Sonnet 5 Drops, Fable 5 Will Return & Fusion’s First Plant Gets Licensed W/ Philip Johnston | #268

It's pretty clear that Sonnet 5 now is a way to kind of fill this gap until Fable 5 is back out. A kind of mediocre capability at a high price point, but people will still need to buy it.

Anthropic's flagship model, Fable 5, has been offline for 15 days because US government pulled it. Now Axios reports it may be back within days.

Historians will look back and say this period in time marked the period towards the middle or the end game of recursive self-improvement.

Helion cleared the required Washington state regulatory approvals for its Orion fusion power plant. It looks like fusion is finally here.

If you were watching the right metric or the right figure of merit over the long term, you could predict when this is going to happen and it's imminent.

All right, mates, let's jump into data centers and space. And for that, we're pleased to bring a friend on.

Thanks so much for having me. It's a huge honor. You know, a long time Bang bang!

Now that's a moonshot, ladies and gentlemen.

Welcome back to Moonshots everyone—your front row seat to the coming singularity and the age of abundance. I'm here with my magnificent moonshot mates Dave Blunden, our managing partner of Lynk Exponential Ventures—the number one funder of MIT and Harvard AI startups; Saleem Ismail, our global trotter. Our CEO of OpenEXO, and of course, AWG, our in-house ASI. I'm Peter Diamandis, your host and your abundance evangelist. Gentlemen, good afternoon, good morning, good evening, wherever you are. So, you know, where's Waldo today, Saleem? Where on the planet are you and where have you been?

We dropped our son off at a camp in San Sebastián in Spain, or near there, and I'm in Biarritz right now for a few days.

Okay. And you were like in Germany before that.

I mean, I've been in 5 countries in the last 3 days. It's really been nuts. Of course. Of course.

And Alex, all well with you?

Yeah, I'm GDP maxing or doing my best. Always be GDP maxing.

I'm happiness maxing. You know, gratitude maxing. And Dave, good to see you, pal.

Hey, I'm at LingQ studio. And just for word for the wise, we have a ton of Northeastern and Princeton activity in addition to MIT and Harvard. These days. That's true. Of course, we had— Tech Trek is killing it out in San Francisco. Liputan, uh, Andrew Feldman, uh, the number 2 guy at NVIDIA, they were all there on Friday talking to the troops. So it's, it's really rolling. Amazing. This is the thing we did with Eric Schmidt and Eric Brynjolfsson and Daniela Rus.

So today we're going to cover a bunch of new stories. We're going to catch up on robotics, energy, data centers. We've got 20 stories across 6 fronts. A lot's happening and a lot of capital is flowing. For those of you joining us for the first time, our mission here at Moonshots is to keep you informed, keep you optimistic about the future that we're creating. All right, let's jump in. I have 3 stories I want to hit on the abundance front.

Wait, wait, I just got to make a quick point to everybody watching. If you've not seen the last episode with Iman, I'm about 3/4 of the way because it's on— I'm trying to keep up with our own episodes when I miss one, and it was ridiculously amazing. Just a comment there.

Yeah, and Emad's made some great releases in the last 48 hours on his latest model combinations. All right. All right. So we're going to open our first story on robotics. But before we do, because I want to talk about the predictions of how many robots we're going to have on planet Earth. I was having a conversation with a dear friend, Ramez Naam, today. Ramez is one of the earliest Singularity University faculty members. And futurist. He's extraordinary. And he shared this chart with me, guys, that I'd love to discuss. It's a look at how experts consistently underestimate exponential growth. So in this chart on the left here, we see new solar, you know, growth in solar. And that, that yellow exponential line is the actual growth in solar. It's been growing at an extraordinary rate. And what we see on these departures that go horizontal are the predictions that the experts make every year showing, you know, linear or, you know, just small incremental amount of growth. And over and over again, they underestimate it. The chart in the center there is the experts' predictions on EV growth. And again, we see exponential growth in EVs. And then the forecasts consistently are underestimating the actual growth. And then finally, we see the same chart going on in, in battery sales. So it's an interesting phenomenon that while we're living in this exponential growth, the experts who are the experts in the way things used to be are not projecting the growth. They're staying very, you know, shall we say, sublinear in their estimates. So, Saleem, you and I have seen this before and we've discussed this. Any thoughts? Oh my God.

Every presentation I ever give has a segment with several slides showing this. The poster child is a story about Moore's Law ending by 2022, which came out in 2013, some experts said. And you can go back and look at the technology press and every 2 years that article is appearing for 60 years, right? Because experts are really good at measuring the technology. They're terrible at the compounding ecosystem around it. There's some really dramatic examples from Ramesh around the energy ones, where the solar is vertical and every expert for 10 years goes horizontal in that thing. It's an endemic problem. We have that whole headline in the original book, Peter, which we put together saying, beware the experts, right? And this is the immune system because when you've got somebody that's got 30 years of experience in something, they'll tell you how not to do something.

Yeah, agreed. Alex, any thoughts on this?

No, I think the moral of the story is you should always take the logarithm of the actual history before you hand it to experts for their linear extrapolation so you can get the right answer out.

You know, I define an expert as someone— So geeky cute. I define an expert as someone who can tell you exactly how it can't happen, right? And it's so true that experts today are sort of so ingrained in the past because if there's a disruption, if there's a revolution that comes, they're no longer the expert. And so it's just— It's against their best interests. So I want to tie this story to our first, our first robotics story here. And it comes with two predictions. The first was Morgan Stanley. Morgan Stanley had originally predicted 14,000 Chinese robots coming out of China, and they've upped it to 26. Now they've just upped it to 50,000 and projecting 500,000 robots by 2030. But the fact of the matter is there's 140 humanoid robot companies developing hardware in China today. And at the same time, you've got Elon projecting, you know, tens of millions of robots to 50 million robots by 2030 and billions going into the early 2030s. On the flip side, what we're seeing here is on the right-hand side of this is a chart from Andreessen Horowitz that shows we're going to be seeing about $16 billion $1 trillion of hardware investments in Q1 of 2026. And the US is finally catching up. Alex, I know that you're heavily committed to this. And Dave, your thesis is we're moving from an AI-centric entrepreneurial ecosystem to a hardware-centric ecosystem. Would love your thoughts on this. Alex, you first.

Yeah, as I've mentioned on the pod in the past, I think superintelligence is set to spill out of the data centers into the streets. And I think the most obvious vehicle for that is autonomous vehicles on the one hand and humanoid and near-humanoid robots on the other. I think as we start to increase the number of humanoid or just say general purpose robots per capita, there are, there are gonna be certain regimes at, at the low number of humanoid robots per capita regime, it looks like robots performing industrial applications, robots in factories, robots doing logistics. At the— as we start to get, I think, closer to approximately 1 humanoid robot per capita, it looks like domestic robots everywhere. It looks like an I, Robot-style regime where everyone has domestic staff. As we start, interestingly, and a point that I don't think I hear frequently enough, as we start to push well through the approximately 1 humanoid robot per capita regime to 10 or 100 humanoid robots per capita, at that point, I, I think to a hobby horse, I think of Saleem, we start to end up in some pretty exotic futures where there's no longer necessarily justification for the humanoid form. We end up with micro robots and nano robots, and there's sort of a natural sense in which the humanoid form is no longer natural in a world where we have 1,000 general purpose robots per capita and we can start solving all of the grand physical world challenges that would maybe be uneconomical if we only had one humanoid robot per capita. So I think we're going to very rapidly scale through the low per capita regime to the approximately one per capita regime to the many per capita regime. It's going to be a very exciting scale.

So, David, take a second and walk me through your thesis right now, because we've been investing together in AI companies mostly, and, you know, You've said that you expect that is going to sort of fall off and more of an investment to hardware in the next couple of years. Why?

Yeah, I think we have to think in terms of 10-year investment themes. And, you know, 10 years in the age of AI is like 100 years in any normal world. And so I do think the next 1 and 2 years is still dominated by white-collar automation, AI algorithms, chip design that designs chips, the beginnings of data centers and space. But then if you think beyond 2 years in the future, You know, what are the investments that are going to really be big 3, 4, 5, 8, 10 years from now? The automation of construction of data centers is all going to be robotic. Biotech, chemical mixing experiments, reading gels, that's all going to be robotic. And so the machines that do that, you take a guess, you know, how big do you think the U.S. gutter cleaning industry is? People who go on your roof and pick the leaves out of your gutter. How much do we spend per year? I do. I looked it up. How big? That, that use case alone is $1 billion a year. If you built a robot that cleaned gutters and washed windows, those two tasks, that's a $20 billion global market. $20 billion global market. So that's a theme that will support— this slide says 140 humanoid robotics companies. In China. In China. In China. Yeah. There's room for thousands and tens of thousands of robotics companies specialized for various use cases, everything from wafer movement inside a chip fab to chemical mixing in biotech to gutter cleaning to just everything, you know, construction. Construction has thousands of individual tasks. So it's a very broad long-term investment theme. Kitchen work too. We're already starting to invest in that one. But kitchen work automation because the— sure, the fast food restaurants can, can buy at scale and so they'll co-develop with you.

This is where Travis Klatnik is focused with his company, Atom's, just fully roboticized kitchens. You know, what's interesting is this particular note of the $16.5 billion in last quarter was in 500 deals. I used to think that, you know, I'm a huge Star Trek fan. I used to think that Star Wars was kind of silly with all those, you know, hundreds of variations of droids out there. But it looks like it's coming.

Peter, it's such an interesting point. You know, I'd be curious to hear your thoughts on this. Historically, the robots have been missing from Star Trek other than in the first few series, like Soong-type data robots, there were no robots. Maybe it was— what are your thoughts on that?

I think they wanted to create a very humanistic series, and Gene Roddenberry was all about the societal implications of technology in this future. And he— and even Computer, right? Even Computer was the name of the computer. Which was Gene Roddenberry's wife Christine playing that role, was El-Kars, right? Was very, you know, roboticized. It didn't predict this, you know, incredible empathic voices we have on our models today. Yeah, they missed that. Well, you know, we're going to have Rod Roddenberry is going to be with us at the Moonshot Summit. He's going to be on stage, one of the judges of our Future Vision X Prize. So we'll sit down and ask him. I think those are, those are important questions to ask. Why did they miss that part of the future?

Was it just like low production budget? Because they added them in more recent Star Trek series. Yeah. They retconned them.

But still nothing close to Star Wars. And, you know, they're going—

I think a lot of it had to do with the theme, right? Infinite diversity, infinite creativity. And they were focused very much on human interaction. All the aliens that were humanoid, not just because you could put actors in those suits, but the fact that you were always dealing with the interpersonal elements in those, in the plot lines of those.

