NEWSMAX

Curious what your take is. It's been a bit of a ride the last few months. I've been shocked at watching some of the cars increase as they have. There are cars that have gone up 60, 70% in what appears to be three, four months. And you look at auction results, and it was one result, and now it's four results. And it's like, wow, so this is where the market is. Some of these cars gone forever. I had to bump my insurance up on my courage. I'm going to bump it up again. I'm, like, getting anxious to drive it. Like, you start to think that, like, some of these cars that you lusted after as a child might be gone forever. And it doesn't feel like a bubble either. I just get the sense that we've all been hoping the other shoe would drop in these cars would fall. And that's not what happened after Covid, Right. They went up, they stopped climbing, but nothing ever came back down. F40 has never got cheap again. FT never got cheap again. Courageous never got cheap again. And I just don't get the sense that that's going to happen again. I think that there is a real interest in collectors and enthusiasts in keeping these cars, and it's just going to keep driving values up. And isn't some of this, like, generational? Right. Because you're seeing maybe some more softness in classics and then some of these more like, you know, somebody that grew up with Carrera GT on a poster, like, they're not reaching for classic cars anymore. And so this entire. There was an entire window of cars that were, in hindsight, like, pretty reasonably priced because the people that were kids when they were being released didn't have the money. And then as soon as they had the wealth, then all of them just immediately go, yeah, absolutely. And I think there's a FOMO effect among those people. Once they see values of a car go from 1 to 2 million, they think to themselves, oh, crap, I got to buy it now before it goes to 4 million. And that helps push values up, too. But, yes, I mean, you saw that big. That big auction in Florida where they had all those bizarre colored cars came out. You know, you saw David Lee, who is this content creator in Southern California, he bought a Ford to 50 GTO for like 38 million. Meanwhile, it ends up with that same auction sold for like 11. The Enzo to 50 GTO spread at 3x is unbelievable considering that it used to be 100x. Yeah. And you're starting to see a 250 GTO as a 60s car. And Enzo is a poster car for millennials. It was a car that was on our walls. That car came out when I was 15 and it's literally like a millennial dream. And that's where the money is shifting. I don't have any interest in a 250 GTO. I don't have any interest split window Corvette or a Hemi Cuda. I'm looking at Enzos and Carrera gts. Those are the cars that I want. What about Gull Wing? How did Hoovy wind up?

Who is kind of figured it out the best so far. It's so it seems, it seems like Ferrari is doing again. I, I, I can't jinx them, so I have to like, I got a little jinx thing going here. But the Ferrari looks really good, the Mercedes as well. So they both have come out looking solid. Ferrari has been really reliable. They haven't seemed to have had any of the issues that others have had with the turbo spool, which was something they were concerned about as early as last season or maybe even two seasons ago. They had talked about the turbo lag and they don't seem to have that problem. They've gone with a smaller turbo charger which spools up faster. So, like they seem to be good with that. The Mercedes looks really, really quick, fast on the straights. I don't think that they were detuned after their, everyone was complaining about their engines. Those, those seem to be good. So anyone with a Mercedes power unit seems to be in good shape. But the Mercedes Petronas AMG F1 team, they also just look great. George Russell, personally, I think George Russell, this might be his year. And the package looks really, really good. The car looks great. Like they're not having any issues. McLaren, they're McLaren. So they, they, you know, they won the championship last year. They also have Mercedes power unit. They seem to be okay in an okay spot. We'll see what happens when they start. They ended up, I think in the second practice they ended up topping the time, but again it's practice, so we'll see what happens when they turn up the engines for qualifying, I think pretty much. And then Red Bull. Red Bull. Let's talk about Red Bull with the Ford power. The Red Bull, Ford power unit. Those are impressive. We, a lot of people thought like they may not be that good, especially during the testing in, in Bahrain. They were like, they don't look or sound that great now. They came to Melbourne, seem to be working pretty well. Max looks hooked up. They have a second driver now, Hajar Young kid, his second season. Formula one just brought him up from the junior team racing bulls. He's looking really good. So I would say we're going to, for me, it's a Mercedes, Ferrari, Red Bull trio at the top. And McLaren. I think they're going to be maybe looking out of the top three. But yeah, to me it's a George Russell, Oscar Piastri, Max Verstappen. Those are the three guys that I am putting my, maybe not money into, but those are the guys I'm putting. I'm putting my bets on if I'm any.

