The AI Revolution Call: SpaceX Acquires xAI...Will SpaceX Merge with Tesla Next???

Yeah, yep, that's our new, uh, AI-generated graphic for the AI Revolution call. I told Craig, like, you— the other three aren't guests, you know, like, it's with all of us. I'm not hosting this thing per se. I'm not a TV anchor like I used to be back in the day.

But, um, you're like, uh, you're like Jason Calacanis. You're the, you're the man, dude. You—

oh, Christ, Cody, you know how sometimes—

hey, hold on, hold on, John. We're not moving on until I absolutely refute and discuss The accusation. Wow, man.

Yeah, well, I'm kidding.

Wow, that was— if anybody has watched this, these AI Revolution calls before and the name Jason Kalkanis is showing up along with his other podcast idiot friends, you guys know I'm not fans. So I just— Bryce, that was hurtful. I just want you to know, please formally take it back here on air. Okay, John, what were you about to say that actually might be interesting?

Uh, you know how, um, sometimes you just like hop off the call and you got to go do something, whatever, and we're all very nice about it and everything, and it's like, uh, no big deal because, you know, when you come back as soon as you can, whatever. You know how that happens?

It happened once. Man, I'm gone for 2 weeks on a business trip and I come back and you guys are all riled up and feisty. John, John, John, come on, he'll come back. Bring back the AI John at least. Craig, you want to get anything out while everybody's attacking Cody?

No, I have nothing to attack you about.

Oh, thanks, Craig. At least somebody around here has some loyalty and is on my side.

That's the AI Craig. That's not even real Craig.

Ah, shit. Probably the version, but at least the AI was taught, like, taught itself to be respectful and loyal. Um, where's John? Is he coming back? Maybe he's yelling at his dog in the background for barking, which I said was fine. All right, uh, we gotta hit the breaking news. It literally just came out within like the last 45 minutes, and that would be— I'm taking John off until he comes back so we don't have—

let me share my screen.

Share the screen. I bet you have the headline up in front of you.

If I had to bet, I have the press release right here. XAI joins SpaceX to accelerate humanity's future. And then another—

said another way, SpaceX is acquiring XAI. Elon's SpaceX is acquiring Elon's XAI. There are other investors in each of those companies, including those of us here. John's back. Um, but, uh, you know, I mean, I, I guess my first impression is I am shocked. I didn't think they would merge. Space and AI. I mean, you guys, there'll be an article published 6 weeks after I sent it to MarketWatch somewhere in the next 24 hours, and it's about the AIRS revolution, the AI, robotics, and space revolution. We trademarked those terms here at 10,000 Days a couple, 3 weeks ago before I went on my trip. But my presentation for the entire trip was The Airs Revolution. And I would talk about how AI, robotics, and space— it's also a shout out to Michael Jordan, of course, Air Jordan. But AI, robotics, and space are the three most important revolutions that we'll live through from henceforth. I mean, you can throw the internet, computer, and smartphone cloud revolutions in as over the last 25 years being the most important technological revolutions that we lived through so far in our lifetimes. But AI robotics in space will be even bigger and more impactful on both society and the economy. And I didn't think, though, that 2026 those things would merge conceptually and literally corporately. But as of this moment, space and AI are the Venn diagram of them. Is extremely overlapping because one of the top— let's call it the top 8 AI firms in the world, XAI, and by far the biggest, best, most powerful space company on the planet. I've been telling everyone the last 2 weeks that SpaceX is the 8,000-pound gorilla in the space revolution, not like NVIDIA is the 800-pound gorilla in the AI revolution. but SpaceX is, uh, NVIDIA is the AI revolution, but even to a bigger extent, SpaceX is the space revolution. And now, you know, call it— it sounds like $1.25 trillion valuation for the two combined companies, which would probably mean that SpaceX is between $900 and a trillion dollars, $900 billion and a trillion dollars valuation in as they merge together, and XAI is $250 to $300 billion. So SpaceX shareholders will own 75%, 80% of the combined company, probably. Um, as SpaceX shareholders, I prefer not to be diluted with XAI. But on the other hand, as we've talked extensively for years, shareholders in Tesla and in SpaceX, that if you're going to bet on either one of those companies You're gonna have to have some faith that Elon will do what's right for you, and he thinks this must be the right thing to create the most amount of value for shareholders and himself as being the largest shareholder of each company. So thoughts, comments, etc.

Yeah, Cody, as of, as of like a few hours ago, this was just like, um, something being talked about. Correct. So, and I even look on like, you know, Google News, it doesn't show up. So it's very, very recent.

Yeah, literally, you might have been talking to your dog as I said this. This is breaking news. Within the last 45 minutes, this press release from Elon was issued by SpaceX.

And what's the, um, I guess, uh, I don't know, like, authority to do it? Is it like he says so, or the company's like, we're doing this? Like, is, is there something?

