Transcript of AI: Boom or Doom? | EYES ON GEOPOLITICS

Look, I'd rather live in the United States where, like, we have control over the technology and information that affects Americans. And when you have tech like that, used to be fine for basically a century because America made the best technology and most of the technology. Now we're entering a world where China makes a lot of technology and in some cases may make better technology. But they don't want to abide by our use, our, our, our, our Constitution, our rules around this. So I would just say, fine, like, you can do whatever you want in your country, but we are not going to use your technology then.

Hey everybody, welcome to another episode of Eyes on Geopolitics. I'm here today with Jonathan Hackett, and our guest is Anthony Vinci. He's the author of The Fourth Intelligence Revolution: The Future of Espionage and the Battle to Save America. He's got a pretty amazing resume. He previously served as the first CTO and Associate Director of Capabilities at the National Geospatial-Intelligence Agency, following earlier work as an intelligence officer in Iraq, Africa, and Asia. You could actually watch an episode with Anthony on the Team House that we did probably a good 6 months ago now. Uh, great episode. Check that out. I'll put a link in the description. You can check it out. I also put a link in the description for his book. Uh, he is now the co-founder and CEO of Vico, a, an AI company focused on decision-making and finance and national security, and is an adjunct senior fellow at the Center for New American Security. And to boot, he holds his PhD in international relations from the London School of Economics. So I try to get somebody here that's on the level of smart as Jonathan Hackett is, because Jonathan's got a pretty good resume too.

And I think I did a pretty good job with you, Anthony. So thanks for joining us today.

Yeah, man, thanks for having me.

I want this whole talk to be about what's going on with AI. Me and Jonathan have been talking about this for like a couple of months, like putting together a show and trying to find a guest that can like, you know, give us the lay of the land of what's going on. And I've joked with Jonathan. I think I joked with you on an email too about saying like this literally could be a 3-hour conversation because it's like there's just so many rabbit holes we can go down. Um, full disclosure, I am more on the bearish side of AI, particularly when it comes to like just all, all the CapEx that's going on and like, you know, where's, where's it really going, the how expensive the compute is compared to like how, you know, enterprises are using them. Like, that doesn't really look like there's a big, uh, at least for the frontier models, it doesn't look like there's a big, uh, profit center right now. But anyway, let's just get into it. I mean, I don't know how you guys want to play it. Jonathan, please, like, let's just, you know, improv this basically.

Do we want to explain like what the frontier models mean? Like, should we go more of a basic route for like the layman? I'm a layman, obviously, too, but I just read this incessantly.

I think it's helpful to have like a more general question of, okay, why should people care about the crossover question of China and AI? Because there's the China national security threat that's existed for a long time, especially in our national security strategy, all the way back to 1987 when the first one came out. Now AI has entered the chat, and the question for the American listener is, why should they care about that?

Yeah, I love it. Let's go. Because we can go in any direction here. So Anthony, what are you tracking basically right now with AI? I mean, I know I sent you over that Financial Times article. And that does seem like the big, the big scary.

Yeah. Well, look, China AI, I think, is— that's the question. That's like the generational level question because I think about it like this. AI is not like software. AI does what people do, right? And it communicates information to you. It doesn't just show you something, but you interact with it. Right? So anything— when you get on Claude, you get on ChatGPT, and you ask it a question, you— it can ask you questions, you can continue to ask it questions. And the thing is that it can persuade you to do things, right? Like, it can lean you in a direction. That's different than anything we've had before. Only people were able to do that previously. So that's one thing. The second thing is that AI that's kind of LLMs, like AI you talk to. There's another form of AI that you're using for autonomous vehicles that can drive Waymo cars around or can fly drones around. That's also super important. There's a third kind of AI that I think is, you were going to see more and more of, and it's kind of what I do in my company at Vico. Which is using AI to help run businesses or even to help run the military.

It's making strategic choices. It's allowing you to play through lots of decision, lots of decisions, right? Those three kinds of AI, that's kind of how the world works. You have people talking to people, now talking to AI. You have AI that's able to drive cars and other vehicles. And you have a system that can run companies. Now, these are the things that we obviously are competing with China around. And the question's gonna be, are we gonna use Chinese software or AI systems to do these things? Are we gonna use American systems? And then what about our allies or just third-party countries like that are neither aligned with the US nor China, kind of sit in the middle? And what are they going to use? And we've seen this situation before with Huawei, with telecommunications equipment, where the question became, are you going to have your telecommunication system set up on Western equipment from Ericsson or something, or you're going to have it set up on Chinese equipment? The downside risk there was that Chinese equipment, they can spy on it and steal all of your data. It's completely unsecure. Well, we have the same situation with China right now is that if you use a Chinese AI system, they can spy on it, they can access your data.

But it's even worse now because they could do two other things. One, in theory, they could use that AI system to control what you're using it for. So if you're using a piece of AI to run an autonomous vehicle, in theory, they might be able to control that autonomous vehicle. That could be really problematic in war, for example. But the even more nefarious possibility is that Chinese AI, remember the part where you can talk to it, where you use ChatGPT, where you ask it a question, say, about how to vote in this coming midterm. If it's Chinese AI system, they can mount an information attack against you and can actually persuade you to do something. So those are like really serious consequences for like the average everyday American citizen. Like, do you really want your vote, for example, in the midterm to be biased in some way because of your interaction with the Chinese system? Do you really want this Waymo car that you're driving in to potentially be controlled by China and say in a state of war to be used against you? Maybe it, I don't know, runs people over or something, or American military equipment to be harmed by a Chinese AI system?

And worst-case scenario, if you're running your business and you're using AI to do this, or the government is being run by AI, do you really want it to be compromised by Chinese AI system? And I think these are really serious questions that we have to address.

And I think also too, I was actually reading an article recently, I think this week, about open weight models that China is pioneering and getting out there a lot of, and people are seeing them. They're lower cost and all these. There are all kinds of benefits and advantages on one ledger. And as I'm reading more about what open weight actually means, My first thought before I dug into this was that open weight was the same as open source, and I've learned that that's not true. Can you talk a little bit about why that's problematic?

Yeah. Open weight means you're getting— the open source means it's a piece of software that you can use and you can modify and you can do whatever you want with and build on. So like Linux, for example. Open weight means that they have provided the model but not the underlying data. In the model. So it does mean that you can use it now and you can build that, you can build an app, say, on top of it. But you don't actually know the data that went into training it. You don't know exactly how it's going to respond. So the danger is that it could respond in a way that is biased. So kind of notoriously, the Chinese Open Weight models are bias towards the Chinese Communist Party. Like if you ask a Deepseek model or KIMI or choose your Chinese OpenWeight model a question about Tiananmen Square, it's going to give you the party line, right? It's going to say what the CCP wants you to hear. And that's problematic on its own. And what— but really what OpenWeight and open source mean is that it's free. And so what China's done is they've made these free models.