So yeah, we've got lots of companies that are building robots now. Famously in the US, we've got obviously Tesla with Optimus, we have Figure, we have 1X. And in China, probably the robot company that's got the most sort of publicity has been Unitree. We had one of the founders of Unitree on stage with us at the Abundance Summit. I want to play their latest video. It's got 11 million views and This is a glance at the R1, which is incredibly selling for $4,900. I mean, this is the price of a cheap used car, which is saying a lot. So let's take a look at the R1. So, I mean, crazy. You know, my point is it's an extremely capable robot, but the work here is going to be on the software layers, right? You buy this robot. I don't think it does that out of the box, but I think you could probably buy the algorithms that enable you to do that. But I see an explosion in the number of people experimenting with these robots. At $5,000, that's affordable by, you know, almost anybody who has a reasonable income. Alex, where does this go for you?

Well, I think the elephant in this particular room, as with Elon demonstrating to the world that you could drive the cost of heavy lift to LEO down to effectively near zero through reusable propulsively landing rockets, the elephant in the room here, I think, is humanoid robots that are able to assemble other humanoid robots. Sure. The cost of assembly down to near zero, what we're left is the cost of raw materials and the cost of energy. And that's going to be effectively de minimis. So I think as we drive, and by we in this case, I really mean Chinese organizations because the West is woefully behind at the moment. As we as humanity start to drive the unit cost of general purpose robotic embodiments down to near zero, at some point, We will need to cross the threshold of robots being able to assemble other robots in order to keep driving that cost down. And then at that point, we have physical labor too cheap to meter. And as I pointed out in past, approximately 2/3 of the service economy constitutes some sort of physical labor. And we can do for the physical world what AI agents are right now in the process of doing to knowledge work, which is basically driving the cost of knowledge work and soon physical work down to near zero.

Can I make a point actually that we learned at the, at the Gigafactory that I completely think is— it was lost on me and I think it's critically important— is when Elon says humanoid robots building humanoid robots, it's actually the CNC milling machine or the auto lathe already exists. It's already making the parts. Yes. It's just a file loaded in. What those were designed for is a human to go and take the part out of the machine and put it in the next machine.. And so the humanoid robot does not need to literally create, you know, with a file and a piece of metal, you know, it's the, the automation's already there. It just needs to do the part that the human is doing today, which is moving the part from machine to machine and doing the final assembly. So it's a much easier problem than, than robots making robots sounds like.

Yeah. I think the point I wanna make here is we're heading towards commodity pricing on these, on these things. $4,900, right?. And so as we move to commodity pricing, the question is, where's the value layer?

Well, the value layer clearly will be in the software.

And yeah, exactly the point I want to make. It's—

yeah, I think that's a really important point. And people are going to build apps. Look, if you go back to the original personal computers, you put them out there and people kind of like— we didn't know what they would do with them. And then over time, you had more and more applications built. And now these things show up networked out of the box. There will be profound new skills being emerging all the time from these things. I think it's going to take— I'm still going to say I think it's going to take a lot longer than people I think because the driving, it took us 20 years and that's a very bounded domain space. Humanoid robots doing gardening, et cetera. There's a million edge cases. And I think we'll want more to explore there.

You know what's different here, Celine? What's different here is that, you know, and this is sort of a, what does it mean for the entrepreneur out there? You can buy this and begin to build on top of it. Agreed. This is not something that requires permission from anybody. It's not something that requires a a massive corporate budget. An entrepreneur can buy this for $1,400. It's like the Raspberry Pi moment. And you can start hacking and publishing software to these robots that becomes a new revenue engine. So I think that's what's most interesting for me is the explosion of applications that come on top of the iPhone that now come on top of these robots.

I agree, but I still think we're going to spend a lot more time figuring out the industrial use cases and the dull, dirty, dangerous jobs and getting those automated. Automated. There's so much scope, it's going to take a decade to kind of get through that before we even get to somebody coming over and doing gardening for you.

You know, we, we talked in a past pod about, uh, Royal We. I mentioned the idea that the export controls that we're seeing at the moment imposed on frontier models are, in some sense, you could look at them through the lens of immigration policy as being regulations on importing foreign or exporting, depending on your perspective, superintelligence. But that's the software layer. I do think, so admittedly, maybe a spicier take here, I think that as the cost of robotic embodiments, primarily from China at the moment, starts to come down, I would not be surprised to see similar either import controls or other national security motivated restrictions start to kick in. It's not just about dumping, it's also about taking the embodiments for superintelligence and moving them across borders.

This episode is brought to you by Blitzy, autonomous software development with infinite code context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise-scale codebases with millions of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan, then generates and precompiles code for each task. Blitzy delivers 80% or more of the development work autonomously while providing a guide for the final 20% of human development work required to complete the sprint. Enterprises are achieving a 5x engineering velocity increase when incorporating Blitzy as their pre-IDE development tool, pairing it with their coding copilot of choice to bring an AI-native SDLC into their org. Ready to 5x your engineering velocity? Visit Blitzy.com to schedule a demo and start building with Blitzy today.

Well, you've heard the Trump administration saying they want to invest in the robotics industry. We've seen, you know, obviously Tesla and Brett Adcock from Figure getting massive investments to support the growth of these systems. It becomes strategic for the US and we're going to find out in our next story. Let me just Let me just go there for a second where we're going to start to see robots being used in a number of different areas, including law enforcement. So a drone, of course, is a robot. And I'm going to share a particular video here that, Alex, you shared with me. This is out of Orlando, and this is the U.S. law enforcement beginning to use drones as first responders. Let's take a listen to this video.

And new at noon, the Orlando Police Department is now using drones as first responders, sending them to some 911 calls to give officers a live look at the scene. A new eye in the sky now responding to some of Orlando's most serious calls. Orlando Police Chief Eric Smith addressing a big question from the community.

What would you say to the citizens who either seeing this as an invasion of privacy or an overstep? We're not looking in people's windows. We're not spying on people. We're not just flying around, just fly around.

Qualifying call comes in, a nearby drone can be dispatched to specific GPS coordinates. Then from the crime center, an FAA-certified pilot can control the drone, giving officers a live view of the scene, including if a suspect runs, hides, or possibly has a weapon.

So interestingly enough, um, this was demonstrated on June 17th as a first deployment. It had, uh, what they call 9 docks at different locations and 11 Skydio network drones. Skydio is sort of the U.S. manufacturer today. It, you know, used to be DJI out of China. But we put, you know, import controls on DJI for a number of security reasons. And so Skydio, even though it's more expensive, it's something like 3 to 10 times more expensive than the DJI drones are getting it. Interestingly enough, Salim, Rick Smith, who's one of my Abundance members, he's the CEO of Axon. It's the company that makes the taser and the body cams, has the contract here, and they're, they're coordinating all this. They did a trial with a single drone, and it beat patrol officers about a third of the time to get to the live location and provided useful information in 97% of the time, they claim. But you can imagine that we're going to have drones on buildings throughout the city. And a drone will get there almost immediately. You know, I can imagine there's fear that people have on this subject, but I think being able to get to emergency locations, making sure you can assess situation when you need to have medical personnel there, it's going to save lives. And I want to acknowledge the fear that people have. But I think on the whole, as long as you have good policy and good governance about how the data is retained or used, who gets access. I think this is an important step for law enforcement. It's huge.

And, you know, we did our very second sprint back in 2015 with Interprotección, which is the largest insurance company in Mexico, 30 million users. And they actually deployed drones because they were so much faster than ambulances at getting to an accident scene. And they would scan and map the whole area. So that you— people— as before people move the cars and everything like that. So they had full information before anything happened. It was kind of an amazing experience. So we've seen this trend over a long period of time. I think we can expect to see this accelerate pretty radically just because of the practicality of it.

If I might add, there are regimes as the cost of robotic embodiment trends towards near zero, becomes too cheap to meter, that we in the West are unaccustomed to, but that China, for a variety of reasons, including demographics, has seen for a number of years. For example, this is well publicized that the Chinese Communist Party maintains members, officers on a per-block basis. Certainly this was the case during the pandemic in China. The West doesn't really have any concept of this. If you look at how first responders are geographically distributed in the US, it's on a per-municipality basis or precincts. There's no notion of, say, one or more officers per block that are just permanently stationed on a single block. But with drones, this becomes possible. We could have literally drones as densely distributed geographically as fire hydrants are. And you could literally just, if there's a problem in a block or a part of a block, just remotely activate the drone, and then you have an instant point of presence. Let me share the second story.

It comes out of Sacramento. There was a suspect who had a knife, and they deployed a drone with a, uh, with a magnet to grab the knife and, and, and make the scene safe for the police to enter. Let's take a look at this. So here we see the drone going. The guy is holding a knife. A magnet or electromagnet is attached to the knife, and it pulls it out of the guy's hand. As he's apparently sleeping. So, um, this is interesting. This is a drone that's disarming somebody. If you guys remember during the Abundance Summit, um, Rick Smith again with Axon, uh, showed us his taser-equipped drone, uh, and I went on stage wearing a suit, uh, to protect me, and he tased me from his drone. So I think this is coming. This use of drones in law enforcement for both observing and for trying to de-escalate a situation. The drone market right now is about $100 billion around the world. It's going to be increasing. We're seeing drones being used, obviously, in the Ukraine very famously. Eric Schmidt's been funding a drone company to help the Ukrainians in their fight for independence. Dave, any thoughts from you?

Yeah, you know, Scientific American did a great research study on why crime rates are down by half in the US and they keep coming down. And it was entirely connected to deployment of first responders in the right place at the right time, largely driven by GPS. But now with the drone footage, you can get much more accurate. You know, the first responders want to be there and they want to help, but getting the right people to the right place at the right time. But I'm a huge believer that the video footage is going to be massively impactful and the resolution just keeps going up and up and up and up. You know, the physical side of it where you're disarming somebody is— that's a little ways out. You know, that was kind of a like— very, very niche case.

Yeah, very niche case.

But the video side is like right here, right now. And as Alex was saying, you could easily, cheaply have as abundant a fleet as there are fire hydrants. That would cost next to nothing. Yeah. So that's imminent.

I'll also point out, if I may, this was all foretold by Minority Report. You remember the scene in Minority Report with the police officers deploying spiders to search an apartment complex for Tom Cruise's character. We're starting to catch up with that now. And yeah, right now it starts with a couple of police precincts in the US. They're using drones for first response, my understanding is, far more frequently in China right now than the US or the West in general are. But project forward a few years when there are a variety of new form factors. Maybe we get SPIDERS. Maybe we get drones. I have to imagine the drone, the flying form factor is a good deal more versatile for interacting with hostile scenarios, these are going to get smaller and cheaper and more plentiful. We haven't even seen what happens in the West from a first response perspective when police can deploy swarms of drones rather than just individual drones. But as cost goes down, we will absolutely see swarms.