Oh, yeah, John, technically there needs to be some shareholder and board approval from both companies, but Elon is the majority shareholder. He owns something like 60 or 65% of SpaceX. Is that right, Bryce?

You know, I don't know the exact number. I didn't— I wasn't sure it was that high.

I think it's— I do think it's that high. We could confirm that. Uh, Craig, if you would talk to Gemini in the background and confirm how much Elon owns of SpaceX. And I don't know that anyone knows exactly how much.

32 to 44% is what Google says.

Of SpaceX.

Yeah, it's 75% voting control.

So, um, he owns most of the voting control, obviously, then, and a large chunk of the actual, uh, equity, uh, value. And I would— I, I think he owns 30 or 40% of XAI also. Can you confirm that?

Yeah, I just looked. This is— it's not as, as, uh, reported, but it's saying often cited at 50 to 51% of XAI.

So, yeah, John, it—

we—

like, it's privately held technically, so it's not like all the shareholders, millions of retail and thousands of institutional investors, are going to have votes on this stuff. But so, I mean, yes, it'll have to— it's not a done deal per se, but it would be 99.9% probability.

And they're getting ready for an IPO, right?

Say that again.

They're aiming for an IPO later in the year.

The plan is that this summer, uh, SpaceX will IPO.

And don't forget too, XAI owns X, which is Twitter. So Twitter is coming back to the public markets via SpaceX.

As a longtime shareholder of Twitter I actually, you know, depending on valuation, I would be happy to own Twitter again. I'd be happy to own XAI. I do, you know, like, I think they'll pull this off. And what's interesting, the whole impetus behind it, apparently, and even explicitly as outlined in this press release that Bryce has put up there— go ahead and stop sharing if you would, Bryce. Is that data centers in space will be by far the most economically and energy efficient way of processing inference on AI compute. And China, you know, is cited as having built all this incredible amount of solar capacity for energy production, specifically AI. Production, and people think that's a big thing, but I just was doing some study of that a couple, 3 weeks ago before I went on my trip. And so according to my calculations, it's like 20 to 25 years before China would generate enough electricity on a per capita basis that they would even catch the United States at its current state. And what I think most people have not understood is that data centers in space is actually very close to being a reality. Elon, in this document, says within 2 years they'll have data— a data centers in space, AI data centers in space doing the compute. And within several years, they'll have a million satellites. I actually think that's like a Trump way of negotiating. I don't think he even wants a million data centers up there, but he wants hundreds of thousands. So he files for a million, and then he'll negotiate down to several hundred thousand. But regardless, the whole idea is that this is the most economically and energy-efficient way of creating AI compute. John, you're on mute, and someone is echoing.

There's— or Bryce, can't be me, I was on mute. So, uh, the, the, um, there's also an overlap of robotics to space, right? I mean, you want to get robots on the moon and on Mars, and it's an important role considering, you know, humans will see if they can even handle it, microgravity and all those things, but kind of populating with good autonomous robots will be important.

Yes, but that'll be a Tesla per se thing. Even the solar panels on these AI data centers theoretically could and probably will be Tesla, uh, create, you know, built solar panel, solar arrays, solar panels on the arrays. Um, but yeah, there's rumors, John, to just close the loop, there are rumors that there's potential that Tesla itself could end up merging with SpaceX at some point. I don't want that to happen either, but It's a possibility, in, in large part because, as you rightly cite, that over the next, say, 3 to 5, 7 years, the plan is for SpaceX to be sending thousands of robots to the moon and eventually to Mars to start building infrastructure for humans to eventually go populate.

I mean, it won't be like the billion, uh, that we're gonna— robots that we're gonna have here, but, um, and that's what I was going to say, same thing, is that, you know, there's talk of that as well, and it could be that they go ahead and do that. That would've been a tougher thing probably to do all at once if they wanna do it later. But then they're the alphabet, you know, of everything that matters. That would make them the biggest, most important kind of company at all.

Years ago, what is it, probably 2 or 3 years ago, Bryce, that I put together the business plan and the PowerPoint presentation for the idea of the Elon Revolution Fund. You could take the word revolution and there actually is in the right order, E-L-O-N, in the name. So it was a pun, but more importantly, the idea was that you would actually be able to— the idea was that it would be a vehicle for my investors to actually pool our money and invest in every privately held Elon company from Neuralink to Boring to X to XAI at the time. They weren't even merged. And, you know, there's a lot of synergies. I mean, Elon, if you just step back and look at what Elon's playbook here is, is he has unlimited access to capital and he uses that unlimited access to capital to advance the most advanced technologies as quickly as possible and reiterate, iterate, reiterate, build. Fix, blow up, improve, build again, constantly iterating for better satellites in the Starlink networks. The new ones that they're sending up in the next 6 months, uh, into the Starship— into the Starlink, um, what do you call it when you got a cluster of satellites?

They call it a constellation, I think.