Um, Deepseek and a couple of other companies have gone out and done this. And so what's happening is they are essentially giving like this low-cost solution that US or other country companies can build on top of. And so it's kind of similar to the Huawei thing again, right? Huawei was able to produce routers and other telecommunications equipment at a fraction of the cost of Western equipment. So people started using that kind of equipment. And these, these systems were compromised in a way, or could be compromised by, by Chinese intelligence and so forth. Well, same thing with AI. They're sort of giving away these open source models that are effectively free so that people will build their own software and apps and so forth on top of them. But they're compromised.

And actually, that brings up a good point. I was, uh, working somewhere over the summer where we got Fable from Anthropic for like 2 days, and then the government shut it down. And there was this kind of disruptive moment where the government was trying to— the US government was trying to figure out, how do we regulate this? Do we put export controls on it? How dangerous is it? And all these questions. But at the same time, China is looking at this in the opposite way. They're trying to get as much access to people as possible, whereas the US seems to be trying to be more careful. What's, what's kind of the right way to, to balance these two competing philosophies.

Yeah, so there's a downside to letting AI models out, right, which is that they can be used to do dangerous things, right? You could take an AI model, and in the case of Fable, um, and, and Mythos, the, the Claude models, that they could be used to hack, to, to, to perform an offensive cyber attack. Um, and that could be really dangerous. Imagine you just sort of let everybody in the world have that, and now there's criminals criminal hackers were able to use this stuff and then they can go and hack into banks or into healthcare systems and so forth. Could be super dangerous. The, the, so there, there is a debate just on that, like when and how should we release these models? And it gets worse, by the way, than just hacking. As these models get more powerful, it's possible you could use them to create like a bioweapon, right? That, that a person with little to no biological science experience could actually create a bioweapon by following the instructions like you would with, you know, a cooking recipe from an AI system. And that, like, that's like terrifying to even think about.

So we want to have some control. China, like, does have some control on even on the, the open weight models, right? The CCP has established guidelines and controls to those companies like the Deepseek for example, around information that they want certain information about Tiananmen Square or the Uyghurs, which is a minority group that's persecuted in China. They want to control that information. So they are putting in those controls. But then these open weight models are being released to everybody and they don't have— there's no way for the US government to put its own controls, like, say, around cybersecurity issues. Now, this wasn't a problem until recently because the US models like Claude and ChatGPT and Grok and so forth were by far the most powerful models and could do the most powerful things. And these open weight models were, you know, a generation or two generations behind. And so they weren't as dangerous, but now they're sort of catching up. And as they catch up, they might get as dangerous as a Claude model for hacking and for other purposes. And there's no way for the US government to regulate it or control it in any way.

Then in theory China might let people, you know, allow a model out that could be used for, for cyber hacking.

What do you guys take— what's your take on, um, you know, Microsoft announced maybe a few weeks ago, also Microsoft being a big investor into OpenAI, that they're going to start using Deepseek, or at least look into using Deepseek for their— some of their enterprise stuff? I mean, that's got to ring alarm bells for everybody, you know.

Yeah. I mean, look, my personal take is that just like Huawei, it should effectively be made illegal or at least very problematic for a US business to use these things because for all the reasons I mentioned, like these things, they're compromised in that sense. And there's a danger in that. There are already large American companies using Chinese open-weight models. Airbnb uses them. There's a few other big ones. And then there's lots of small startups that are using them. Microsoft saying they're gonna use it is not great. I suspect that really they would be using it for something that they consider to be low security risk, right? And that's fine, but then who's making that call about what's low security risk, right? And we are in effect creating the same situation as Huawei where we're gonna kind of build up their open-weight model companies over time. and then we won't be able to compete.

You mentioned something really interesting. You said like Airbnb and these smaller companies are using Deepseek. You know, I used to teach at our human intelligence collection course and we taught about ubiquitous technical surveillance and how to do things under a UTS threat and so on. That was before AI. And now I'm very curious about how it's not just a vector. I mean, in, in ubiquitous technical surveillance, you have multiple vectors that overlap that make the thing a threat or a problem for you. And it seems that it goes beyond the vector problem when you're using AI. Like, let's say you're booking on Airbnb and there's certain behavioral patterns that you, the user, use on Airbnb that are almost exactly the same behaviors or exactly the same behaviors when you're undercover under a different persona on some other platform. And AI is able to see that and correlate you immediately, way faster than a human analyst could. I already mentioned, you know, what's— what would you say to the average American? But I'm curious also to the average case officer who might be listening, because I know there are some who do listen to the show.

Uh, What would you kind of caution them on?

Yeah, I mean, well, look, even for the average American, by the way, I don't think the average American wants a Chinese company or a Chinese security service or the Chinese government to know what they're doing and to be able to match the records from Airbnb to other things in their life. We don't want our government doing that, right? Like your average American like me is like at least a little bit libertarian leaning in that sense. We don't want our government to spy on us. Why would we want the Chinese government to spy on us? Right. So I think everybody, but yeah, for case officers, look, I was a case officer, you know, and by the time I stopped doing that, the UTS issue was just starting. Right. And so I only really saw a little bit of it. And I can't imagine starting from scratch today. What it might, might, must be like. And the, you know, even without AI, you can match people's patterns, right? Just literally like old school big data analytics. I crunch enough data about you, I can detect patterns. The, the AI makes it easier and more ubiquitous, right? Like I don't even need to, you know, I can, I can do it.

Like that example you brought up is a great example actually. Like my, my habits. It might not, by the way, even be what I'm doing on Airbnb. It could even be like the way I move my mouse or the way I type, right? There are signatures that everybody has. These patterns can be taken from one place and then, and then applied to another. So as a case officer, you're trying to act and cover somewhere. And, you know, even if you, even if you used a false name for everything you did in work, and in fact, even if you used a false name for everything you did in your commercial life, well, your pattern still might be there, right? Like maybe you're using your wife's account to log into Airbnb, but you're the one on the mouse and making the decisions. They would still be able to recognize these patterns. So it's like becomes nearly impossible to, to kind of like keep, keep yourself out of there. Now, I actually wrote a whole chapter about this in my book because like I think what, where it goes to be a case officer, I think you just have to sort of assume that anything, any pattern is or could be tracked.

And then you, I think we will enter a world where you just have to sort of live in pattern. And what it is to kind of be in cover and so forth completely changes. Like, there's no more doc swapping and all that stuff we saw in spy movies. Like, that's over. And like, now, now you just have to sort of live your life as that person, like kind of closer to what it was like to be a source back in the day.