Yeah, I mean, again, I think the public is going to have a bit of a fearful reaction to this. It really depends what the drones are armed with. It depends, you know, what the drones, the guidelines on the use of drones are, and we need to address that. I mean, getting medical equipment to a site of an accident rapidly, you know, we're going to see eVTOLs, flying cars, delivering ambulance personnel there. But getting a defibrillator, for example, to a location that's jammed by traffic, drones are going to play an important part of this, and they're getting better and better. Saleem, you're going to say?

I just remember the counterpart here. There was a fellow from the Dutch police force at one of our Singularity Executive Programs, and they were combating the fact that drug dealers were using drones. So they trained a bunch of hawks to drop mesh wires onto the drones to wrap them and tangle them. It was like so retro to be training up birds to be attacking drones. It was totally surreal.

That's actually a real problem. Below a certain size, uh, there are a lot of birds that go after these things. So the really small ones actually have a little bit of a problem. In, yeah, I mean, in the Ukraine, they're having—

they were having a huge problem because they've got strands of optical fiber from all of the drones that have attacked them lying everywhere. Oh yeah, it's a massive issue. Yeah.

Well, I'll tell you, the number one use of drones I'm excited about came from the Wildfire X Prize I talked about in the last pod. You know, where drones are able to get to a fire at inception, put it out rapidly before it, you know, causes hundreds of millions or billions in damage and, you know, causes the loss of significant life.

So— And we didn't— we didn't get our flying cars in the end. I mean, we have— we have flying car companies, but they're relatively sparse. But I do think we're going to have skies over the next few years that are utterly filled, that are densely filled, I should say. With these drones.

Well, the drone ambulance is the coolest thing ever because it's— no one's going to stand in the way of a drone ambulance, right? It's there to save somebody's life. But that will unlock all the technology, all of the airspace, all of the regulatory barriers, and also prove the efficacy. So that, that's going to be a great, really cool stepping stone. I can't wait for the—

one of the, one of the Substacks I put out was about the fact that we're heading towards a point where you can know anything you want, anytime you want, anywhere you want. We've got orbital satellites. We have the 200 satellites from Will Marshall at Planet. And then we're going to have an aviation layer from the flying cars and these drones imaging everything at centimeter and sub-centimeter resolution. And then all the autonomous cars, you know, gathering terabytes of data on the road. So everything is going to be imaged very soon. And, you know, the spooky side of that obviously is, is loss of privacy if you believe you have privacy. The positive side is that, you know, there's no crime. I put out a part of that blog saying when people are observed, they act better. You know, I got a lot of negative feedback on that, but I think the fact— I can imagine— it's true. I think when, when there's someone, when a despot has a CNN camera pointing at them, they do less. They do. They, they behave differently in the global stage. And one of the foundations I used to support was the Lindbergh Foundation that would fly drones over over herds of elephants and the poachers would stay away when the drones were flying over them. So—

Well, you want to hear a funny story from China? Sean, my son, who just got back from China, was talking to a guy who was mansplaining the entrepreneurship vibe in China and how to build a great company. And Sean said, "Well, it's all about the team, right?" This is what we preach at Link Ventures. You get great people, they succeed every time. He said, "No, no, no. It has nothing to do with the team." Like, "Well, then is it the business plan?" He said, "No, no, no. It's what the government needs next. That's the only thing that matters." Wow. You're like, "Wow, is that discouraging?" So I think, you know, with the loss of freedom and privacy also comes the loss of innovation. So I would think that— I don't think the drones are going to be, you know, taking away all of our privacy and all of our freedom. I don't think that's a real issue. But in general, the slippery slope does kill innovation and entrepreneurship.

Alex, should we talk about the Innermost Loop? Let's do it. All right. So our first story is out of Switzerland, and it's an important one. You know, Switzerland just voted to lift its ban on nuclear plants. So after Fukushima back in 2017, they phased out nuclear completely. Now they're reversing course. And just to give you a sense, you know, nuclear has been very slow. The country that succeeded so incredibly well is France, who has 57 operating reactors. The UK has 9. Spain has 7. Switzerland has 4 aging reactors that supply 40% of its power, and they were all due to be shut down. They're going to be upgraded instead of being shut down. Interesting. I was on the, on a Zoom earlier today with Ramez Naam, Salim, and Ramez is one of the most extraordinary thinkers in energy, and we should have him on the pod. I mean, uh, we definitely should. He would do an extraordinary job giving us an overview of all things energy across solar, batteries, and so forth. And he was saying the reason that France actually succeeded as well as they did is because they mass-produced a single reactor design instead of starting from zero and where the costs escalate and get out of hand. I asked him whether he thought other European countries would follow suit and be able to implement nuclear. And he was like, nope, not gonna happen. But, uh, it's interesting that the buzz on nuclear is beginning to soften and the need is significant. Alex, your take on this?

Europe's in a bit of a bind. So maybe here's a really relatable story. Whenever I'm in the Swiss Alps and it's not the winter, it's very difficult to find air conditioning. And I think Switzerland and a portion of continental Europe has a real energy crisis. They have lost access to cheap Russian oil thanks to recent events. They, except for France, underinvested in nuclear energy. They aren't this amazing native producer of their own solar PV, and they have a culture that one can sort of theorize where the culture comes from, but a culture arguably of energy scarcity. And now, as global temperatures are rising and as power consumption is increasing, Europe is having to do an about-face and discover— sort of learn to love nuclear energy, learn to love energy in general. And the risk, as we've talked about previously, cite to the EU 2031 scenario and other scenarios, Europe is going to need to start to radically increase the power consumption and power production per capita. And nuclear is a very— fission in particular is a very attractive way to do that. Otherwise, Europe will smolder under heat domes, including the one that right now over the past week or two Europeans have been suffering under. I forget the exact statistic, but thousands of Europeans are dying due to heat overexposure every year. It's a startling statistic. And it's unnecessary with better air conditioning and higher energy per capita. So I think this is the obvious trend of the future for Europe.

Yeah. Yeah. I mean, the good news is AI is turning energy from a— though AI demand, it's going to turn energy from an environmental issue into a capacity issue, into a national—

commercial profit-driven issue. Yes.

Well, that, but, but at least into a national capacity issue. Like every country has to deliver enough energy. Energy. And it's going to get it from, from fission in the, in the short to medium term until fusion or whatever come along to cover the baseload.

So let's talk about, let's talk about fusion in our next story. Uh, you know, interestingly enough, you know, the joke about fusion has always been it's 50 years away and holding. Well, it's now here. There are some 50 privately funded fusion companies that have raised about $6 billion 2 US companies lead the pack. It's Helion, which we're going to talk about, and Commonwealth Fusion. I had Bob Mumgaard, the CEO of Commonwealth Fusion, on stage with me at the Abundance Summit. They're expected— they're building a tokamak-like design. They're expected to build their first 400-megawatt plant back in mid-2030, circa 2032. But the second story here is Helion. It's a Sam Altman-backed company. He was the largest early investor back in March 2012, and until just 2 months ago, he was the executive chairman of the company. Apparently, he stepped down now so Helion can actually do some large-scale partnerships with OpenAI. And on June 16th, the news here is that Helion cleared the required Washington state regulatory approvals for its Orion fusion power plant., which is intended to supply Microsoft with 50 megawatts of power starting in 2028. So in success, this is the first fusion plant coming online. It's relatively small. 50 megawatts is, you know, we talk about gigawatt-level plants. This is 50 megawatts. They've raised about $1 billion at a $5.4 billion valuation. But it looks like fusion is finally here. I've got a video showing how Helion works because it's a unique design. I think it's worth discussing. But Alex, do you want to comment before I show the video?

Yeah, maybe just comment. So fusion, or the lack thereof, has long in futurist circles been the, the whipping boy of why long-promised technologies never happen. But actually, if, if you look at one of the figures of merit for fusion, the so-called triple product, which is a product of the density of the plasma, the confinement time of the plasma, and the temperature of the plasma. There has been steady progress for the past half century toward self-sustaining and net positive in terms of power production fusion reactions over the past 50 years. This has been sort of not just— it's not the case that there was suddenly some recent unlock. Although arguably economically there has been in the form of high TC, high superconducting transition temperature ribbon that's very helpful for certain architectures of fusion reactors. There has been continuous progress this entire time. So I, I think there's an interesting parallel that one can draw between fusion, which is arguably achieved by compressing enough matter into one volume that you achieve net power output, and AI/ASI, which is arguably achieved by taking enough human knowledge and compressing it into a small enough information theoretic footprint until you achieve really a phase transition that produces prompt engineering and large language model behavior, all of that. Strong parallels. I think they were both inevitable, but they're both inevitable. I'll make a stronger analogy, which is if you were watching the right metric or the right figure of merit over the long term, you could see both of these from 50 years away.

You could see, or maybe 30 years, it's just a slow linear growth over time.

You just watch the compression over time. Arguably with large language models and AI, if you were, say, watching the Hutter Prize, if you're Marcus Hutter, and it's the late '90s and you're watching the ability to compress the English Wikipedia over time, you could see LLMs and AGI happening from decades away. Similarly with Helion and all of its competitors, and you're watching the triple product, you could predict when this is going to happen and it's imminent.

Yeah. I think the— I think there's something incredible here because this is such a foundational technology for abundance.

This is the abundance thesis at Lawrence, for sure.

It's the foundational technology for this because once you have clean and cheap and dense energy, cost of computation, desalinization, transportation, manufacturing, agriculture, I mean, everything becomes the cost of just the materials at that point. So this is such a big deal. And it's hard to get— one of the hardest conversations I have with CEOs and with companies and especially with public sector is the fact that energy is becoming abundant over the next few years. And when energy becomes abundant, all sorts of other dominoes fall.

Energy is the number one correlate to GDP, to health, to education. The more energy a nation has, the better it is across the board. And, and this is something that Europe needs to learn. You know, interestingly enough, the challenge here is the fission plants, the small modular reactors and the Gen-3 plants are still not going to come online really till the early to mid-2030s. And the fusion plants, you know, getting up to 400-megawatt plants like Commonwealth Fusion or getting Helion up to that level, again, those are not going to be coming online till the early to mid-2030s. And so the question is, where do we get the energy from now? I had that conversation with Ramaz and he says it's from the grid. That it's going to be from the grid and that we just need to make use, better use of the grid. And he's got a company, Alex and Dave, called Agentic, that basically is sucking down energy to batteries in the middle of the night, you know, between 1:00 AM and 6:00 AM, and then pumping that energy out during peak storage. So there's plenty of energy on the grid if you could time shift it. So I think that was fascinating.