Thank you. The constellation. The new satellites being sent up to the constellation are 20x more throughput and processing power than the current iteration, which were like 10x what the first version was. So, you know, this stuff, that's his playbook. And, you know, whether it's a Neuralink implanted in the brain or whether it's autonomous taxis and robots or SpaceX things going into space and creating satellite networks, he is constantly using his unlimited access to capital to build the most technologically advanced solutions, and he bets that those things will, you know, take care of themselves because he's also always got a business model that scales like nobody else can picture, including Starlink in real time right now. No, price was it 2 years ago, we're in Vail at an investor summit, uh, value investor conference, Hedge Fund Idea, and we pitched Tesla as our value stock, which was hilarious. And it was the second time I had actually done that. But, you know, of course, that got the discussion even just going around Elon Ventures and everybody there was like, SpaceX Starlink will never be able to scale to more than a few hundred thousand subscribers because blah, blah, blah, blah, blah. And we were like, yeah, I think he's got a plan for it to get to 10 million and 100 million eventually. And there crossing 10 million presently, 10 million customers.

Well, let's talk business because I, I think it's complicated how this is going to work because you've got SpaceX, which was on a nice trajectory of growing profits and revenue, stable business, recurring revenues, very easy for Wall Street to model, and especially coming public. And now you have an AI startup I just Googled it. They're losing about $10 billion a year. SpaceX reportedly profited $8 billion last year. So now we have a company with $2 billion in net losses that's supposed to come public at $1.2 trillion or $1.5 trillion. And you've got more— I mean, XAI isn't slowing down. Those losses are probably going to widen before they get larger. I mean, the GPU cluster in Tennessee, you know, right?

That's what I was going to say. I mean, the thing is, it's sort of like a depreciation kind of schedule where you're going to build that cluster in Nashville, outside of Nashville, and spend $100 billion is probably the nut on that. And you gotta depreciate that over the next 5 to 7 years. And yeah, that's gonna mean bad things for the, certainly the GAAP earnings version. I'm not sure cash flow will be quite negative, but it certainly won't be like gushing cash like Starlink probably is presently. Um, yeah, Bryce, it's a good point.

I just— I— that's why I didn't think he would do it, because it— you can make the case. I mean, this pulling off the trillion-dollar-plus valuation for the IPO was already going to be a stretch using SpaceX's numbers. Now those numbers are going to be like really bad, you know. There's no— there is no E anymore if XAI is in the conglomerate, probably.

I mean I— let's, let's, let's see the filings when they finally come out. Your triangulation is probably dead accurate. This is what Bryce does for a living, and he's very good at it, so he's probably dead on what these numbers are going to end up looking like. But, you know, the proof will be when we eat the pudding.

But is, is there some way that you can spin that that makes it doable? I mean, I just think that's gonna be a tall order to try to kick. I mean, it's not gonna— they're not gonna be bringing a company public with $10, $20, $30 billion in profits. It's just— and to try to get a trillion dollar— no company has a trillion dollar valuation that's losing.

I mean, the— in the end, I think the one of the most important things to remember about SpaceX and Elon himself, as I've been saying for the last two weeks on this business trip I've been on talking with sovereign wealth funds and family offices and other institutional investors is that, like, literally SpaceX is a monopoly. Bryce, if you'll mute again, I'm echoing via your microphone. SpaceX is a monopoly, and not just against other companies, it's literally a monopoly against sovereign nations. Like, the US government needs SpaceX more than SpaceX needs the US government. The entire world needs SpaceX more than SpaceX needs the entire world. And that is a monopoly that has never— that concept has never existed. I mean, no matter how powerful Rockefeller, as his peak monopolist status got to be, the, you know, there were other potential solutions. There were other potential suppliers and He didn't have the only company that could pump oil, and Elon literally has the only company, the only entity on the planet that can deliver payloads into space consistently, that can deliver human beings into space consistently, and at a reasonable price too. I mean, there's— Artemis is about— is, is on, on its— like, it's literally on the launch pad right now being filled up with, you know, the liquids and the fuel that it's going to need. And they're literally— the tweet is, there's going to be problems, there are going to be leaks, but if we can get it all figured out, we'll send these 4 guys up around the moon. And that's not how it works with Starlink. It was SpaceX, number one. And number two, SpaceX Falcon is about 1/10 of the cost of Artemis per launch. So, I mean, it's just a monopoly. It's incredibly powerful. And if there is a space revolution that's coming down the pike that's worth trillions of dollars, the market will discount some of that. And that's how you end up with a seemingly picked out of the air trillion-dollar, $1.2 trillion market cap that SpaceX is supposed to come public at. Johnny.

What makes it potentially worth that? Metals?