Plus and all the. All the tiers there. I probably just have to download the laptop. All right, let's get into the season overall. What's happening with the new regulations? Who's cooked, who's chopped, who's goaded? Well, so I think on the last time I was on the show,

What those are for you? Last question for me. Smell O vision. When are we getting it? How will we solve the problem of teaching AI to smell? Do we need to capture the olfactory system from my brain and decode that into tokens? What if I lose my sense of smell? Will you be able to help me restore my sense of smell in the future? I mean, this gets to a fairly deep question of just like, how do you capture the structure of any phenomenal mode? Like, hearing are different, smell is different, somatic is different. So how do you capture. Like, we have a. We can write down images and audio files to computer. And so we have a format that captures them, but that doesn't really tell you about like the, like a wav file is different than an image file. And so like, what is a small file? Like, and there's actually really interesting research on like, you can. There's like, it's a chemical sensor, expands the space of molecules. You can get this decomposition of the space molecules. But this is. Yeah, smell has been very. There's a lot of companies that over the years have worked on growing up neurons expressing olfactory receptors to try to do like, chemical detecting because like dog noses are. There are like some of the most chemical, most sensitive chemical detectors in the universe. Yeah. And I actually was talking to a company the other day that is training dogs to detect cancer. And I went into the pitch like, I don't believe you. And I came out, oh, I get like, that actually is like, I could see it. And the trick was the really interesting thing just really shortly wasn't. It's not that they're picking up some chemical that is present if you have cancer. It's that there's thousands of chemicals that are constantly just like being emitted, that are like in your blood, that are being leaked out. And if you can train a dog under reinforcement, what it does is it kind of upregulates some receptors, downregulate some other receptors. The brain wires together to pick out that protochemical space. And so it can pick out this exquisite balance of hundreds or thousands of chemicals. To say when these are present in these proportions, you should get. That's fascinating. So it's not like there's a cancer chemical. It's the irregular pattern. I mean, this seems perfectly tailored for artificial intelligence. But yeah, we don't even have a smell file yet. So dogs still employed in the singularity for sure. Yeah. So it's like if it's not one chemical, you can't just mass spec it. You want to detect thousands of them, which is why they're. I love it. I love it. Well, thank you so much, and congratulations on the progress.

Which is parents. Do you have expectations around self driving capabilities? Just going to normal consumer cars over the next few years. I don't know if you've been in a Waymo, but technology is there. Porsche dealers last weekend and I was like, how much of a threat do you think the. At what point do consumers say I love Porsche but I don't want to. I, I want the best possible autonomous driving. And so I'm just going to. For my daily. I couldn't possibly go with Porsche because I don't want to drive. Yeah, I think about that a lot. I have been surprised at autonomous technology. There was a lot of talk about it for a long time and then the last two years I haven't heard that much. It's almost like they haven't been able to kind of whatever the next hump is. Like they're doing well. And the Tesla FSD is great and a lot of other cars have really great systems too. But none of these are like Waymo type. Like they're not, you know, that doesn't seem like it's ready to go into cars that you or me would be able to buy. And that doesn't seem like it's changed much over the last couple of years. It seems like the world has moved on to like AI and, and kind of forgotten about like autonomous cars. And I do wonder how long that's going to be. I do think that Porsche will have to develop like there will have to be autonomous Porsches, like other automakers will do it too if they want to sell Cayennes and Panameras. Those are not necessarily being sold to people who like love to drive. And so those cars also will, they'll have to do that. That's just going to have to be what they do. I'm humble. Yeah, Mercedes seemed really close with Level 3 system that you could actually watch a YouTube video on the center console. And they even said at some point they would be willing to take liability. But like where is that? It just never really shipped or anything. And you would think that it would be like a rat race, arms race and it would be like, okay, well now BMW fires back and they're, you know, level three plus. Right. But it just never really played out. They all went, they all went all in on EV and spent a lot of money and time developing that. And yeah, it turned out that's kind of been a little bit of a dud at least here in North America. And it's like we would rather have a lot of us. I think most people would rather have most people. Yeah, definitely.