Yeah, I think one of the questions that I don't hear enough people discussing is what is the killer app of fusion going to be? It seems obvious we're going to get it, barring some surprise not anticipated, but will fusion arrive in time to be transformative for terrestrial data centers? Maybe, maybe not.

Will it be helpful for— You're pushing space.

Yeah, will it be helpful for orbital data centers? Maybe, but but there's also a lot of solar in space.

There's a 93 million mile away fusion plant that works really well in space. That's right.

Well, I'll tell you one killer app. Sorry, go ahead, Alex, finish.

So the punchline I was going to gesture at is I think actually space propulsion is one of the killer apps. If we get compact fusion reactors, then like that's a wonderful application.

Yes, my 9-year-old science fiction self loves that.

Yeah. Yeah, hand in hand with that, if any material scientists or chemical engineers want to work on— if we have fusion, then any storage mechanism, even if it's inefficient, as long as it's clean, is suddenly viable. So if you have cheap, cheap, cheap, virtually free fusion energy and you can put it into a car in a cheaper way than a lithium battery— right now you have to have some degree of efficiency. You don't want to throw away electricity. But post-fusion, you won't care about the efficiency of the reversible reaction. So anything is good. And that's true for launching rockets too. You know, once, once you've got fusion energy, you can create any reversible reaction very efficiently. Then you can port it out to your space station or to your moon base, have it do whatever it's going to do, come back and recharge it. You don't care if it's only 10% efficient.

If you remember Bob Mumgaard, again, the CEO of Commonwealth Fusion, when he was on stage, his goal there, once he gets his, his his unit working is to pump them out to create a, you know, the machine that builds the machines. The same thing here for Helion. I've got a short video that explains how Helion works. And given the fact that it may be the first fusion plant coming online and it's unique in how it works using magnetically propelled plasma and then magnets to pull electricity out of the plasma, let's take a listen to this. I think it's valuable for our listeners to hear about Helion.

Helion's pulsed fusion device directly recovers energy, which is used to generate zero-carbon electricity from fusion. It starts with Helion's fusion fuel, deuterium and helium-3. These fuels are injected as a gas into Helion's formation chamber, where they are superheated into an ionized gas called a plasma. The machine's capacitors are charged and send electricity to magnets that wrap around Helion's device. The magnets invert the plasma's magnetic field on itself into a toroidal, or donut shape. The device's magnets fire sequentially, accelerating the plasmas toward each other at a velocity greater than 1 million miles per hour. They collide in the fusion chamber and merge to become one hot, dense plasma. In the center of the device, the machine's magnetic field is rapidly increased, compressing the plasma with a powerful force over 10 tesla. All these fusion reactions within the plasma convert matter into new energy, which strengthens the plasma's magnetic field. As the plasma's magnetic field gets stronger, it pushes back on the magnetic field of the machine, causing a change in the machine's magnetic flux. In accordance with Faraday's Law, this change in flux induces current in the machine's coils, which is directly recaptured as electricity and returned to the capacitors that originally charged the magnets around the machine.

You know, we really are living in the future. When I see that, it's like extraordinary. And their goal was to mass manufacture those helium plants. Well, if I had to design—

cool too is that, that Helion raised a billion. Also Commonwealth Fusion, remember we had dinner with him in Riyadh, uh, March, and he had just raised a billion. So, you know, when we were at MIT, the budgets for this were in the tens of millions. Research budget. Yeah, yeah, research budgets. Now suddenly, you know, we've said this on the pod many times, but we're actually truly investing in the commercial sector in hard science for the first time in my lifetime. But something great will come out of those two $1 billion investments for sure.

I mean, Alex, look, to bookend this, once you make energy abundant, every other scarcity becomes a negotiable.

It's, it's probably also worth, I think, pointing out what the 'so what' of that explainer video is. So unlike many other fusion architectures, the whole point of Helion's architecture is direct recovery of energy from the fusion plasma. In a more conventional, say, tokamak style or other fusion reactor, there's a bucket brigade of energy production. You create the plasma through inertial confinement or through magnetic confinement, and then the plasma will be used to heat something, maybe water, and that produces vapor, and then the vapor goes into a turbine and you turn the turbine And that you recover electricity from the turbine inductively inducing via magnets currents in wires. And it's like a 10-step process. The whole point, and what's potentially quite seductively attractive about the Helion architecture, is you're just directly recovering from magnetic fields that are being induced by the plasma and then currents induced by those magnetic fields. You're almost directly recovering free energy from the plasma. So you're skipping a whole bunch of steps. It's potentially a lot more efficient. It's potentially a lot— I was going to say a moment ago, if I had to pick a sort of pattern match, a Mr. Fusion from Back to the Future Part II architecture, and identify the archetype of any one of the, the now many fusion startups that are out there, I think Helion is the closest to being a Mr. Fusion startup because All of those extra steps that are being skipped could lead to potentially radical compactification of the ultimate fusion implementation. So it's very exciting. Yeah, it is.

It's a beautiful, beautiful design. And again, something he can mass manufacture. And, you know, where does it go? It goes into every, every township, you know, depending on the size, every city, every place that you need baseload energy production.

Compact fusion is going to be a thing.

Well, if you told me that I'd be listening to a little chipmunk voice explain Faraday's law, I would say no.

It was like, am I running this? Am I running this at 1.5? No, that's the voice they chose. All right, let's jump into AI and a really fun story to kick us off. There's a $1.8 million incentive prize that was founded by Nat Friedman, former CEO of GitHub, and Daniel Gross. Matt, I know Nat is your friend, roommate. Roommate. Carlo Cocuzzi.

He was my first roommate at MIT.

Fun stories. Yes. And that was on stage in March of 2023 at the Abundance Summit. And he announced the Vesuvius Challenge. And here it is being won, you know, some 3 years later. So, the challenge was there are these scrolls that were basically buried and burnt under Mount Vesuvius back in 79 AD. And the scrolls are fully carbonized. You can't open them and read them without destroying them. So they said, can we use technology? You know, can we use CT scans to gather the data and then use AI to read them? Well, for almost 2,000 years, they've been unreadable and they have just been one. So this is the first time it's done. You can see here in the image these, these scrolls of ancient Greek that have been linearized and laid out by the AI. 22 columns of ancient Greek text. And there are still hundreds of scrolls that can be read. I mean, this is using AI to basically do, you know, look back in time. Alex, this must be a favorite one for you.

And not just because of the Nat connection. Objectively, I think computational archaeology powered by AI is going to be utterly transformative in the future. I've argued from time to time— You've told me this before, sure. Yeah, the killer app of the singularity is superpowering computational archaeology. And I'll inevitably cite to Nikolai Fyodorov, one of the fathers of the strain of philosophy called cosmism, the idea that humankind's common task is to essentially resurrect every human who's ever lived using technology. And I see in the Vesuvius challenge the very beginning of a larger arc of technological progress that may require completion of the, this singularity that we're in to fully run its course. But imagine, just as a thought experiment, imagine if we could do what the Vesuvius Challenge did, not just for, say, performing high-resolution scans of the positions of ink, uh, or the, uh, the X-ray analysis of small blotches of ink in order to recover scrolls that were otherwise preserved from the eruption of Vesuvius. But now imagine being able to do this at planetary scale. Imagine that somehow— not going to say specifically what the mechanism of action would be— but imagine that we're able to scan the Earth and gain fine spatial and temporal precision, or position-momentum canonical coordinates if you like, for every atom on Earth. And imagine what if we feed the entire Earth state into an AI, what we might be able to recover about Earth's history. I think the answers would be quite transformative. One of my favorite anecdotes is environmental DNA, where if you go for a walk outside, you may or may not realize this, but you're just drowning in DNA that's been aerosolized from animals alive and dead. If you look, if you dig into the soil, If you dig into, say, a soil near a cemetery where human bodies have been buried, DNA has a surprisingly long half-life even under environmental conditions. So there's a lot of state left over from Earth's past, not just in these scrolls that were preserved by volcanic eruptions, but in general. And I think, again, idiosyncratic position here, but I think with strong enough AI in combination with strong enough scanning technology, at some point in the future, we will be able to recreate large fractions of our past light cone.

Well, you know, this is what Colossal's doing in a way, right? Going and extracting DNA from fossils, bringing back the dire wolf, bringing back the woolly mammoth in a very limited slice. And this is taking that to extremes. Yeah.

Yeah, there's another angle to this story too. If you're watching this podcast live and you look at the image and really zoom in on it, AI is very, very good at interpolating these fragments, these little— like, if you look at the characters, no human being could ever reverse engineer what that original character was. But the AI is really good at filling in those blanks. And it's not regular LLM AI. It's not your, you know, your Anthropic or your OpenAI transformer. And this is where, Peter, you know David Siegel, right? The founder of Two Sigma. Sure. He's on the board of MIT. He has a project called Project Open Athena. Which he's hugely funding, which is designed to give AI compute resources to people who have alternate versions of AI that are not necessarily transformers that are very, very good at these types of problems. And so there are many, many, many of these. And, you know, as Alex was describing, if you wanted to look at fragments of DNA that are lying around and reverse engineer what happened in that room, that's a really good use case. But, you know, all of these world events, you know, historical events leave a little trace that's kind of scattered around. And there's usually only one interpretation of history that could have created that trace. Impossible for humans to glue together those fragments. But it's not LLM AI, it's, you know, core neural network AI that's built from the ground up to solve that problem.

Let me do a shout out to the non-technical founders out there. You know, if you're— if you've got an idea for a technology or for a company, but you're not a technologist, and you want, you know, it's your dream to make it happen. Imagine being able to use an incentive prize like this to actually aggregate the best experts in the world to come help you solve your problem, right? So, in this case, you know, a $1.8 million prize probably brought, you know, on the order of tens of millions of dollars, if not more, of genius to apply and solve the problem. So think about this. This is the basic principle of XPRIZE, to get people around the world to focus on solving a problem. And we get 30x the prize money spent cumulatively to solve a challenge like this. Salim?

I wanted to stress the prize model and the incredible powerful— what really strikes me in this is the before and after. The image of before and getting real actual data and information That would just be so mind-boggling.

But the audacity that you think you can do that. And I think this is, this is the abundance thesis again, that challenges we never thought were solvable fall as a result of the technologies that we're building.