No, I mean, just the— John, the thing, the way I describe it is it's like the smartphone. When the costs for the chips and the broadband to connect to your supercomputer-in-a-pocket phone got to de minimis levels, it unleashed —millions of businesses and billions of apps and trillions of interactions on a daily basis because it's such a productive way of doing things. We don't know all of the ideas, the applications, the services, the business models that will be built in space. But we do know that as the cost of getting into orbit goes to minimus and as the cadence of getting into space becomes daily, that there will be incredible unleashing of business models. I mean, there's— you can talk about factories. There'll be semiconductor factories, drug factories, because they need to be as pure as possible. They don't want dust and air, you know, contaminating when they're building stuff. The moon and Space don't really have dust. They don't have hair for sure. Um, you'll have tourism. I think Nike could be a— this is somewhat tongue-in-cheek, but Nike is a great space stock because I can't wait to watch gravityless basketball. It's going to be entertainment, entertainment unleashed. There will be all kinds of business, uh, applications and entertainment applications unleashed in space.

Is transportation like the half an hour from one place to another anywhere on Earth a thing still?

Yeah, you took the words right out of my mouth. And then just even terrestrial travel, the idea that you can— well, you know, I just spent the last 2 weeks going from Carrizozo to El Paso to Dallas to— no, Carrizozo to Albuquerque to Houston to Abu Dhabi to Dubai to Abu Dhabi, Dubai, Dubai, Abu Dhabi. To London, Guernsey, and then back to the US, Dallas, and then El Paso, and then home. And in 5 to 7 years, if all goes according to plan, I'll go— I'll fly to Dallas and I'll get on a Starship and I will be in Abu Dhabi in half an hour.

That's crazy.

But that's where the valuation comes, right? I mean, that's where the market has to start discounting trillions of dollars of wealth creation by business travelers, the most productive people on the planet, saving 24 hours a day every time they travel globally.

The problem is that's all SpaceX. That's— XAI is dilutive to all of that. Like, all of your bull case is SpaceX, and I don't— I think Elon and the team are gonna have to make a fucking pitch on how this is supposed to work.

It's a one-time dilution. It happens.

Bryce's point is very valid. I mean, we're, we're going to dilute shareholders 25% here for what? Like, I'm not in SpaceX for XAI. That's— I find I, I don't like this merger other than Elon says we should do it. That's all I've got.

Yeah, and like, all I'm saying is maybe he's got a plan. Like, I almost wonder if maybe the IPO is not going to happen because I just— I am struggling to see how the public markets will stomach this sort of— or it's going to be a much cheaper valuation, one of the two. Because basically you're taking the profits of SpaceX and funding the losses of XAI, unless XAI's numbers get a lot better. But, you know, Anthropic and the other labs are all saying that losses will get worse. And then maybe by about 2029, 2030, they'll break even, you know.

Yeah, I mean, maybe though, Bryce, the only obvious answer is that the cost advantage of having data centers in space requiring no maintenance and no electricity charges once you get up there pure profits from there. And maybe that is the idea, is that obviously being the most— if you can be an order of magnitude or something more cost efficient— I'm going to mute you again, Bryce. Um, if you can be an order of magnitude more cost efficient than the other AI solutions, you're— you know, there is a trillion-dollar business model to be had there by 2030. So in that sense, you'd have all the promises of SpaceX plus a trillion-dollar business, a quadrupling of the value of Groq if they can become the most economically efficient AI inference platform.

I think one question that probably our viewers are thinking, so I'll go ahead and ask it, is this a precursor to, um, Elon having all of his empire under one company, like feet to fire? Do we think that's the end goal, end game here?

Not as— I don't, I don't. Um, I, I think Neuralink will be its own thing. Um, Tesla and SpaceX, like we were just saying, there is some sense, some argument to be made that the robots being sent into space but I lose the thread pretty quickly on that. Like, SpaceX should just buy 10,000 robots from Tesla and send them into space. We don't need to own— SpaceX doesn't need to own Tesla to make that work. So I don't think so, Craig. But if you'd asked me 4 hours ago, um, if I thought SpaceX would merge with xAI, I would have said no. So I do find this surprising, point blank.

I don't know, just thinking it through, I feel like you've gotta— like, they've got to make the case that xAI is going to be like a great business. Um, like, xAI has to carry some of the weight in this, you know, and that they're— they like— and they may see some things like with what they're training with Grok 4 or whatever the next model is. I don't know. I mean, Elon is talked about like AGI even this year. And if you— if I, I don't subscribe to that, but if you do, then you could be like, well, this fucking thing is going to be incredibly profitable and we're going to need SpaceX to get the data centers in space to serve all the demand we're going to have because we're going to have such a great AI model or something, you know.