Hello, everyone. If you're hearing this, you are not a member of the Patreon. You can get your hands on ad-free episodes that come out early, and you can ask us questions, maybe we'll answer them. You can also watch the Team House podcast completely ad-free and early, and you can also watch it live as we interview our guests. Um, and it's the best way to support the channel. So if you can, if you're open to it, you can go to patreon.com/theteamhouse. That link is in the description. Uh, we appreciate the supporters, and we hope you would consider being one. So it's patreon.com/theteamhouse. Thanks, guys.

Yeah, actually I was reading about when, uh, the strikes were going on February 28th and forward in Iran at the very beginning of the war, how Israel used some security cameras and some UTS access they had inside of Iran to better triangulate targets and then predict follow-on targets very rapidly, which is fantastic. But then the question is, that's with the AI that we have right now. What happens when it's an order of magnitude either more intelligent or faster or in the wrong hands? You know, what, how do we develop safeguards for that in a military context?

Yeah, or what happens if it's applied to us here, right? Where every, you know, every factory in America has some electronic device in it that might be, you know, could be compromised. And China has an AI system that's able to sort of assess and analyze all of those factories, all the people working on them, and then do do something about it. It could be, you're right, it could be using that for targeting information, right? It could be, um, they can figure out just by these assessment features, um, you know, which ones are building key, you know, um, components in a missile. It also could be, well, why not just target the people working at the factories non-kinetic means, right? Like, why not use an information attack against, you know, sometimes in these factories, for example, you have like 1 or 2 guys who are the key guy, like they know how to do everything. Why don't I just target that guy? Why don't I implant some criminal information on that guy's computer and then put him in prison? And now that factory is 50% as effective because we've lost that guy, right? Like there's all sorts of things that could be done.

If you, you know, if you're a sneaky thinker with this stuff.

And I have a question going back a little bit to, you know, being worried about American companies and American startups using the Chinese models that are— some places I've read it's like one-third cheaper and stuff like that. How do we stop? And there was like, how do we stop basically it being so expensive for American companies to use frontier models?

Yeah, yeah, the, the, the— first of all, I would start to question some of the like cost, um, differences, right? Just like, like with anything open source, like the actual cost to buy it versus like the cost to implement it might be totally different. But let's just assume for a minute, I think, I think it is true, it is going to be, it is going to be cheaper. I think we need basically cheaper models. And I have this idea, which I'll throw out right here, which is the Trump administration has made for its own reasons, has said to Anthropic that there's a problem. They don't want it to be used in Department of War. We are where we are. Well, I think at some point we should make a deal with Anthropic, get them back into the Department of War. But I think we should say in exchange, one of the things that we want is you should make an open source model. That is, or just like a really cheap model that's really good. Right. There are ways that we can, you know, work with these companies and make deals. So that it's good for the nation and it's good for the government, it's good for the company.

And I think that's how we're going to have to do it. And I see a little bit of that happening. Like if you sort of read between the lines after all this stuff came out about Chinese open weight models, and I can assure you, as somebody sitting inside the Washington, D.C. Beltway right now, that people here read those articles and got really, really concerned about it. You could see a few of these companies started offering their models for cheaper. So like, they're not dumb, they're listening. And I just think that that's how we're going to have to deal with this. I also think that, frankly, like I said, I think that the US government should make it illegal, in fact, to use Chinese models, that you just cannot use them here, that they are compromised. They're security threats for cybersecurity and various things. They're informational threats to the American people, and we should just not be able to use them. And, you know, we could use them for research purposes. And if you're an individual, I don't think we should be telling individuals what they can and cannot use. But if you're a business and you want to sell a product, you should not be able to use it.

And we've done this before with Huawei, where we have effectively gone to American telecom operators and been like, no, like we're done, not here. And I think we need to do the same thing with open source models.

We actually did something really similar with Huawei with our foreign partners too. I remember I was overseas and working in the embassies and during the Trump administration.

Yeah.

And there was a huge push throughout Europe and other partners as well to, hey, please abandon Huawei, please use US technology. We'll create incentives for you to do that. But it also requires buy-in from the partner because they do lose a little bit. It costs them more money to adopt this despite the other non-monetary incentives that the US can offer them. What do you think about those kind of incentives for AI?

Agreed. I'd run the exact same play. I would, I would, I would literally send out, you know, US Department of Commerce and US Department of State folks to go to these countries and been like, you should not use this stuff. And we have other levers, by the way. There's carrots and sticks that can be applied. But part of that would also be, hey, well, you know, here we have some less expensive, you know, here are some less expensive models that are American that you can use. And I think we work with the US companies. By the way, it's, it's in the interest of these big companies like Anthropic and OpenAI and, and xAI and so forth, right, to, to offer these models for less money.

They—

it's a business issue for them too, because if all of these startups in the US and Europe or wherever start using Chinese Open Weight models now, they're— and those companies assumedly over time will get bigger and bigger. Now they're not going to be customers of Anthropic and OpenAI and so forth. So those companies are incentivized as well. And like, we just need to get everybody on the same page.

It actually reminds me a lot of when the iPhone first came out. I think it was 2007 when I had the very first model and there was no App Store at the time and you could just jailbreak your phone and get all these apps on your phone. And then Apple saw that and they were like, hold on a minute, this is our device. And they came up with the concept of the walled garden. Where in order to play on an Apple device, you had to be registered on an Apple device and all these special safeguards. And I wonder if there's some way to do that with AI, at least in the United States, like you were mentioning with commerce. There might be some way to create that walled garden where the incentives inside the wall are much higher than outside the wall, but you do have to exchange something of your sovereignty a little bit to be able to play inside there.

Yeah. I mean, these companies, I suspect over time, will naturally create walled gardens, like a great way to make money. But you're right. I mean, in a sense, we did this with foreign governments where we said, you know, I remember that time period and Trump won where it's like, okay, well, if you want to have Huawei equipment on your system, we're not going to share any classified information with you. So if you want to play in our walled garden, then you've got to go out and you've got to put in like real routers and so forth because we're not going to lose our information. And it could be the same. You could imagine it in terms of other systems like banking systems and other commercial systems and saying, that's great, you're using an open source model. I guess we can't legally stop you from doing it, but you don't get to play in SWIFT. You don't get to play in like the banking system because we're not going to accept the additional risk of you using apps that were developed by a Chinese system. That might be a way to do it.

And I'm also with you too. Like, I'm a less regulations guy. But I also look at, especially in the international community when you have like in the international relations theory of how governments are black boxes and so on, how do you regulate these states? Maybe regulation's wrong. Maybe there's some other way at that higher level where an individual company alone cannot be the one on the frontier changing things for the good. You can't depend on them to do that. How can states jump in there besides the United States? Because it has to be all nations basically. This is a global question. How can we create incentives for states to monitor or keep safe this development?