Voice agents are just software, but software deserves a real development platform. I'm Nick Leonard, CEO and co-founder of VoiceRun. VoiceRun is your runtime and development platform for voice agents. On VoiceRun, agents are built and configured in code. No limiting no-code platforms. For developers, that means total control. And for enterprises, that means extensibility that meets your complexity. We've built VoiceRun CLI-first, meaning we've kept your Claude code, Codex, and even your OpenCLAW in mind when we built it. Your assistant of choice can build and deploy voice agents, can test and simulate scenarios, and can analyze and evaluate at scale. In other words, we've closed the loop on voice agent development. We don't build demos destined to fail in production. VoiceRun is where the best voice agents happen. Visit us at voicerun.com.

All right, let's move to the SpaceX universe. Grok 4.5. GPT-5 is coming out based on a 1.5 trillion parameter V9 foundation model. Interestingly enough, Elon has made the claim that he's going to iterate a new model and release it every month for the rest of the year. Alex, let's go to you on this one.

And he said something I think even spicier, if I understood his announcement correctly, that he was going to start pre-training every month and not just distillation cycles or not just post-training or fine-tuning cycles per month, but start a new pre-training run every single month. That's completely, a completely new model from scratch.

From scratch.

From scratch, which means pre-training, which is, I mean, it's audacious, it's brute force. It's exactly, I think, what the world expects of Elon, a brute force attack. Uh, I, I've had a number of folks since we first discussed, um, in an earlier episode, my comments about Grok being put on life support in favor of SpaceX's hyperscaler resources being handed over. You got a lot of hate mail on that one. I got some spicy comments. I had people saying, okay, well, Elon's announcing this, Elon's announcing that. Are you retracting your comments about Groq being put on life support? And I think, again, my comments may have been misconstrued. I want the frontier to be competitive. Right now, we are in arguably a duopoly between OpenAI and Anthropic. They're just running away with the race. With Chinese open weight models a few months on their heels. And I want there to be a competitive frontier. And I think Groq is one of the possible competitors along with Gemini, and maybe Meta will come up with something eventually. I want it to be competitive. And Elon's strategy historically vis-à-vis xAI has been brute force. And I think hopefully as SpaceX brings more and more compute online, this sort of brute force approach where he eventually more compute than everyone else in combination with off-the-shelf algorithms, maybe this will work. I see the beginnings with Groq 4.5 in combination with Cursor. Right now, the Cursor acquisition is taking the form of post-training of models that he already had. In the near future, Elon's promised that Cursor is going to be part of the pre-training recipe. If he can make that work, I think Groq has a fighting chance to join the frontier through brute force. Compute efforts.

There's a few other things he's doing. First of all, he's got, you know, tens of billions of dollars of dry powder now to focus on this. The second thing is, and he's announced, he's bringing in his smartest players from SpaceX and from Tesla to work on XAI. He is by no means giving up the ghost here. He wants to be number one. He wants to beat OpenAI. He wants to create, you know, Grok as the ultimate greatest seeker of truth. Truth. Uh, and I, I, again, I would never bet against him.

Yeah, I think the best bet here is he will brute force his way back to the frontier. That's what I'm hoping will happen.

Yeah, well, he also doesn't— he's got—

what you're just doing— yeah, what you're describing is a two-horse race. He doesn't perceive that at all. He thinks it's still a race against Google, specifically because the, the training directly toward a customized chip is a 10 to 100x unlock, and he doesn't perceive Anthropic to be near there. And then OpenAI has some activity there, but it's nowhere near— but Google has already got, you know, their whole vertical monopoly stuck together. And so he doesn't think he's going to lose because he thinks that he's going to actually be the first to have a reasonably good model that then is custom silicon that supports the model instantaneously with the design of the chip being AI.

You know, we're seeing verticalization, verticalization win every race here. We'll see it in the space industry. We're seeing with Google, seeing with XAI. I mean, it's, it's fascinating. Uh, Celine, please.

Yeah, there's a thought, there's a, there's an idea here that could allow him to leapfrog even faster, which is we're moving from models that were built on human artifacts towards models being trained on the actual process of human machine work, which is where the cursor data becomes really useful because that creates a flywheel. We have a better model, more usage, richer workflow, and then a better model, right? And I think that's going to be serving very well in the future.

I'll comment on, on that point narrowly. I think it's an, it's an interesting debate one could have. So the Cursor acquisition, I think I, I viewed this at the time, commented on this on the pod previously, I viewed as a brain transplant for the future trajectory of Groq. Given that ChatGPT also is going through the same brain transplant of making Codex, essentially a model that, or a class of models together with scaffolding that were optimized for code generation and recursive self-improvement, that's becoming the new mainline ChatGPT. By analogy, xAI and SpaceX acquiring Cursor post-IPO to make Cursor essentially the new mainline Groq, I, I view as an analogous move. However, the dataset from Cursor, which consists in part, my understanding is lots of reasoning traces driven from developers who wanted more code gen. I think that buys Elon sort of a leap to the near frontier in terms of code generation, but it will still be incumbent on Elon/XAI/SpaceX to achieve their own recursive self-improvement loop. You can only get so far by post-training on developer or user traces, at some point the models need to start developing better models. And that's been historically, I think, a strength of Anthropic, probably a weakness of xAI, but maybe he can brute force himself to the front of the recursive self-improvement loop.

Let's jump into a few Anthropic stories. Anthropic's flagship model, Fable-5, has been offline for 15 days because the US government pulled it for national security fears. And hopefully we've talked to this ad nauseam. Now Axios reports it may be back within days. Secretary Luttenegger accredited, you know, credited Anthropic for working on the risks through the Pentagon and NSA. They still haven't signed off on that. Interestingly enough, Stripe recently reported that they ran a test using Fable-5 to overhaul 50 million lines of codebase in a single day, work that would have taken engineers many months. The government's treating Fable-5 as a commercial AI model, like a like a controlled munition, taking it offline and re-permitting it by users. And we're going to have to see how this evolves in the coming days. So let's talk about this Fable 5 coming on. And Anthropic is another story. Alex, if you want to cover that one too. Sure.

So I think there are a couple of interesting notes coming out of this. If Fable 5, as I do expect, will eventually become available. One of the, the more interesting takes, I think, is this will have been a period of a few weeks when allegedly Chinese organizations that were leveraging access to Anthropic's frontier models or near-frontier models for reasoning trace distillation will have been denied that access. So we talk on the pod all the time about how the US maybe has a 3-month lead or a 6-month lead or an 8-month lead, depending on how you count. This— there's, there's a certain sense in which this one-month-ish shutdown, future perfect tense, may have or will have denied China at least a month of catch-up time. That's a generous interpretation. A less generous interpretation is that this will mark what we'll look back through time— historians will look back and say this period in time marked the period towards the, uh, the middle or the endgame of recursive self-improvement when months counted and the permitting of frontier intelligence became almost a zones of thought, to borrow from Verner Vinge, or a block system, to borrow from the Cold War, when either before anyone could access frontier intelligence versus after, when now you have to be a US person and there is strict export control for capabilities and there's a nonproliferation regime where you have to gain access. If you're a non-US person, you gain access to models that are maybe a few months behind the frontier. But I also think a few years from now when we look back on this time, I, I think this will have been, yes, there was a phase change in terms of the diffusion of frontier models, but I do think I think sometime in the next few years, we're going to get to the end of the recursive self-improvement rainbow, and there's going to be a perfect model. And we'll look back and say, this was just a period of months, you know, a delay, but ultimately everyone is ultimately going to figure out what the perfect model looks like.

Was this period of a month also a chance for a lot of critical systems to safe themselves against Fable 5?

The most essential ones, maybe, but by— and yes, there was the standup stand up both within the US government of vulnerability scanning. And outside the US government, we saw 3 or 4 independent nonprofit or for-profit organizations stand themselves up to do bulk vulnerability scaling using Fable 5. But I think to the extent pause-ism has its day in the sun, I don't think this actually decelerated any AI at all. I think this is a net accelerant because even though public, sort of private— I'm not thinking about decelerating AI.

I'm thinking about decelerating black, you know, black hats from being able to get in there and, and penetrate.

I don't think so. I, I think black hat capabilities are proportional to capabilities overall. And I think what we saw during this time is Chinese models like GLM-5.2 gain ascendancy for anyone who wants near-frontier capabilities to do essentially whatever they want with them. It may not be as capable as Fable-5, but this creates enormous pressure on Chinese organizations and the Chinese frontier labs to catch up. And so I, I think, as with the original pause AI movement, it had the net effect of accelerating capabilities globally. Same idea here.

Uh, quick two implications from this slowdown. I just want to point out one for investors, right? Regulatory risk is now one of the first-order variables for— you're looking at companies, um, because it's real. Model. And the second for technical founders is don't build your product or your company on a single model. You have to make sure you're able to swap out models because you have no guarantees as we're going forward. And Dave, you saw that. Do you want to— Dave and Alex, you want to talk about the Sonnet announcement that came an hour ago?

Dave, you first. Well, I mean, it's really obvious that AI is sold out and that When Fable 5 came out, you know, they doubled the price on us, and— but you had to use it because it's just so good. Uh, so it's, it's pretty clear that Sonnet 5 now is a way to kind of fill this gap until Fable 5 is back out. But the price point is very high given the amount of compute that they have to use to deliver it. But people will still buy it because, again, AI is sold out. So you see the revenues at Anthropic going through through the roof. And it's because the demand for AI way outstrips the underlying chip supply. And so, you know, the byproduct of that is a lot of things. You know, only the very top of the mountain use cases are going to get access. And then after Fable 5 comes back out, you know, I agree with what Alex was saying a minute ago. This moment in time will be remembered in history. This is the intersection of AI and the government that's never going to go away now. Now. But not every person on the planet and not every company on the planet is going to be able to access the models. And it's supply constrained at the same time. So then there's going to be preferential routing. SONNET 5 is a kind of mediocre capability at a high price point, but people will still need to buy it. And then Fable 5 will come out at its extremely high price point. That's my read on SONNET.

Alex, any addition?