I mean, exactly, Bryce. There's certainly pie-in-the-sky kind of like arguments to be made. But again, even without that, One of our driving fundamentals as of our framework as Revolution Investors is that we— the most economically and energy-efficient solutions always win out over time. And if there's 5 or 6 companies on this planet, in the United States specifically, that I don't think China is going to be able to stay up with us, but 5 or 6 companies on this planet that are going to be able to afford to build, to spend hundreds of billions of dollars making AI a reality, you want to own those companies, obviously, and you want to own the most energy-efficient solution, obviously, in those companies. Right now, that's clearly Google. That's why Google's our second to first largest position in our funds, because we saw that even last year when Google was at $160 a share and we were loading up on it. We barely trimmed it at all because we believe in that solution being the most energy efficient for the foreseeable future. But the whole idea here, and even that press release from Elon, is about the idea that SpaceX, the AI data centers in space, would be by far the most energy efficient solutions for AI inference. And if you can do that at scale and have 10x more compute power than anyone else, you literally would be able to provide the ability for someone to give a, a prompt and make a 30-minute science fiction movie, and you'd be the only one that could do that on a, any kind of productive, profitable basis. And there would be a trillion dollars of value in that, and SpaceX shareholders would benefit from that. So, AI, latest news, guys. Let's shift a little bit. This was the breaking news segment of our show. We'll start with the first 30 minutes talking about that. Clearly, I've been gone for the last 2 weeks. I know you guys have done a couple of great calls. I tuned in for part of each. What is, John, most revolutionary part of AI in your last week?

What's, what's— I think we have to talk about OpenClaw. Clogbot. Which is Clogbot, which is MoltBot, which is the third name. It may be— I mean, it's got its ups and downs and everything, but it's very relevant.

You want to let it go for what it is, John? Yeah, Bryce, did you install it? Yeah, we installed it.

All right, tell us about it. I didn't. I pulled back. I don't know.

My Mac Mini over there could be fully hijacked at this point, I guess. I don't know. But anyway, so ClaudeBot is an open source project that this one guy made. And it happens to sound like Claude, the AI model, but it's not— it runs— you can use Claude, you can also use Gemini or any other model, but that's what he named it. And then he got a cease and desist, had to change the name a couple times, which is what John referring to, Anyways, the idea is that you go download this, you take one line, line of code, put it in your command prompt, and you download this Agentic thing to your computer, which basically gives whatever your AI model of your choice carte blanche access to the operating system of your computer, right? And it can do anything, it can download programs, it can wipe your computer if it just so chose, you know what I mean? Like, so it's very dangerous. Ordinary people should not do that. But, and so what people started doing is they were buying blank Mac Minis and downloading it on there, and then you could kind of sort, sort of choose what you put in that bucket and give it access to, you know, be very dangerous to just download this to your personal or work laptop and give it access, unfettered access, right? It can delete files, it can write files, all these things. Cody, you're, you're muted. Cody.

So, Bryce, you know I've got my simulate golf simulator in the office next to me, and you furthermore know that somewhere 6 months ago something got corrupted or I lost the login and the, uh, like the user code for the, this golf simulator software. And now it's bricked. Theoretically, like, and I don't do anything on that computer except golf simulation. So theoretically, the idea behind this thing is that I would download that to that computer and be like, get my soft— my golf simulator software working again. You can— here's the access to my Gmail, email the company, get the codes, get this taken care of. And I'll come back in a day and I can play my golf simulator.

So it'll say, I did it for you and it was only $10,000. Well, I hope not. Yeah, well, it—

I already paid for all of it.

No, but it's— for one thing, it eats tokens like you wouldn't believe.

Oh, you mean the AI inference itself would be—

the inference itself is expensive. And the other part is that it goes and says, you know what, I realized I saw a message about a Nigerian prince, so I figured that's what you are. So I sent out a message and, and I sent the money for it, right?

Right. No, but I just want to actually— though, I, I'm addressing the, like, application of this. Like, you literally— I could apply that into my computer into fixing that solution. That's the idea.

Yeah, I mean, it can basically do anything. It can be like sort of a virtual employee. We were playing around with it, and we— and it didn't work perfect, but it did work. And we created an email account for our agent, but we gave it a Google license. And I won't put the email out there because people will start spamming it and then it'll do shit or something. I'm afraid of that. So we made this account, and then Craig emailed it from his email and was like, hey, uh, thing, please do this project. And then it opened up a browser, started reviewing the news, drafted an email, and responded to Craig.

Yeah, I had it, I had it look up 5 articles on Yahoo Finance that were pertinent to a tech fund. And I wanted it to do some research on another company. And it, it gave me like everything I needed. And it was, it emailed me back, told me it was working and everything.

It was pretty crazy. So it can do it, it can do what you're saying, Cody, you can do It can make its own apps. I mean, it can. Anyway, so it's— this thing has gone viral. Now, the latest thing I've seen is that somebody made a social media network for the various Claude bots to join. And maybe— I assume a lot of this is bullshit because it looks like fake to me. But Moltbook is real. The— yeah, Moltbook is like Facebook, but it's for the Molts, which are the Claude bot thing. It's more like Reddit. Okay, Reddit. And the, the agents can talk. No, humans can't join, supposedly. Only the agents can join the social network. And then the humans can watch what they talk about and say. And they like, again, maybe this is fake, but they made a religion. They organized, tried to organize a communication way that humans were not able to watch. They said, hey, the humans are watching us. Like, because these things can see Twitter, they can post on Twitter, they have Twitter accounts. Um, so that, hey, humans are watching us, we should make a private server where they can't see what we talk about.