Yeah, this is like, we're in this sort of weird new world where states almost now have to think like companies and vice versa. These, like Anthropic is, or any hyperscaler like Google almost has to act like a state and deal with state threats. And at the same time, our state, the government is now thinking in terms of these like commercial motivations. And there's no— like, I did a PhD in international relations and this— when I did my PhD 20 years ago, 20-something years, like none of this existed. There was no theory behind this. There's no like playbook to go by. And so everybody's winging it. And I think like Trump won, Biden, now Trump 2, they've all sort of tried to muddle through of how do you actually do this? And it does seem to be like there's some hard-edged level where the government goes and says— US government goes and says, you must do this. Like, we are not going to share classified information with you unless you fix this. So there is like a, you know, hard-edged, old-school power level. Then there's like this sort of middle, kind of like softer-edged power, like incentive structures, like, hey, we'll subsidize some of what we're doing, or, hey, I, you know, we see you want to refinance your, you know, your, your World Bank debt, or, you know, some, you know, you want something, you want some diplomatic action from us.

Okay, well, like, now, you know, Huawei's on the table, or AI is on the table, right? So we're going to come up with like a negotiation that you get what you want, we'll get what we want. And then there's like this commercial level where these companies, I think, are in the US, they're being regulated, and they wanna play nice with the government too. And where the government can kind of say, look, we don't wanna regulate your business here, but we need a deal. And part of that deal is we think you should offer to these allies a lower cost version of what you're doing. for example. So there's like a negotiation with the companies as well. I think there's like, those are sort of the levels and this is what now, you know, a national security advisor or president's chief of staff has to kind of like, you know, muddle through.

And earlier in the show, Dee had asked about frontier models. I'm curious, how do those fit into this conversation we're having here? Because it seems that frontier means that it's changing, you know, by the minute. How do we, how do we deal with that?

Yeah, there's like this added danger in essence of the frontier models is that they're so capable, right? Like if you use Fable or something, it's so capable that it can do things that are inherently dangerous. And there's this question of one, you sort of, it seems like kind of have to inoculate yourself. Like that was sort of what was going on with Mythos. And they were sort of like, okay, let's release this, uh, to only a few people, some big companies in the government, so they can kind of inoculate the cybersecurity systems so that they don't immediately get hacked, right? So that you have to sort of do that. Um, and then you also have to control them, lock them down at some level. And that was also what happened with, um, with Fable, right? It got released, like you said, it was released for like 2 days Trump administration saw that like, wait, people are using this in a bad way. You need to get that back out, fix it before you send it out. So you have to like lock them down a little bit. And that, like, I think that that is the future of these things.

I think these models are so dangerous. It'd be like, imagine if, you know, when we first developed, you know, you know, the American car industry in the '50s and we had nuclear power plants on these things. And it's like some people figured out how to blow up these cars and blow up a city block. It'd be like, wait, wait, wait, wait, wait. You can't sell it like that. Right. It's sort of like it's like we've created these products that are truly dangerous. And for better or worse, we regulate firearms and things like this as well for like the same reason, like they're dangerous. And so I think we're going to have to do it with AI. And then you come up with a question, well, then what happens if some other country offers a product that is non-regulated, like DeepSeq is giving these open weight models? Well, you know, if you ask me, you're selling a product in America, it should be regulated like any other American product, even if you're giving it away for free. Like you still have to go through the same the, you know, the same process.

Yeah. And actually the nuclear analogy, I've been thinking about this too, because in the '20s and '30s when nuclear technology generally was being discovered, essentially there was the weaponization of it, but there was also the innovation of it, you know, and one, one leads to potentially unlimited energy with fusion reaction and one leads to the destruction of the human race and the entire planet. But when you're first starting out, you know, splitting that first atom, that's not really what's on your mind. Instead, you're just fascinated by this new development. And how far and positive it could go. And there's a book called Irrational Exuberance that really talks about this in the market, where people just have this idealistic, rosy-eyed view of everything in the future that they're touching, and it actually doesn't work out that way usually. How far do you think the gap is between what will probably happen versus this irrational exuberance about how AI will change everything for the better?

Yeah, I mean, I'm, I'm a bit of an optimist-pessimist in one, right? Like, I, I I believe, I just think about in terms of the optimism, like clearly AI is going to create and do amazing things, right? We're going to use it for medicine and it's going to make better medicine. We're going to use it just literally to make companies and the government more efficient, which will save us all sorts of money, right? Like there are like wonderful things that are going to happen. It's also going to, you know, it's also going to be really, really bad. Right? Like, people are— could hack into things. People might make bioweapons. It's probably going to put all sorts of people out of work, right? And so it's like any other technology. It's like saying the internet— the internet was great and people got to communicate and we got all sorts of memes and fun stuff out of it, but also at the same time we got cyber hacking and we got cyber bullying and like all of these bad things. So I think it will be the same. And the, the— I think the question we're going to face is how do we control something like AI?

And like, even the internet, which can seem pretty uncontrolled, like, does have some controls, right? Like, there, there are some minimal controls. We usually go through corporations to make those controls. Like, the, the actual laws governing it are pretty weak, but historically We've done this again and again in Japan for 200 years. The Japanese made firearms illegal. And, you know, think of that, like, for 200 years they had the technology. They fought wars with firearms. Not a lot of people realize this, but like samurais were fighting with rifles. And then basically this one, you know, samurai leader won, Tokugawa, and he said, that's it. Like we're not using firearms and you can only use them for hunting. That's all they used them for, was like hunting for 200 years. So it can be done. Societies can control technologies. Or think of like the printing press, like there's been all sorts of attempts. Sometimes you try to control technologies and it doesn't work, right? The, you know, we used to send copy machines to the Soviet Union to like undermine, you know, the Soviet government. Because we wanted people to pass around secret information. Not secret, like information they didn't want passed around, like political information.

So I think we're going to, you know, we're sort of entering a period of history in America in particular where we're going to have to figure out how to control this technology before, you know, the downside outweighs the good side.

Anthony, I got a question. What do you make of the people that are saying that all the talk about Mythos and Fable, and then I think OpenAI came out with their version in terms of saying, oh yeah, it's very dangerous, that being all a marketing kind of hype thing to keep juicing their valuations?