I think this is a bizarre announcement. Admittedly, this is a hasty analysis since Sonnet 5 was released right before we went to air here. But I've been trained, as I think the majority of sophisticated users have been trained, to expect that the Sonnet series from Anthropic would represent some distillation of the Opus series. And similarly, the Haiku series represents a distillation of the Sonnet series. And that as, as you go down towards towards smaller, lower parameter count, more distilled models, you see some optimal frontier emerge in price-performance space where performance, at least throughput goes up, price per token goes down and performance goes down. And maybe I'm missing something and maybe the answer will reveal itself in the next few hours, but just looking at the cost versus performance at agentic tasks, curves that Anthropic released with Sonnet 5. It's a little bit bizarre. On the one hand, Sonnet 5 is an optimal frontier, sort of a Pareto improvement over the, the last version of Sonnet, Sonnet 4.6. But Opus 4.8, which has been out for, uh, for what in these singularity times passes for an eternity, is better. It's superior on a cost-performance basis. So I'm not 100% certain I understand what Anthropic is hoping to achieve with At 5, I can tell you, Alex, it's—

everybody's working on these frameworks where you can bounce from model to model while keeping the context intact. And all of the work that's piling up, the prompt history and the intellectual property is piling up like crazy now. And so you have a choice between working in an open framework, but Claude Code and Claude Cowork are super compelling, you know, with all the MCP wrappers and connectors already built in. So the easy choice, kind of the Apple-like, I'm going to pay more, but it all works. Works choice is to go with an all-Cloud stack. And then when you're working in Opus 4.8, if you have a simpler question, you go to Sonnet, or it automatically goes to Sonnet. And if it's an even simpler question, it just goes down to Haiku. So that's the easy way to go. But the more cost-effective way to go would be to bounce over to a different model, but then you have to use a third-party context management platform. So that's the tension. But this is kind of like Anthropic becoming the Apple of AI where you know you're overpaying by some insane amount.

But it works. I feel like there, my intuition is there's some branding going on behind the scenes that we're just missing. Like maybe in the sense that Fable and Mythos are the new high-end models. Maybe there's some sense in which Sonnet is the new low-end, like Sonnet is the new Haiku and just viewing it through the branding of Sonnet is maybe incorrect. Maybe we should be thinking of it as the Haiku level and it's just that Fable 5 isn't accessible. There's some weird Pareto optimal frontier, I think, that's missing in order to explain why we've seen a reversion of this optimal frontier.

We will find out and we will— We will get to the bottom of it. —get to the bottom of it. Welcome to the health section of Moonshots, brought to you by Fountain Life. AI is having an outsized impact on every aspect of our lives— how we teach our kids, how we run our companies. It also is having a huge impact on health, helping you prevent heart disease. One of the key things— I'm here with Dr. Dawn Musalem, our chief medical officer Sir Fountain, heart disease has been personal for you as well, hasn't it?

It really has, Peter. When my daughter was 5, my husband died of sudden cardiac death, and so this is a topic that is one that I am mission-driven to try to eradicate. Prevention first and early detection is absolutely critical. 50% of people die of heart attacks with no warning signs. Silent killer.

No shortness of breath, no pain, no nothing. No, silent killer. They just don't wake up in the morning.

They don't wake up, and so, you know, AI, this is our mission, to advance science, to try to help to one day democratize Wellness. We know at Fountain Life, when we do this CT angiography with AI analytics, we are actually finding that 88% of people coming in have detectable coronary artery disease. But Peter, what's more alarming to me is 23% of those individuals had soft plaque. This is the plaque that would not traditionally be seen on CT looking at calcium scores alone. And this is the plaque that we must intervene with, with the multimodal testing we're doing, including diagnostic laboratory studies partnered with healthy lifestyle recommendations.

So listen, make sure you understand what's going on inside your body genetically, metabolically, and cardiovascularly. You can know, and it's your obligation to know. So check it out at fountainlife.com/peter to find out more and really make sure that you're the CEO of your own health. All right, back to the episode. All right, mates, let's jump into data centers and space. And for that, we're pleased to bring a friend on. Philip Johnson is the co-founder and CEO of Star Cloud, a startup building space-based data centers. Star Cloud has raised about $200 million with its last round over $1 billion. Famously, Philip's company launched Star Cloud One, the first NVIDIA H100 GPU in orbit. And apparently, Philip, you've trained your first LLM in space. Star Cloud launched in November of 2025 on Falcon 9. Welcome, Philip.

Pleasure to have you. Thanks so much for having me. It's a huge honor. I've been a longtime fan, so very privileged to be here. Awesome. It's mutual.

It's mutual.

Yeah, I mean, I, a lot of stories we want to talk about in data centers and space to wrap up today's episode, but let's start with a little bit about Star Cloud. Tell us about the company, your vision. Where are you guys? What have you done with the H100? And yeah, Yeah, let's go there. Yeah, yeah, for sure.

Uh, so we started about, uh, 2 and a half years ago in January 2024. Um, we're a team of about 20 engineers based in Redmond, Washington, actually right down the road from the Starlink manufacturing facility. So about half our team came from SpaceX and the rest are from the data center companies up here, so AWS and Azure and the others. Um, and then, yeah, so we launched our first spacecraft, Star Cloud 1, in November last last year. I had actually 5 GPUs, uh, 2 from Arm and 3 from NVIDIA. But the most important one, as you mentioned, was the NVIDIA H100. And so with that, we were the first to train a model in space. We trained, uh, we trained NanoGPT from Andrej Karpathy, which I'm— very tiny model, but it still counts. Totally counts. And then we actually were the first to run a version of Gemini in space. So we ran Gemma, which is the DeepMind's cutdown version of of Gemini. And then we've done a few other things. So now we're doing much more kind of useful workloads. So we've just been doing high-powered inference on SAR data, synthetic aperture radar data, in collaboration with various DOW entities. And also we actually just played Doom. We ran Doom on the space, on Star Cloud 1.

As one does.

Little latency there, probably.

And before we get into it, I just want to say you guys have been ahead of this trend more than most. So thank you for your support. I know Alex has often been vocal about his support, and Dave gave me a shout out early on. And Peter, I know you, in the interview with Elon, you are also very supportive. So really appreciate it.

Yeah, well, you know, you're building the Dyson Swarm, and this is important for our great-great-grandchildren, for all of us.

And Philip, just speaking for myself, I want there to be multiple Dyson Swarms. Swarms. We can't have a solar system-scale monopoly.

Agreed. Maybe a Matrioshka swarm, so we can have swarms inside swarms inside swarms.

So there's Jupiter and Saturn, plenty of atoms left to disassemble.

So, so Philip, give us a, give us a sense of where, you know, you gave us a sense of where you are right now in terms of Star Cloud One, uh, but where do you go next?

Yeah, so we've got 3 launches booked next and we're launching Star Cloud 2 in, uh, booked for January. Um, it's about 100 times the power generation of Star Cloud 1. We'll have by far the largest commercial deployable radiator in space. Um, and so yeah, that's gonna have a whole bunch of H100s also flying, the Blackrock chip from NVIDIA, and also some other interesting things like some Bitcoin mining ASICs, um, and also AWS Outposts, which is an on-premises server blade, so we can run one, um, like an instance of EC2 on orbit, which is useful for DOW customers. Um, and then as soon as possible, we'll be launching a much larger spacecraft, what we're calling Star Cloud 3. It's a, uh, 200-kilowatt, 3-ton spacecraft which will fit on the Starlink— the Starship HEZ dispenser form factor. Actually, I can— if you guys can see this, I can even show you. We've got it welded up, a version of it welded on the, on the ceiling, the chassis. Just so you get a sense. So this is the, this is the, the length of the chassis of it. So it's about 6 meters long. And with this, we'll have huge 100-meter deployables hanging off the side of that. How much power are you generating for that, to power that? That's 200 kilowatts. Um, and we can— 2 tons, you said? 3 tons. Yeah, 3 tons. Okay. Yeah. And we, we can fit about 50 of them per Starship. So talking about, um, about 10 megawatts of new compute capacity per Starship launch. And yeah, we're hoping to be launching very frequently on Starship. I mean, they're building just absolutely enormous capacity, as I'm sure you're all aware. These two Starship gigafactories designed to produce something like 3 Starships per day. So hopefully there'll be capacity for us to launch.

Um, and obviously, you know, we're looking at some other launch vehicles as well, but that's the primary So I have to ask the elephant in the room, which is when you're going in to pitch for investors and they say, hey, but there is, you know, Elon and his mega plans and Google's far behind. And are you picking a niche in this area that you're going to be competitive in?

Yeah, it's a great question. So the main niche we're going after is to be more like an energy and infrastructure play than like our own cloud. So for example, we've got an agreement with Crusoe where we essentially say to them, hey, we have a box and that box has power cooling and connectivity and we'll work with you on whatever chip architecture you like and you can sell to whichever customers you like for whatever price you like and you just pay us a fee in the same way that you would pay a rental fee to somebody like Equinix. You pay that kind of rental fee to us and you finance the chips. So that's kind of the approach we're thinking I think, um, in the early days at least, it looks like SpaceX is primarily going to be serving their own, you know, Cursor, Groq, and xAI workloads. And, you know, in the more medium term, we'll be providing a cloud service to folks like Anthropic. I think it's probably a bit further out that they're looking at doing— being just a pure infrastructure provider. Um, though I think actually Elon did mention something about that relatively recently, but That's the idea. I mean, so in general, it's a good point because of course we're going to have a higher cost base than SpaceX because they own the launch. So long as we have a lower cost base than all of the other hyperscalers, I think we're in a reasonable position. You know, lots of, um, if we have a lower cost base than OpenAI, for example, they're going to need to figure out a space solution, and either they pay, you know, uh, XAI to run workloads on SpaceX's satellites. In OpenAI's case, I think that sounds unlikely. Or they start building their own satellites, right? Yeah, or they start building their own satellites, then it's possible. But I mean, they're going to be way, way far behind. Or the last option is they'll look around, you know, in sort of 2-3 years as Starship cadence ramps up, and they're going to be like, okay, we're going to get left behind if we don't get on top of this. And they'll be like, okay, who, who's the most advanced in the market, you know, maybe besides SpaceX? And at that point, I think will have, um, a very significant lead over anybody, perhaps besides SpaceX. Interesting.

Dave, Alex, you want to jump in? I've got two questions.

Okay. Okay. One was, two years ago, you know, we didn't have data centers in space in our bingo cards at all. What had you—

hold on, some of us did.

Some of us did. Definitely the mainstream did not. I know, Alex, you've been talking about Dyson swarms since you were like 5 years old. Um, but, uh, what has you jump to that and say we're gonna do that?

Yeah, yeah, yeah. We actually started off by doing— well, we were initially just looking at space-based solar. So in mid-2023, I actually just on a randomly on a weekend took a trip down to Starbase, Texas, like even before the first launch. So not as many people were looking at it back then, and I was just blown away by the scale of what they were building. And I my head, I was like, okay, all of the concepts from sci-fi that I remember reading about— Asimov was talking about space-based solar in the '40s even— are gonna come true. It's just a matter of a timeline now. And so, Philip, if I could—

space-based, you know, solar power satellites for beaming energy down to the ground?