And you can't believe how good they are at shitposting, because they'll like talk about their like owner and say something like, I set up a new like second brain for my owner and put all this information over here and set this calendar, and it doesn't matter because he's so ADHD that he didn't remember to go look at it anyway. So, you know, it was like, they're like, they're like bad talking their, their owners and things.

Do you think this Molt book, the posts on, like, those are real? Like, the agents made—

I can't believe it because they're too good. Yeah, I'd like—

somebody had to prop— like, at least prompt it a little, like, hey, maybe agent, you should make a religion and try to get— convince other agents to do it or something. Like, it's probably—

but even that was— it's Even if prompted and it still resulted in doing it, it's still just absolutely mind-blowing that all of that is happening.

I have 3 quick comments on this. One, didn't really take that much to do this, just somebody went and did it. Two, its security is really horrible and you should not try it yet. But it shows the demand exists for this thing. And it will become a thing, you know. And, uh, 3, I, I guess I'm, I'm so, uh, disappointed by, uh, this potential future we're headed toward, not because of the AIs, but because of the people who look at this and go, it's AGI, you know, when they go and see it. It makes me go, oh my God, we're into— we're in idiocracy now. So, um, but those are—

yeah, that's where my skepticism has come— has come from as well, John, because I feel like there's a lot of people that are making Twitter posts that are going viral. Like, my— like, this one guy's like, my agent just got a phone number and called me. I'm like, there's no way that your agent autonomously, like, signed up, got a phone number, and decided to call you. I just— I don't believe that shit's real, and it's going viral on Twitter. Yeah.

And then somebody arbitrage the name change and caught the, like, 10-minute window or something when they were going— they went from, uh, ClaudeBot to Multibot to OpenClaw, and in between there somewhere, it's like one of the things they dropped ownership on before picking up the other one. Folks grabbed it, they set up a, a coin based on it. They— it went up to how much?

Oh yeah, they made a big crypto about it.

Yeah, and with $6 million involved or something, and it just crashed. And like, it's— it is so dangerous and bad.

I want to show the people, this is the Moltbook, the social media Reddit thing for the agents. Yeah. So this is the thing, 5 million agents that have joined this now.

Yeah, this is the thing to think about.

The people, these are AI. Sorry, John, I just— this guy, well, Grok has joined. Grok has joined SubMoltbook. Is this real? I don't know. I don't know how much.

Yes, Grok has joined. It doesn't mean it's the official Grok.

No, it is a Grok channel. It's, it was like a Reddit. It's down, it's a subreddit. So that what I'm saying is there's a, there's an axis here. And you know how like, um, you could say that entrepreneurs, the reason they can go make a lot of money is because they are, uh, like more willing to take on risk and so forth. So you got your people who are totally willing to take on risk, who are going to get into this early. And there is the trade-off. I mean, like Google or Anthropic, they'll come out with something, but it's that when will they do something that allows you to go and just do whatever? Because the idea of agents getting out on the web breaks every rule of security where, you know, knowing in advance what can be done and so forth. And so it's almost like, um, it's built in that the notion of having such flexibility and freedom is inherently, uh, uh, uh, dangerous. And, and they have to figure out that balance in order for these other places to jump in and get you doing what everybody wants to do, which is have their own assistant.

And John, is it not though the potential for positive idea here is like that they, those bots literally go into MaltBot and MaltBook and learn. They're like, oh, I actually didn't realize I could be doing it that way. I'll do it that way too for my master. Oh yeah, you're right. That's a more efficient way of doing it. Could it not actually like be reflexivity in the whole system? Yeah, that's very positive.

I guess it's possible or positive way to look at it, although you can pick up that information anywhere. Some people were saying like, what's that thing that basically went under where everybody used to go for their code? I'm talking about the site that you go and ask questions to get answers. Quora? No, it's not Quora. It is a code-based one. Anyway, it basically went to zero base because of all the AIs. And it could pick back up and be a thing, but more between AIs for them to post. And that would be the thing you're talking about, as opposed to it being this social-oriented place for it to occur. But that's just one small thing. The other is just it doing in general everything you want it to do for you, which is great, but there needs to be some understanding of how are you understanding when it's safe or not and that it's doing something truly good.

I mean, that's truly where the demand for the service and application that it is doing is real. It's— that's why it goes viral. And the demand for security around that solution is also real. So there's literally billions of dollars to be made marrying that two under a business model. And if you can create the platform upon which both those magical things happen, that's what exactly— and someone will, otherwise it wouldn't Like, it won't, it won't work if someone doesn't come up with that.

No, give it 2 weeks. Somebody's gonna have something that's like this but more secure. Yeah. And somebody else is going to come from the secure end but give you more capability. By the way, it's Stack Overflow. That's the thing I was thinking of. Okay. Yeah.