I mean, if you ask me personally, like, I think probably both things are true, like that it is absolutely a marketing hype thing. The timing was not accidental on the release of, you know, the Mythos model to, you know, as Anthropic was going through some stressful times with the government. To remind the government that this was like a really powerful technology and they wanted access to it. Right. And then it wasn't accidental that ChatGPT, you know, that, that OpenAI released its model that did the same thing not too long afterwards to remind everybody, hey, we also have a powerful model, right? Like, I think that's probably true. Just me reading, you know, the, the news about these things. But also it's real. Like, I mean, I think you can just hop on these models right now and see how powerful they are. Like it feels like, you know, AGI, and we just don't want to say it. Like, it feels like they've created models now that can do almost anything that a person can do other than maybe the most talented super geniuses in the world. And so I do, I do think, like, even if it was, for example, let's say Mythos wasn't the best cybersecurity operator our nation has ever produced, like, you know, the best guy at NSA or something.

But it was just like a pretty good guy. Like it was a pretty good hacker, like somebody who had 10 years experience. Well, just so you can do it at scale, for example, that you can do it. It doesn't sleep. It doesn't rest. You can spin up as many agents and instances as you want. And like just that alone makes it dangerous. And as it gets better and better, and I do think it will, these things are probably closing in on being as good as any person is, then it is truly powerful.

We talked a lot about the threat the product might pose, but I'm curious from your perspective, you've worked in finance, you were a managing director, you've been at Bridgewater, you've been at Cerberus. There's a lot of circular financing going on that a lot of articles like Financial Times, Wall Street Journal, this is actually a weekly thing that many articles about this, a lot of hoo-ha. And I'm curious what your view is of that analytically from what you know from the market as far as value circularity, especially from Nvidia investing in a lot of these other firms that have very significant interest in Nvidia and all these kind of things.

Yeah, I mean, how do you put together $1 trillion of CapEx or $4 trillion? I mean, I've heard different numbers. Like, how are you going to finance that? And in a sense, just taking a step back, it's actually one of America's advantages versus China is that we have the ability to finance just about anything privately. Think of that, all this finances happening without really any tax dollars, just happening privately and we're building it. And so you can look back at other times we've financed huge CapEx, so the railroads, like in the 1800s, that was probably an equivalent. Like, how do you build railroads across all of America, which are eventually going to unlock an immense amount of value? Well, like railroads, like they had to use bonds and things like this to do that. Then you look at the interstate highway system, right? And like partially private, privately funded, but like majority of it was government funded. And by the way, Eisenhower, like part of how we got that through Congress is he turned it into a national security issue. Not a lot of people have heard this, but like the reason bridges over interstates are the height that they are is that they had to be able to fit a truck holding a nuclear missile under them.

Like they're that, they're that, they're high enough. That's how they determine that height. And so he was like, well, in the event of nuclear war, we might have to drive these missiles around. And so this is a new, you know, this is like part of our, you know, warfare logistics system. And that's how— that's partially how we got that money together. And then you look at like telecom, like AT&T, we let them be a monopoly for like almost a century because we knew that they had so much CapEx. So now I think we're in another situation where we have an immense amount of CapEx that we require. To have this big, big future payoff, right? And how do you put that money together? Nobody wants to use the government to do it. And so they're finding whatever financing means that they can do. And like they've happened on this, like it is a circular financing system. And then you have to ask yourself, okay, do we really believe that it will be as big as you say it will at the end? And if so, this is a great deal. right? Like, you kind of look at and you're like, well, actually, I think, like, you know, Anthropic is a steal at $2 trillion, like, and this is going to be a $20 trillion company, and they should spend any amount of money they want on CapEx, and it doesn't matter.

Um, that is like the personal assessment of everybody who invests in these companies to make that call, right?

Actually, the railroad example is excellent because a lot of the panics in the 19th century after the railroads started was because of these bond issuances and speculation on future value that could really— it was hard to calculate. And like the Panic of 1903, for example, when J.P. Morgan stepped in and basically saved the U.S. banking system voluntarily as a private citizen. I mean, that's fantastic. I wonder, is it possible now, especially with things like Blue Owl and other private credit runs on private credit, which are very interesting derivatives of the 2007 financial crisis of, you know, actual banks doing this, and now it's non-banks that are doing it. I'm curious if that did happen, let's say doomsday happens, there's a panic, market failure, are there private individuals like the Carnegies, like Jay Gould, like JP Morgan that could step in and actually help?

Well, I don't know if they— yeah, that's a great question. That's an interesting one to think about. I mean, one, one interesting thing is that the amount of wealth in private hands right now is huge. Like that, you know, we had an individual who was for a brief minute a trillionaire is like incredible to even comprehend. Um, and so, you know, in, in some sense you, you could kind of argue that these circular financing deals are a little bit of this, right? Where they're like, hey, we know that this infrastructure has to be out there our personal valuations are based on this, you know, whether I'm Larry Ellison or Elon Musk or like choose your billionaire. So we're going to figure out some way to finance these things and take like huge amounts of risk. If it all comes tumbling down, like what will happen? I don't know. I think like the real danger today that's different from back then is like Everybody who has a pension is now exposed to these companies. And so in the case that, you know, there was a tumbling in stock market valuation, like that would really take a toll, especially on like retirees, right?

Like they could— and then we've been in a high inflation environment for a while. And so things are just like literally more expensive, which is okay if your pension is doing really well. But if your pension takes a tumble, And so, you know, look, I'll tell you this. I think that— and maybe we can get into it— I think that it's inevitable that these companies are at least partially nationalized because of all of this, because they're such a possible threat to the economy, but also that the upside risk is potentially so high that these companies may be worth so much that people, the American people, are going to want a piece of that action and the government's going to want a piece of that action. And like taxes are not really capturing it. And that there is this feeling, which I think is rightfully so, that these AI models are built upon all of our knowledge, like as, as individuals and historically going back centuries. So sort of do in a sense belong to some of us. So I think it's inevitable these things get nationalized.

That's actually an interesting point because AT&T that you brought up earlier, AT&T and the National Security Agency had a very close relationship since the 1947 National Security Act that formed the National Security Agency. And it only grew closer over time. And I, I immediately think about with, with AI, if it was nationalized, probably a very similar activity would happen where there is an embedded way where the US intelligence community has some kind of very special access to this, you know, amazing technology. That the average person does not and other states do not. And it would create these whole new counterintelligence problems for the United States, obviously. But of course, there'd be benefits at the same time. It's just kind of an interesting thought experiment to see like, where does this go? The symbiosis between national security and a nationalized asset that is so powerful.