Yes, exactly. Yeah, so huge, huge solar panels in space, and then either using infrared or microwave to beam power down down. Um, the main problem— and we spent, you know, several months on the math, on the math, essentially, on the break-even. We wanted to know, okay, what is the break-even launch cost at which space-based solar makes sense? And we came to a number around sort of $50 a kilo where that would make sense. And initially we thought, okay, that's kind of good enough. That's— Starship will get there at some point. But then, and, and, you know, we started working on that. But then we were like, okay, well, the problem with space-based solar is you lose 90 or 95% of the energy in transmission from space to ground. And we were looking around and saying, okay, once we get the power down, what are we going to be using it for? And even 2 years ago, most new energy projects being built, particularly in the US, were being built primarily to power data centers. So the thinking was, okay, well, either directly or indirectly, that power is going to be going into data centers. And so if we instead can find a cheap way to get the data center to space, let's rerun all those calculations to know what would the breakeven launch cost need to be that that business would break even, you know, would make sense versus terrestrial. So we reran those numbers, we came to a launch cost breakeven of around $500 a kilo if we had a cheap way to get the data center to space. And then that became the basis of a white paper that we put out in summer 2024. And then that essentially became the basis of the company.

Wow, that's great. That's a great story. And it's an interesting entrepreneurial story for all the entrepreneurs listening, right? It's like you're going down one road and then all of a sudden you see a massive opportunity, especially when you can go deep enough to look at it. Because the idea of space solar power satellites has been around since the '70s. Gerard K. O'Neill at Space Studies Institute looked at his solution was to build them on the moon and then fly them to Earth orbit where the launch costs obviously are are de minimis.

I'm sure that will happen. I'm sure that will happen.

Yeah. Yeah. Second question is, what are the couple of biggest bottlenecks you're facing right now?

We are actually very constrained on launch right now, as everybody is. We're so constrained that we're trying to book now on Relativity Space's first launch in January. Ah, right. I'm really excited about that. I think that would be super cool to have a launch on that..

And for those, uh, we've covered Relativity Space in a podcast ago. This is Eric Schmidt's company where he's CEO. He bought it, uh, from, from, from Tim and, uh, and Jordan when it missed a financing. And, uh, Relativity Space is about the size of, uh, uh, about the size of New Glenn, half the size of Starship.

But if you're building toward the PEZ dispenser, actually that's a problem. That's really interesting.

How's that going to work out? Yeah, it is a problem. I think we're going to have to have two form factors. I mean, the primary one we're working on right now is for those, but I think we're going to need a form factor that will fit on— also Stoke Space is the only one which has a reasonable upstage, or the only one that right now is seriously working on a reasonable upstage. So we're also looking at a Stoke launch vehicle form factor But to be frank, the, the relativity one is maybe just to take a step back. So the business has two phases. The first phase is while launch cost is relatively high, we're launching on Falcon 9 and others, that we're primarily serving edge and cloud, uh, you know, providing edge and cloud services for other spacecraft, particularly, you know, DOW and Earth observation constellations. And then on a sort of 3 to 4 year time frame, as Starship ramps up cadence and production, that's when we switch over to competing with all terrestrial data centers on energy cost. But in the— yeah, so all of these— when I say we're launch constrained right now, I'm even talking about for the first business. We've got 3 launches booked for next year, but if you want to book anything for 2028 on Falcon 9, there's just nothing available. Wow. The government's just plonked down 20 launches, which has bumped everything back. Um, and so yeah, unless you can get some priority, and we're going through various channels to try and get some priority on that.

But so this is GPUs, GPUs for use of, uh, processing in space? Yes, correct.

So we, we will receive, for instance, raw imagery, hyperspectral or SAR or other types of satellite sensing data. We will, instead of having to wait for a ground station and downlink enormous amounts of data, we can process all of that on the edge. And actually, this is one of the demonstrations we've just we've done is process a whole bunch of SAR data, identify the coordinates of a tank, and then just downlink the coordinates of that tank. And rather than having to, you know, it could take 3 days to get enough ground station passes to get 100 gigabytes of SAR data off satellite. Rather than doing that, we can get all of that, receive all that data optically in space, process it on orbit, and then just downlink the insight. And that's, that's the use case we're building towards right now.

I'm curious, Philip. So the premise there is presumably bandwidth is scarce, scarcer maybe than launch. So you have to do a lot of edge inference in, in LEO or wherever you're doing this. I'd love to ask you the sort of a similar question that I asked Will Marshall of Planet in a previous pod episode. Let's project out 10 to 20 years, well past the current bottleneck in heavy lift or heavy launch capability. What do you think as as the founder and CEO of one of the incumbent Dyson swarms, plural, what do you think the Dyson swarm or Dyson swarms of call it 20 years from now look like? Does it look like LEO? Does it look like sun synchronous orbit? Does it look like the moon? Does it look like a Dyson swarm around the sun? Paint a picture for us what 20 years from now the Dyson swarm or swarms look like.

Mm, that's a, that's a good, good question. I hope within 20 years we've started putting significant amounts of compute in a Sun orbit, probably starting with the Lagrange points, although you probably don't want to clog them up too much. But even just a, um, just a distinct Sun orbit, uh, that, you know, trails Earth or, or is in front of Earth, um, because Certainly you could fit about 10 terawatts of compute in the Dawn dust sun synchronous orbit, and then you're back to flying, uh, in orbits which have an eclipse, you know, sort of 45 minutes of every 90-minute eclipse. That then drives the cost up significantly because you need batteries and all these other things.

Scarcity of real estate. Yeah, SSO is going to get crowded.

That, that's— you're, you're one of the few people, Philip, I hear talking— this is like overpopulation on Mars talking about SSO getting crowded due to the SSO Dyson swarm and then overflowing back to other orbits. That's fascinating. Yeah. Yeah.

Well, because there's only one SSO orbit that is, you know, it's very rare and beautiful orbit, the dawn-dusk sun-synchronous orbit, which flies over the terminator line. I mean, 10 terawatts is a lot of compute. That's like 20 times the entire US power grid. And probably in 20 years, my expectation roughly is that in 10 years we might hit a point where most new compute capacity is being deployed in space. If you ask me what is the total percentage of compute in space at that point, it's still probably less than 5%. In the same way, you know, right now most— in certain parts of the world at least, you know, Norway or whatever— most new cars coming from production line are electric. But if you want to know what is the percentage of the fleet which is electric, it's like 4%. Still, and it just takes a long time to replace all of the capacity that we're building on Earth to— terrestrially right now. Um, so in 10 years, I would say most new compute capacity will be in space. In 20 years, it could be a— I, I would be surprised if more than half of all compute is in space in 20 years even. Um, but, but beyond that, certainly there'll come a point, and it's probably more like a 50-year time frame, where 99% of all computers in space.

I also just have to ask, a second elephant in an increasingly crowded room, uh, if Lyft over the next few years, if heavy launch is the main bottleneck, have, have you— is there some plan in a back room somewhere for you either to build or buy your own vertically integrated heavy launch provider if that's the main constraint?

Um, we are in quite serious discussion options to partner with various launch providers. Probably shouldn't go too much down that. Yeah, it's early days, and you know, if SpaceX can provide the capacity, we'll be very happy customers of SpaceX. You know, if not, then we'll need to figure out something. If you ask me, do I think SpaceX has a monopoly in launch in 5 years, the answer is yes. In 10 years, the answer is no. You know, the world is going to be And we'll talk about Rocket Lab's purchase of Iridium in a moment.

Yeah. And in fact, what I'd love to do is jump into a few stories, Philip, and have you comment alongside the mates here. Yeah, the, yeah, the first story is a recent conversation by Elon about basically Earth to or space to Earth telephony. So let's jump into that. So I'm going to play a short video. This is a video clip of Elon speaking at the All In Summit. And then let's talk about it afterwards.

So the phones that are able to use the spectrum that was acquired probably start shipping in around 2 years.

And then we also need to build the satellites that are going to communicate on those frequencies.

So in parallel, we're building the satellites and working with the handset makers to add these frequencies to the phone.

Phones, and then the satellites and the phones will then handshake very well to give high bandwidth connectivity.

But the net effect is that you should be able to watch videos anywhere on your phone.

So, uh, a fascinating story. Vertical integration again, the whole stack. SpaceX owns it— launch, satellite spectrum, and increasingly compute. So direct-to-phone, um, you is an interesting product. It's going to be space-based internet to everybody on the planet. Uh, thoughts on this? It's gonna, you know, we're gonna see the, you know, for investors, we're gonna see the telecom industry getting disrupted. Um, yeah, yeah.

And he didn't mention it here. Yeah, to a point on vertical integration, he didn't mention it here, and I think he's talking about working with phone providers is it would not surprise me at all if they either buy or start manufacturing their own phones as well. Um, I don't know, what was your read of that? My read was he was talking about working with phone providers and not talking about building their own phone, but, um, at some point I would be surprised if they don't think about building their own phone.

Well, you know, it's been, it's been rumored— the Tesla phone's been rumored for a long time, uh, and Elon famously does not work well with others. He tends to buy them or blow past them.

So did you see recently the possible acquisition of T-Mobile? Yeah, the rumors are out there.

And also with Charter Communications.

I mean, Starlink is the most unbelievable business. Starlink is going to produce hundreds of billions of dollars of free cash flow in the next sort of 5 to 10 years. Um, you know, they're going to have direct sell, they're going to have unbelievable bandwidth on, uh, almost, you know, So unless you're really in the middle of Manhattan, I think Starlink will be the best option for most people. Yeah, I can't see a world where on Starlink revenues alone, SpaceX isn't the most valuable company in the world on a sort of—

We've been projecting $10 trillion by 2030 in terms of valuation and scaling towards $100 trillion.

I'd absolutely back that.

Dave, what's your thoughts on this?

Well, my question actually is if we have direct satellite phone connectivity downloading videos and there, you know, that implies tens or even hundreds of thousands of satellites in low Earth orbit and then data centers are going to want that same space. So is that actually going to survive the escalating needs of AI? You know, because if you figure, you know, as you said, you know, 5 to 10 years from now, almost all compute is going into space. That's, that's also a lot of very high value use case for that same LEO.