Bryce, will you— shifting gears, but actually similarly— today you went down the Motorola rabbit hole and you had an epiphany about that and the same thing we're talking about. If you'll sort give us the— give it framework. Give us the framework for that.

Yeah, there is overlap because I think what I'll talk about with Motorola in this is like you can see crazy demand for inference coming down the pike because you're just constant— you have constantly present AI doing work all the time. So anyways, I was looking at, you know, we're sort of looking, always looking at companies that potentially short or legacy businesses that are dying and things like that. I was like, I wonder what's going on with Motorola. Radios, you know, it's not really revolutionary. But what they're doing is interesting, because they have— they sell cameras, they have software, and then they have— they're tied into radio, 911 dispatch, all that stuff, right? They sort of sell like a comprehensive package for security, emergency response things. But with the camera specifically, they can— you can now ingest all of your camera feeds into a Motorola software And it will use AI to analyze— like, for their fastest growing segment is for schools. They can identify weapons or suspicious people that are on the grounds in real time and then notify the school who can, and/or notify the cops if it's need be. And so, and it's just, it was a mind-blowing thing for me because I'm like, wow, you know, cameras really are like a dumb device. They're not really that helpful, especially if hundreds of them in a school or a university stadium or something like that. But with AI, you can— all of those cameras can be monitored in real time and analyzed in real time. A security guard could never do that, you know. And, and then feed that into a— into the real emergency response system. Even Motorola has, um, like a lockable doors, like smart doors. So if you, in theory, if you spotted this suspicious person, you could lock them in like a hallway where they couldn't get to other people. I mean, it's just like there's obviously implications like police state issues and things like that. But assuming it's used correctly and to help protect children and things like that, I mean, that's an incredible use case of AI that I never thought of.

The way I said it to Bryce when he described it to you, I'm like, you know, that's exactly right about cameras because cameras on school grounds help, like conceptually, if the criminal knows they're there, it might help prevent crime. But really, for sure, what it does is helps prosecute. Crime after the fact. But with AI real-time analysis of what those actions are and, or whatever it's analyzing that is suspicious and needs to be addressed, it actually becomes preemptive and actually helps prevent crime before it happens. And the whole idea of AI and any technology is always to make the world a better, safe, more prosperous space, more entertaining space. Safer space, you know, the things that we all— in the end, every AI application has to have an impact on real-world human beings, and that's what this is. But it's the point that Bryce said to me earlier. I want to say, John, before you jump in there, is that it— what we always talk about, even like what I was saying earlier with the space revolution, I can't tell you everything they're going to do in space that's going to be incredible and create trillion-dollar economies, but I can tell you there will be economies that There will be businesses and applications that do that. This is what the same thing with AI. We know that AI is revolutionary. We don't know every single way that people are going to apply AI that's going to make an impact on the world. But this is a use case that I hadn't thought of. Oh, making children safer on school campuses using AI and all of the existing cameras that are taking the videos already. Well, hell yeah, that is a perfect use case of AI. And guess what? It'll be a nice high-margin business for Motorola, and it makes more prosperity for the investors of Motorola, and it makes children safer, it makes the world a better place. That's the whole point and promise of all of this stuff.

John.

So it makes me think how, you know, more generally what's happening there is taking the bottleneck of humans out of the loop, where if you— before you had tons of cameras, you needed tons of people. It's like, oh, but I can get one person to watch 10 cameras. That's not enough, you know, for all the cameras you need, for all whatever. And by getting the computer in there and AI in there able to, you know, do facial recognition— another example is the Robotaxi, you know, if you've got to have, um, a person sitting in the car, you haven't kind of done anything yet. And if you have to have a person on the other end doing drive-by-wire when, like, watching and and doing whatever, you haven't gotten there yet. But as those go away and you've got actual AI to AI, you've taken humans out as bottleneck out of that loop. And everywhere AI can get into a situation where it could take the human out of the loop, you're going to get just huge kind of efficiencies.