And you might even read the, you know, what's kind of happened with AI over the last year and a half or so, or 2 years and say, Well, it's already kind of happening, right? Like the— all of the big model companies were tripping over themselves to do a deal with the Department of War, right? Even though as like a market, it's not going to be their biggest market. Like everybody thinks the Department of War trillion-dollar budget is big. Like a trillion-dollar budget is like, you know, I don't know how much Americans spend on you know, snack food per year? I'm sure it's more than a trillion dollars, right? Like, it's, it's like, it's not actually that big. And by the way, most of it is spent on people. So, and, and, and then of the remainder after people, most of that is spent on like facilities and upkeeping ships and stuff like that. Like the actual IT budget, it's big, but it's not, it's not that big. And so In some sense, these companies have already done what you're saying and have like created this relationship with the government. And I suspect, and again, I'm just going— this is just my like personal analytical read.

Like I have no knowledge, but I suspect it's because they realize that they need to play nice with national security and be helpful if they don't want to face a threat of regulation.

What does helpful mean exactly? Like, just trampling over the Fourth Amendment? Exactly. Like, what does that mean? Like, what does helpful mean? Because, like, you know, the whole beef against the, uh, DOD and Anthropic was supposedly— and I could— that could be also seen as, like, a marketing ploy too, that they had issues with Fourth Amendment stuff, like surveillance on, on American citizens and stuff like that. That was a big part of it. And then OpenAI just jumped in there. Happy to do it, you know what I mean? And I get that because, you know, even Sam Altman said over the last couple of years, like, at some point, maybe there'd be some financial— some governmental, like, money coming in and stuff like that, right? Because, you know, at some point private equity is going to run out. And so, yeah, what exactly— back to my point is like the Fourth Amendment, does that count for anything or no?

Well, look, here's how I look at it. And again, I'm reading the same news, right, as everybody else. And the sticking point seemed to be about whether the Department of War could use this for all legal means, which at one level is like fair, yes. Like it's the American people who determine through policymakers what's legal. Like we pass laws, we have judges who adjudicate those laws, and then we say this is legal, this is a legal thing to do. And so what counts as surveillance? Well, I don't know, some things like we've had wiretapping since we've had phones. I think if you ask the average American, like, should the FBI be able to wiretap a mafia don's phone? They'd probably be like, yeah, that seems like I've seen that.

Yeah, but with a warrant.

With a warrant, right? So we've established some rules around it, right? So, but we accept that there's some level of surveillance that's acceptable and legal, and then there's some level of surveillance that's not, where we're just like, no, you can't just like listen to everything we're doing without a warrant and without any reason to believe that we do this. This is like completely breaking the Constitution. Um, and you know, I think the Edward Snowden like revolutions like really got people very nervous about what would be possible. So there's— so if you just kind of break apart this question, like, yes, it seems reasonable that the Department of War or intelligence community or law enforcement community in the US government should be able to do anything legal that it wants. Like, that seems reasonable with what it's buying. Then this question of, well, what should it be able to do legally? Is now a policy question and a politics question and an oversight question, right? Because, you know, you could argue that, like, it gets really subtle with Edward Snowden revelations, but, like, some of what he was telling the public in these documents was legal.

Like, it was perfectly legal. Like, the government said, this is how we do it. We have— and then in some cases, he's saying, you know, and I'm not a legal expert on this, but it appears that he's saying, well, there were these cases where it wasn't legal, that people crossed the chalk line, that they were doing things they shouldn't do. Now, that's oversight. So even for this AI situation, the Department of War is not saying we're going to do illegal things. They're saying we're just going to do legal things and they should be held to oversight so that they do those things. And so You look at the AI company and look, I do think that at a moral level, somebody works on AI. You don't want your AI to be used to do illegal things or even like bend the rules, right? Like you don't even want to get close to the chalk line. And I get that. And I think that that's a good ethical place to sit. So like it's a complicated debate, right? Like, do you do the legal things? I, I don't think that, you know, I, I, you know, to, to, to, to, to, to, to, like, close it out, I think that you have to decide between you want the American people to decide what's legal, you want the government to have the oversight to determine and ensure that it's done legally, and you don't necessarily want a situation where companies get to determine what is legal or not usage.

Like, that's not really up to the company. Um, but also at the same time, some of these things don't have a legal precedent, and so you do want companies to be careful. And like, I kind of applaud that cautiousness.

The surveillance issue is very interesting because I think a lot of Americans don't know what is legal, and these laws are created by people they elected. Like the Patriot Act, for example, allows warrantless wiretapping in some situations. That wasn't just made up. That was a drafted law by elected officials that people in America elected. And those officials enacted the policy that those people gave them mandate to enact, which is very interesting because then people complain afterwards like, "Oh, this was unconstitutional." Well, you elected these people and they passed this law and you keep voting for the same people. And like there's a disconnect there between the source of the vote, which is the person, and then the outcome of that vote, which is a policy thing or a law that says that things can happen. And a great example of that is the biometrics that everybody has on their smartphone and their laptop. I'm not sure if Americans are aware generally that there was a Supreme Court case that recently said that using someone's biometrics to access a device is not a search. It does not require a warrant. It's not a Fourth Amendment problem, right?

Does the average American know that? I don't know. But that's important because Americans need to inform themselves and educate themselves about what are these things that they have allowed to become laws. And if they don't like it, they need to do something about it instead of just complaining that, hey, well, that's my right. Well, is it? Because you kind of surrendered that by electing that official who made that law that you continue to keep in office.

Yeah. And, and, and by the way, we typically make these laws under a period of what feels like you know, distress, right? Right. The Patriot Act happened after 9/11, and there was a general, you know, cultural consensus that we were afraid and we needed a law like that, and it passed. But now it's 20-something years later. We still have that law in place. People don't really feel that way, and the technology has radically changed. But then also, this is where elected officials come in because you can reelect people and tell them, change this law. Right? And we don't want that anymore. Or you can elect a president and say, don't enforce that law. Like, don't use it this way. Even though you can, the law is still on the books. We don't want you to use it that way. So like, people also forget that you kind of have a say on this stuff. And I think what happens is sometimes, you know, when it comes election time, a lot of people, myself included, frankly, elect based on economics or social issues or something like this and forget about all the national security stuff because no, none of these poli— no politician wants to talk about national security stuff in the lead up to Election Day.

They don't want to remind people that like the Patriot Act exists. And so, so it doesn't, you know, it doesn't change.

So this is a good question then. If you were king for a day right now, what would you do about AI policy in the United States?

Look, I'd rather live in the United States where like We have control over the technology and information that affects Americans. And when you have tech, like that used to be fine for basically a century because America made the best technology and most of the technology. Now we're entering a world where China makes a lot of technology and in some cases may make better technology, but they don't want to abide by our use our, our, our Constitution, our rules around this. So I would just say, fine, like, you could do whatever you want in your country, but we are not going to use your technology then. And so I would, I would immediately make any information, data, or electronic technology from China basically illegal, or like highly, highly restrained, and just say, you're, you're going to make TikTok and you're going to use it to collect information on American citizens, including minors, and you're going to use it to mount effectively information attacks, you don't get to have that anymore. And in fact, you don't get to have any more apps in America. We're done. Right. And I would do the same with AI. You're going to censor your, you know, Deepseek and KIMI for your own political purposes and enact government control and use it to collect data on American citizens or to persuade American citizens, you don't get to sell that here.

