the stocks or you get them before they do the last pre IPO round. But if you're buying at the top of the market and you're paying a premium, the nav and then you get marked down by another 20% that's not a recipe for success. So that's the one thing I'd just be concerned about is just be really feel good about the portfolio and the, and the mark. I think these guys are probably in a, in a pretty good position. Last thing I'll just say is I think their fee structure is also a little on the high side. These are the, the major players and their net expense ratio. There's no carry by the way in any of these funds. So that's the good news. But the bad news is you're paying 3 to 4% as an annual expense ratio. Destiny Tech 100 is on the high side of that range. Power Law right behind. And then you may be getting a bit of, a, bit of a deal with VCX or rvi. But I think overall the portfolio to me looks like a pretty good portfolio at reasonable marks. As long as you're okay with SpaceX and OpenAI.

Dave: Well I think the team is also a big plus. I think Ben uh, has been running secondaries for over 10 years. I think closer to 15 years at this point. Uh, he's certainly Experienced in venture capital. Uh, Mike, who is the, you know, the other partner there as well on the operating side. Um, I think, you know, their list of companies is interesting for many reasons. It's not just based on, you know, sort of top visibility companies. It's got a bunch of other companies that maybe, you know, most people wouldn't have heard of before. Um, the other, you know, us uh traded, you know, entity that's not on your list, USVC I think is also maybe interesting for probably other reasons. And that I, I think, you know, based on the angel list, access to a lot of early stage uh, managers and companies. Some of the USVC bet is not just based on pre IPO companies, it's also based on funds. Um, and those funds might have companies that are earlier than the rest of the later stage companies that are in these other indexes might even have, you know, series A, B and C companies, uh, in some of those portfolios. So I think for a couple of reasons I'm really bullish, uh, in power law, you know, Ben Black's you know, experience and ability to pick, you know, interesting later stage secondary companies. I'm also kind of bullish on USVC's you know, different thesis really. Um, you know, those other, other three alternatives, I think I'm a little bit nervous about, you know, where the market is pricing their multiple as you mentioned, you know, whether that's 3, 4 or 5x, you know, price to nav. That just seems a little rich for me.

Aman: Yeah, right. The distinction with USVC is they're not a um, traded fund, they're an open end fund. It's uh, you actually get in at the nav, which is fantastic. Um, the downside is that you're a bit more locked up. You can't do redemptions, uh, you can't sell on the NASDAQ tomorrow. You have to do accordingly. Redemptions, you're limited to 5%. But if that's your strategy and I guess you have to like naval and you know, the angellist portfolio access that he has, if you like those then I think that's a great one. Would you say Ben Black and you know Mike are the. Of the teams that I just showed, is that do you think the premier team or do you think they're based on their experience?

Dave: Probably put Duval in. The list of syndicate managers in angel list is pretty wide up there too. But both then and naval are people that I would bet money on. The rest of the folks, I think certainly the Destiny XYZ folks are experienced, uh, that's some ex forge X equity, folks. The multiple the market is pricing for that nab is maybe my concern there, not the team.

Aman: Yes, exactly. Well, there you go. There you have it. Those are. That's the, uh, that's the breakdown.

Dave: All right. Thanks for another great, uh, valuation corner, Ahmad.

Aman: Thank you, Dave.