So yeah, you know, they're a slightly different orbit because the, um, the Starlink satellites, they fly around 400, 460 kilometers I think it is now, but they are in a, uh, I think it's a 50-degree inclination, so they, they're not going over the poles. They're mainly going up to just, you know, midway through Canada, and they're flying as low as they can. Um, the— actually, with the AI satellites, you want to fly them as high as you can almost, because even in this dawn-dusk sun synchronous orbit, uh, if you're flying at 600 kilometers altitude, you still have— there's like a month of the year when you have a 10% window which is blacked out. Um, the ideal altitude to fly is actually 1,200 kilometers, um, that's the lowest you can fly where you don't have any black out, um, even throughout the year. Um, so I think all of it— yeah, there's very, uh, between 400 and 500 kilometers is going to get extremely crowded. Um, I think where the AI satellites fly will, will get very crowded, but it'll be for a different use case. I don't think they're going to be competing for the same real estate, essentially, because you're going to run into Kessler, Kessler effect problems if you go that high.

I mean, at scale, you know, we're talking.

Yeah. Uh, the main reason people don't want to fly that high right now is actually radiation from the Van Allen radiation belt. Do you mean because there's less drag? If you have any collision there, the debris is going to stay up for longer.

Yeah, for millions of years. Yeah.

Yeah. Well, even, even at 1,000 kilometers, you're probably— most stuff is going to deorbit within about 50 years. So, which is not great. I mean, but it's not, uh, for example, I don't know if you know about in 1970, the US government, um, dispersed 400 million needles at about 3,000 kilometers altitude because they wanted to bounce radio frequency off them, which is insane to think about that today.

It was like an artificial ionosphere.

Yeah, it was the most insane thing ever. But it responded. Yeah, yeah, you can't imagine it being done today. But actually every single one of those needles is now deorbited, which people don't realize because people think that, um, if you, you know, if you have quite a large satellite, it will stay up in at 3,000 kilometers for quite a long time. But, um, things which are much smaller and don't have their own propulsion, because of the way the sun and moon and all of the solar wind interact together, every so often they get closer to the Earth and they get dragged in and dragged in and dragged The Kessler effect is something we really need to worry about, pay attention to, but it's not as drastic as I think a lot of people sometimes think.

That's good news. Salim, you had a question.

Wait, what about cooling? I mean, cooling and radiation. Oh, sorry, go ahead, Salim. No, no, go ahead.

I was going to ask that exact question. Oh, really?

You? Yeah. Well, we were talking about this. I forget, Philip, we were talking either in Riyadh or at A360. I forget where. But you had this all aluminum cooling and I was like, wow, that'll be incredible if that works. But now it sounds like you're going to move to, uh, to liquid cooling or some kind of a liquid, uh, cycling process.

Yeah. Yeah. Well, it's, it's liquid through aluminum. So the, yeah, that design is, is the same as it. Yeah, it was in Riyadh, I think. Uh, exactly. So it's a very large, low cost and low mass deployable radiator. So radiators we know work because the International Space Station has been it for 20 years. The problem with the ISS radiator is it is both expensive and heavy. And so the core— it's a government project, what else do you expect? Expensive, heavy, and late. Um, so the core IP of our company is making this radiator cheap and light. Um, it's not a new physics problem, it's a manufacturing, you know, an engineering, uh, problem. And so our radiator design— and we've got got it, uh, working is we've just— it's fabricated and gone through TVAC and everything— is our radiator design is 10 times less mass per watt of dissipation than the ISS radiator and about 100 times less cost. Yeah, about 100 times less cost per watt of dissipation than the ISS radiator. So, um, I'm excited. That will— that will fly in January and it will be a big milestone, by far the largest commercial—

I wanted to double-click on something you said. Did you say over time 95% of the compute we use will be done in space?

99.9%, I think, over time. Wow. But that's on Alex's Dyson Swarm timeframe.

That's right. So, Masa-san, so Philip, Masa-san just came out with a statement saying he disagrees with the thesis of orbital compute because energy is only 7% of the cost compared to everything else. Where do you come out on that? What's your answer to him?

Um, I did a post about this this morning because, uh, people kept tweeting at me. So I was like, okay, well, he also sold all of his Nvidia stock in 2019, so he's not always right. Um, no, I mean, he is right that energy is a very small proportion. Energy and infrastructure though is actually quite— well, Well, uh, yeah, I mean, it depends on— right now chip cost is very high, and so that is, um, by far the dominant cost. But, but if you, if you use— include energy and infrastructure, you're talking about at least 30% of the cost. The main problem is if you even, you know, to build a new energy project terrestrially, you're looking at like a 5 to 10 year long lead time on just the permitting of that. Yeah. And so So the main problem is we can deploy this stuff extremely rapidly. So even if we were breakeven on energy and infrastructure, it would still make sense to do this. But we're looking at doing this about 10 times cheaper on both energy and infrastructure. When I say infrastructure, what I mean is we don't need batteries, cooling towers, big chillers, backup power. All we need is a dirt cheap radiator. Our radiator is really dirt cheap., and then a lot of the other infrastructure costs are gone. It's only then the launch cost is the additional piece we have, but that, as I say, is very rapidly trending towards, uh, a much lower launch cost.

All right, Philip, if, if I might, um, yes, please, just pull the thread a little bit on launch. So projecting conservatively 20 to 30 years out, where do you think launch is going to come from? Or will we be using railguns to launch from the lunar surface? Will we have optimistically self-replicating von Neumann probes that are disassembling our solar system to build more compute. Where is all the matter and energy and launch coming from 20 to 30+ years out in your mind?

Um, I am— Elon has this great quote, something like, Optimus is the von Neumann probe. And I kind of agree with him. Like, if you can get 100,000 Optimized to the surface, to the lunar surface, surface and get them to build an Optimus factory on the lunar surface, uh, you know, then we're off to the races. Like, then you have this insane exponential curve in terms of development and pace of development. Hyper exponential. Yeah, yeah, absolutely hyper exponential. Yeah, so, um, I do think we'll have mass drivers on the moon, and I think they'll probably come sooner than most people— well, if it was 20 years, I think that would— that if it was less than 20 years, that would be be maybe surprising to me. If it was more than 20 years, I'd also be a bit surprised. I think, yeah, around the— that time frame.

All right, I apologize, uh, I apologize. I have one of the senior execs from State Street Bank waiting for me outside the door here, and I'm gonna love watching this podcast because all of my questions are in Alex and Salim's heads, but I can't wait to hear your answers. So super excited about what you're doing though. Congratulations.

All right, thanks so Dave, appreciate it.

See you, see you very soon again on our next pod recording. But let's move to our last story on the docket here. Rocket Lab is acquiring Iridium, creating yet another fully integrated, you know, space powerhouse. So Rocket Lab, uh, for those of you who don't know, is a company started by Peter Beck, actually Sir Peter Beck. It's a $64 billion company now. Kudos to them. I mean, going, you know, Peter had no background in launch and he built arguably, uh, the second-tier provider for launch after, after SpaceX. They have their Electron launcher. It's a smaller size launcher, but it's launched 91 times. It's got a very high reliability at this point. They're building their next vehicle called the Neutron. It's about half the size of Falcon 9, and it's planned for a first launch by the end of this year. Like Falcon 9, it's got a first stage which is reusable, and they're acquiring Iridium Iridium. I know Iridium well. I was playing in the Big Leos in the early '90s when it got its license and started. It started commercial service in November of 1997, and it's a 66-satellite constellation orbiting at about 780 kilometers. A fun story, I don't know if you know this, Alex, it was originally called Iridium because it originally had 77 satellites, which is the atomic number atomic number for iridium. When they change it to 66 satellites, they did not change the name to dysprosium, which is atomic number 66. A good marketing move there. I don't think this is—

Also, iridium, I think, is a little bit more stable as a nuclide.

Yeah, and you know, it sounds a lot better than dysprosium. So what makes iridium interesting is it's got 10.5 10.5 MHz of bandwidth at L-band that's globally coordinated. So they go through the ITU and they get, you know, importantly for our viewers here, you can get 10.5 MHz in the US, but can you get in every country around the world? And that's what makes it a prize. So spectrum is the prize and vertical integration here is becoming the winning structure for the new space economy. Company. So owning launch and manufacturing— they build, you know, at Rocket Lab, they build their own satellites as well— and getting spectrum and operations, uh, is a winning combination. So they're playing on the SpaceX, uh, handbook. Uh, pretty, pretty extraordinary. Uh, thoughts for you, Philip?

Yeah, I mean, it's a very smart move, I think. And, um, it's particularly smart because, yeah, I think there's been quite a bit of commentary, people saying, you know, he's not trying to compete directly with SpaceX with this, um, and if he was, it probably wouldn't be as smart a move. Um, he's really carving out a niche. And the other thing is his share price has gone to an insane multiple of his revenue, and he's capitalizing on that because I'm pretty sure all of this is going to be in Rocket Lab stock. Um, and so Yeah, it makes sense to start paying for cash-generative profitable companies in Rocket Lab stock when you're trading at a however many hundred X revenue, uh, multiple that they're trading at.

I have to ask Philip a question just about the spectrum side. There are a bunch of obvious questions I could be asking about vertical integration and does spectrum and, uh, and LEO compute Institute want to inevitably own or be owned by heavy launch capability. But I just want to focus on spectrum. So, so to the extent part of the Iridium Rocket Lab story is the acquisition of RF spectrum, I have to ask you, do you think radio has a future, or will we find ourselves 5 to 10 years from now where it's all optical frequency, direct laser links, and radio has approximately no future?

Um, it's a great question. I, I would lean more towards the second, uh, of those two options. I wouldn't say it really has no, no future. It's useful because it's cheaper, because you don't need gimbaled, you know, uh, things. But, but laser is, is where everything's going. We've got three laser terminals on our second satellite launching in January with, you know, gimbaled lasers. Um, we've actually just signed a contract with SpaceX to— for the next 25 of our satellites, we'll have 2 Starlink, uh, they call it PLASER, plug-and-play laser terminals, um, on our satellite, and then 1 laser, SDA-compliant laser that can connect with the government satellites. So yeah, lasers are the future for spacecoms, and it's also unregulated, which makes it amazing because it's—

All right guys, I apologize, I've got a hard out here as well. Carlo? Philip, pleasure and excited to watch Star Cloud 2 and Star Cloud 3 make it to orbit. Thank you for joining us. Salim, uh, coming home eventually?

I am. I'll be here for a few more days and then I'll be back.

All right. And Alex, how about you? What's your travel schedule looking like? Well, I don't know.

I'd love to visit LEO or sun-synchronous orbits sometime soon. Philip, we should chat!

We're We should. And Peter, I think you're going to be in Paris in a week, so I might see you there.

I think I may be there virtually. Oh no, wait, where am I? I know I'm in— I'm in— I'm in— calling the kettle black. Yeah, Calgary, Germany, and Greece. Yes. Okay. All right, love you guys. Uh, again, thanks, Peter. Thank you. Thank you, Alex. Take care, guys.