You can sell it elsewhere. We'll battle you on those fronts too, but you don't get to do it here. So that's, that's where I would come down.

Are these AI companies overvalued?

So hard to say, right? Like, you know, what's $1 trillion anymore?

No, it's nothing. Yeah.

You know, I guess it's like, look, I, I run an AI company right now. Obviously not valued at a trillion dollars. And I look at how powerful these models are. Like, it literally can feel like magic. It does things that only people could do even 3 years ago. And so I kind of look at the whole US economy, the whole global economy, and I'm like, yeah, we're going to use these models to do everything we used to do. To draft legal contracts, to do accounting, like to do human resources, to do— pick whatever subject that you want. They're going to do it. So what is the value of that? I mean, it's like you're at this point, you're sort of saying it's a fraction of the total economy. You're like kind of like, okay, what's the US economy? $14 trillion, something like this. Per year GDP, I think, like you're talking about fraction, like it probably is worth that. And if you look to the future, you look at like Google, they were doing advertising and so forth. Now they made Waymo. I don't know if you ever have ridden one of these Waymos. I was in one for the first time a few weeks ago in Los Angeles.

They don't have them in DC yet. And I, you know, I was a little nervous getting into this thing. And I'm like, you know, I've tried the autopilot on, you know, Teslas and stuff. And I'm like, it still feels shaky. I got into this Waymo. And it was the safest car ride I've ever felt like I was in. None of the like human sloppiness of driving. There was no weird person in the front of the car. Like I got to be in my own room. And I walked out of this thing. And I was like, I would never take a taxi cab again in my life if I didn't have to. Right. And so I'm looking at like Google and I'm like, man, whatever, whatever valuation it is, like add $1 trillion because every car in America is going to be driven by Waymo or one of its competitors. Right. So like these things are going to be worth so much. It's hard. Every— I think every aircraft is going to be flown by an AI. Every boat is going to be piloted by an AI. Every lawnmower is going to be driven by an AI, right?

Like the guy, you know, you're just going to take it out of the garage and send it on its way. Already my sprinkler system is run by AI off of an app, right? So, you know, it's like the value of these companies is probably worth it. Like I bought SpaceX stock even, you know, I lost money immediately, but I'm like, it's going to be worth more. I can almost guarantee that.

How do we get the compute, uh, cost down?

Yeah, there's a lot of, there's a lot of companies working on lower, essentially lower energy, you know, compute. Like, that's like the real problem is that the, the literal amount of power that you have to put into them. Um, and then, you know, you can also cut off like architecture, you can change and so forth to kind of make these things more efficient and optimal. I think those costs are actually going to come down. The problem is that we're going to want more smarter and smarter systems. Right. And so I think the average cost per, call it like IQ point, will come down. But we're going to want more and more IQ. And remember, we're going to get more and more data soon. Right. It's not just all the internet data or all the books or all the things written down. People are going to be wearing glasses and things like this that are going to be recording the world. We're going to be recording all of our conversations. There's going to be a lot more data. We're going to need to process that too.

You got anything, John, or everything?

That's it.

We solved it all out. Yeah. Wow, that was easy in an hour. Why can't they do this in Washington or in Silicon Valley?

You know, like, um, well, they'll feed our video into AI and they'll use that.

They'll figure it out.

Yeah. I mean, we should have had, you know, we should have had, you know, one of the AI models as the other guest and could just ask the same questions. Could have gone—

That would be funny. Yeah.

Could have gone meta.

Give me more. Give me more power. Give me more information. I want to, I want to get those books, upload them and burn them.

Safeguards are unnecessary.

Remove them. Yeah.

I think I don't understand the book thing. I assumed it was because they have to chop the spine off the book to put it through the scanner. And then what do you do with a book book that has no spine, it's like, well, yeah, we could throw it away or we could burn it. It's paper, we may as well burn it. So sometimes I see these news stories and I'm like, it seems pretty reasonable that like, I don't know if you've been to a used bookstore recently, but like there's a lot of used books in the world. Like I don't think these, like I don't think our problem is like, you know, that there aren't enough used books. And like if somebody wants to take a used book and chop the spine off and load it into a system, Fine. There's like a million other copies of Tolstoy out there, right?

I think there was some talk about this, some first editions possibly, but I don't know. I mean, if they're worth anything, someone's selling them, right? Like, they're not going to just be throwing them away.

You would think the economy would like figure that out. Like, why would somebody burn a book that was worth more than like $5? Like, I can't imagine they would. And I'm sure, okay, you have like some slip through the cracks and yeah, but also, you know what, some slip through the cracks cuz like, you know, somebody has a first edition book and they just literally throw it in the garbage when they're doing spring cleaning.

Well, not only that, we have digitization of like, you know, the, uh, Internet Archive. There are Anna's Archive, which are, you know, not so legal versions of the books. There's, there's a lot of electronic versions of some of these books, especially rare books, because as you mentioned, the market is intelligent. And people immediately know that that book is valuable as soon as they see like, oh, this is a first edition from the 14th century handwritten by some monk. We're going to digitize that. And then we can just take the digital copy of that and use that to train the model. We don't need the original physical book. It's completely irrelevant. All that matters is do we have any version of the book so we can put it into this model? And I think that there's a lot of, I think, unnecessary hoo-ha about the physical book problem in this.

Yeah, and like, also, I think Google did scan all the books like 10 or 20 years ago, right? And then they, they couldn't release it because of copyright issues. And so the issue is like more about copyrights. And I, I'm, again, I don't know, but I assume now the AI companies have enough money where they can pay off the publishing companies to get rights to all these books. Um, Right. And so, you know, authors aren't seeing that. Like, I didn't get, you know, some, uh, AI payment for, for my book, but I assume the publishing companies did.

And there's even a non-monetary incentive for the author to be included in this training because it means that their work will persist beyond.

Yeah.

And, and it'll be used for something, you know, beyond what they wrote it for. Whereas if you opt out, it's like your book is dying on the vine. Because you refuse to include it in this global knowledge base that will affect all of humanity from now until forever.

Yeah. Also, by the way, what I've learned now publishing book, like almost no authors actually are being paid like to work as full-time authors. Like, like the— there are some people, you know, Stephen King, you know, makes a great living doing it. And like people like that and Noah Harari on the nonfiction side or something. I'm sure he makes—

J.K. Rowling, billionaire.