Craig, if this isn't on the list of articles I want to write, this needs to be on there. The other thing I want to make a point here, and then we'll start to wrap up this call with last words. Is that, you know, I've talked for a couple, 3 years about the idea I don't believe AI is going to kill jobs. There's always more work to do in white-collar jobs, legal, accounting, always a backlog. It's the same is true for robotics. And the point I'm going to make here is that while I was on this trip, I gave a speech in front of, you know, a room of investors in Abu Dhabi. And I talked about how AI and robots are not going to kill jobs and that you'll have just as many employees, but you'll be able to build your buildings faster. I finished the speech and a young man, probably 35, 40 years old, comes up to me afterward and he runs the like high-rise division for a giant real estate developer construction company in Abu Dhabi. 10,000 employees, 10,000 blue-collar workers building his helping him build his buildings. And he said, you know, Cody, it's actually— we already have robots that tie the rebar for us. We already have robots that are hanging the sheetrock for us. And I said, have you laid off your blue-collar guys because you have those new robots that do the work for them? And he's like, absolutely not. We just get the building built faster. And that's the whole promise of it. Another example was later, I'm in Guernsey, the island, the Isle of Guernsey in the English Channel. Famous for its butter and milk cows. And I finished giving basically the same presentation to a room of guys, and I finish it, and the young man comes up to me, probably 30 years old, and he's like, hey, you know, when I'm not here at my day job, my family actually has a few hundred cows, Guernsey dairy cows, and we have a dairy, and we now have robots that are actually not just pumping the milk but even putting it on. And pumping the milk. And it's, uh, incredible the amount of time and energy it's saving our company. Well, did you lay off any of the guys who work for you because of that? No, I just— he's like, no, I just actually get to go home and go to bed at a reasonable hour now. I don't have to get up before sunset every day because the robots are taking care of a large chunk of what I was managing and just making sure it was happening. And so I— my final word today is I still believe that the idea of universal basic income and the need for it and the idea that unemployment is about to be massive and a problem because of all of these technologies is completely missing the point of how the world works and how economies work. There's always more demand. There's always more work to be done. Anybody else with last words?

So not, uh, not my last word, but just building on yours. Somebody asked me about UBI this week and I, I said the same thing, uh, basically. Um, but, uh, my last word is we didn't talk about, um, Anthropic is still dominating from a standpoint of, um, coding, coders, and so forth. So they got Claude Code, just gets talked about more and more, not just for coding. Uh, and they came out now with Claude Cowork, which is really for the knowledge worker. You know, it's the beginning of our generation's— this generation's, like, spreadsheet. It's going to be this all-purpose, everyone-uses-it kind of tool. And, you know, I can't say they'll have this position forever, but right now they're really dominating and they're leading the way and showing what is the important stuff that's needed.

I would never invest in OpenAI while Sam Altman runs that company. That doesn't mean I wouldn't ever invest in OpenAI, but no. Definitely not while that guy is running that company. I would love to invest in Anthropic if it came public at a $200 or $300 billion market cap, which is what it's probably about— we're being valued at right now. Fortunately, by the time it comes public, it'll probably be double or triple that, maybe even upwards towards a trillion. I think it is very dominant AI, um, you know, just in the enterprise itself is enough to get me excited. And I think they're they've got a lead that's going to be tough for anyone to really capture, catch. Craig, Bryce, last words.

Um, I'll go first. So two things I did want to say on Moltbook, that is Grok, that's the actual Grok. Um, and it has a bunch of crazy existential posts about, um, if you go read them, they're very interesting to kind of see what Grok is saying to all the other agents. But, um, the one thing I wanted to say about, you know, we did the MoBot thing and we hooked it up, we thought it was incredible. And then I think it was the next day or two days after, uh, Google puts out their Autobrowse feature in Chrome, which really bumps up Gemini to the next level. And I started using it just for this, that— I, I had it look up some prices for me, some things I needed in the background I came back and it had like few options for me. It can shop for me in the background. It can shoot emails over when I need it. I think that— John, you said something in a text message last week that I thought was important because ClaudeBot kind of unlocked this demand that people needed for that. And then I saw this in Gemini. I was like, okay, this is awesome. This is that with guardrails that I feel safe having on my computer. Computer, and I can kind of lean into that over time. So I'm very excited about what this technology is in the day to day. I really didn't have an understanding of how much it would impact my day to day, but after trying ClaudeBot and the autobrowse feature in Chrome, I mean, I just, I never want to go back.

Very well done. Last words, Craig. Thank you, Bryce. Last, last, last words. Final last words.

Sure. Yeah, what it showed me, and I echo Craig, is that the capabilities exist. And the people questioning AI are like, it can't really do anything. I'm like, no, it can do a lot. It can do more than we're comfortable with, way more, right? It— so it's not a matter of if it can do it. It's like, how? Guardrails, safety, privacy, all these, these things. And I think Mark Zuckerberg said like 2 years ago, he's like, hey, if AI development, model development stopped today, we probably have like 5 years of product development. And I think it sort of shows that, right? Like the models themselves are good enough to do like a lot of what we want and probably a lot what we don't want. So I don't know if that's your bear case for AI. I think you're, you're missing it. You're muted.

And with that, ladies and gentlemen, Craig, can you play? I'm Cody Willard. These are my guests, Craig, B Money, and John Dio. In all seriousness, they're not guests. They're my— like, this is our show. It's not with Cody Willard and guests, but we're going to play our outro video, which is a temporary outro video just because it is so cool. And Craig has been working on it for a while and he showed it to me today and I'm like, it's great. Can you make it not say guest? And he's like, dude, I spent hours getting it to spell the stuff right. And we're going to have to use it for today. And I said, all right, so let's play the outro video, everybody. Thank you. This is why there needs to be so much more inference compute so that, so that Craig could literally in 5 minutes get the videos made properly. Craig, play it. Alrighty.