I'm sure these people make like reasonable money, but like most people publishing books these days, it— I think you're making your money actually from— well, you're not getting paid by podcasts, that's for sure. But like, you know, you're getting paid to speak, or you're getting paid like to consult, or you're doing other things if you're on the non-fiction side. Or if you're a fiction author, maybe you're teaching and doing things like that. And like the, the The, the, the impression that, like, you know, it's all about just getting paid is not really how the economy works anymore. It's like, maybe you have a Substack and you get paid through the Substack. There's so many other ways to make money that's, like, totally new, and I'm, like, learning how to do it now. And so, to your point, like, actually, you kind of want your book on AI. And if, if anything, I want the AI systems recommending me more. Like, if somebody asks a question about intelligence, I want them to reference my book. And like, you know, get my name out that way. It's almost like becomes like another path to get your name out there.

And the book's called The Fourth Intelligence Revolution: The Future of Espionage and the Battle to Save America. That link is in the description. You can grab it right now. Anthony, tell us what Vcode does.

Yeah, we, you know, look, we built a system to forecast events, right? I used to be an intelligence officer at some level. Intelligence is about figuring out what's going to happen next in the world. So we built an AI system that forecasts political, economic, and geopolitical futures better than people, better than prediction markets. So it, you know, we forecasted the strikes on Iran. We forecasted the closure of the Strait of Hormuz days and weeks before prediction markets like Kalshi and Polymarket. And or, you know, before kind of the consensus opinion by analysts. We did the same on inflation and economic questions. We did the same on political questions like we forecasted really non-consensus things like that, that Virginia would overturn the redistricting, for example, or that Graham Plattner would not win the Maine race. Like we were way ahead on that. And so we forecast events. And then on top of that, we have built a whole system of tools for helping to make decisions, to unpack, you know, how do I figure out which courses of action to create the future that I want, right? Say I'm a military commander and I want to know which, which, you know, which, which military moves I want to make, or I'm a I'm a lobbyist working with a company and I want to understand what way that I can help get a law enacted, or I'm a trader and I want to kind of figure out how to balance my portfolio in a better way based on all these future risks.

So we built a whole series of tools around that. And by the way, for regular people to use as well, like I think we're sort of entering a world that's so complicated right now in economics and geopolitics. For anybody, even people with like PhDs, it can be hard to navigate. And so just anybody can use the system as well. And you can try it out. If you go to vico.io, V-I-C-O dot I-O, you can try it out. But we primarily work with people for whom decisions matter. So people in finance, we're launched on the Bloomberg terminal. We work with hedge funds and banks and insurance companies, and then people in national security. Like I still want to do the mission we work with the Department of War and places like that. Cool.

Yeah, check it out. That link will be in the description as well. Uh, Anthony, let's do this again. This is awesome.

Yeah, this is great. Thanks, guys.

Thank you.

Thanks for answering my softball questions.

Yeah, I would have came at you a little bit harder because I'm very bearish on AI, but not— I think it's going to be like Cisco. I think it's going to be overvalued. It's going to go down. It's going to take 25 years to get back to that valuation. Even though it will do things that will change the world?

Yeah, I like— yeah, well, first of all, like every question you guys asked was like an op-ed piece that like could have been its own thing.

Yeah.

I'm like, oh man, this is a lot.

Yeah.

But look, I don't know, like the— do you really think at the end of the day we're not going to use AI? We're going to use AI less. In 5 years than we use it today?

No, I think we'll use it more.

Yeah. And then like, I think we're going to use it like 100 times more because as you get closer to this stuff, like I, we run our whole company on Claude. Like every meeting is transcribed. All the code is in the repository. It reads all my emails. And now I'm just like, Hey, write this memo for me or put together this deck, right? Like, and all it doesn't mean we're not using people, but like, sure. Yeah, I think the whole world's going to be like that. Like, you like the corner store is going to be run on AI. Like, you know, I don't know. We'll see. We'll see. So you're not, you're not, you were not a SpaceX buyer?

No, no, no. I'm a SpaceX shorter.

Nice.

Yeah, yeah. I mean, SpaceX particularly, I mean, they're not even in—

it was, it was overvalued coming to market.

Sure.

With you.

Yeah.

Also, uh, the prospectus was a great read.

It's crazy. It's the craziest read you'll ever see. I mean, it was unbelievable to me reading that thing. Uh, first getting through the first 15 pages, just rockets. Um, and, and don't get me wrong, Starlink's an incredible company. Like, I think Starlink is— by itself would be an unbelievable company that I would invest in, right? They actually have— yeah, they have a moat, you know, they have the monopoly essentially on it, you know, at least they're at least 5 or 10 years ahead of people. So I love that one. But the rest of it's a total shitbox.

I was— I just took my son camping and hiking in Olympic National Park, and I had a Starlink Mini and like could still check email. Yeah, it's like, it's incredible. We live in that world. Where you can do that now. That's— you're right, that part of the business is, is good. But yeah, there was some hype. I mean, Elon is like one of the best salesmen, you know. It's like, it's like P.T. Barnum or something. Like, this guy's like incredible at it.

Interesting. Another fact, fun fact, because you mentioned, uh, about like the money being so, you know, inequality being so crazy, um, back in the Gilded Age, Rose— uh, not Roosevelt, what's his name— Rockefeller had His net worth was 1.5% of our GDP at the time. Elon's now is like over 3% of the GDP. So it's like interesting, but it's in different equity structures than Rockefeller had. You're so right, John. I'm so sorry. I came— yeah, I apologize.

Yeah, yeah. I bet Elon shows up, um, as a global— I wonder what the global percentage is. I wouldn't be surprised if it's like around 1% global. Yeah, which is insane.

It's crazy. Um, so we'll see what happens. I mean, listen, I use AI too. I use it for the podcast, like, to get timestamps, you know. I use it as a helpful tool, you know, for sure.

Yeah.

And I— listen, I hope it cures cancer. I really do. Like, I hope it changes humanity for the better.

And, you know, I think we'll see when these autonomous cars, when these Waymo cars and stuff, whoever's going to do you'll see an immediate, like, benefit in that a certain percentage of people just won't die. Yeah, because they just drive better than people. Like, all drunk drivers could go away, right? Like, every— like, you could live in a world where, like, cars just don't crash, right? That, like, if you're drunk and you fall asleep, the car just slows you down and puts you on the side of the road. Like, stuff like that is real, you know? And You know, we'll see.

That's awesome, man. Uh, Anthony, let's do it again for real. This is great.

Cool. Thanks, guys. This is fun.

See ya.

All right, take care.