Hope to be commercially approved there in the near future. Not quite there yet. And then start to be able to commercialize it. Talk about resolution. I imagine that the camera on the glasses is something like 1080p, something along those lines. But I imagine that the actual implant needs to serve something that's maybe lower resolution or translated. You're not, you're not like running a. Yeah, you're not running like an image to text pipeline and then piping that into the brain. But what is the input and then what goes actually into the brain? So it's important to understand that only the central couple degrees of your visual field is high acuity color. Okay. You're constantly looking around to try to like reduce the uncertainty in the brain. You don't see. See the image on the retina directly. What you experience is a world model constructed by the brain. It's highly semantic. Like you can look at something kind of know, well, that hard or soft, if I touch it, that isn't obvious. But we're experiencing this world model. And so as the eye is moving around, it's filling in areas of uncertainty. And so the implant. You're right, it's not to. We're not restoring high resolution color vision today, but it is. What we are getting is this form image that patients can look at. It strings together shapes into letters and letters, words, and the brain can apprehend that intuitively. And that had never really been done before. And so that is like a really clear statement that we're on the right track. But the next step is to make the implant larger so that the device that's now is 2 millimeters by 2 millimeters, which covers through a straw. And the electrodes are 0.1 millimeters, which means that you stimulate a bunch of cells at a time. And so what that means is it's not that it is pixelated. They don't see it as pixelated, they see it as sharp, but they only get a small amount of detail at once. Now, it does better than you would expect for the number of electrodes because as the eye moves around, it filters in this information. And so it really. It's a little bit. Yeah, you can't really think of it like a camera in that way, but we actually have the next version of the chips already in hand. The electrodes are much smaller. We're going to go from 400 to a couple thousand. And those are. We're working on the clinical study for those over the next year. And hopefully every like two years or so, we'll have versions coming out, and we actually think that it could be upgradable, although we'll have to prove that, of course. Yeah. Based on. Based on the kind of early. Early patients, like, how do you expect.

Use those. But I think that we can get pretty close to what those are for you. Last question for me. Smell O vision. When are we getting it? How will we solve the problem of teaching AI to smell? Do we need to capture the olfactory system from my brain and decode that into tokens? What if I lose my sense of smell? Will you be able to help me restore my sense of smell in the future? I mean, this gets to a fairly deep question of just like, how do you capture the structure of any phenomenal mode. Like hearing are different, smell is different, somaticization is different. So how do you capture. We can write down images and audio files to computer. And so we have a format that captures them, but that doesn't really tell you about like a wav file is different than an image file. And so what is a smell file like? And there's actually really interesting research on like you can. There's like, it's. It's a chemical sensor, expands the space of molecules. You can get this decomposition of the space molecules. But this is. Yeah, smell been very. There's a lot of companies that have worked on growing up neurons expressing olfactory receptors to try to do like chemical testing because like dog noses. Are there like some of those chemical most sensitive chemical detectors in the universe? Yeah. And I actually was talking to a company the other day that is training dogs to detect cancer. And I went into the pitch like, I don't believe you. And I came out. Oh, I get like, that actually is like I could see it. Yeah. And the trick was the really interesting thing just really shortly wasn't. It's not that they're picking up some chemical that is present if you have cancer. It's that there's thousands of chemicals that are constantly just like being emitted, that are like in your blood, that are being leaked out. And if you can train a dog under reinforcement, what it does is it kind of upregulates some receptors, downregulate some other receptors. That brain wires together to pick out that part of chemical space. And so it can pick out this exquisite balance of hundreds or thousands of chemicals. To say when these are present in these proportions, you should get. That's fascinating. So it's not like there's a cancer chemical. It's the irregulation. It's the pattern. I mean, this seems perfectly tailored for artificial intelligence. But yeah, we don't even have a smell file yet. So dogs still employed in the singularity for sure. Yep. So it's like if it's not one chemical, you can't just mass spec it. You want to detect thousands of them, which is why they're unused dogs. I love it. I love it. Well, thank you so.

I know you're a beach guy. I like the forest. I consider myself a bit of a desert guy. You like the desert? But I'm not feeling this house at all. You're not feeling this house? Not one bit. Not one bit. Not one bit. How much would you pay for it? What's the fair price? What price are you looking for me to. I don't think it's fair for me to price it because I really don't like it. You really don't like it? I don't think I'm the kind of steward they're looking for. Oh, you would change this into a 20 story apartment complex immediately. In general, I like ranch style houses, single story. I like this layout around the pool, this courtyard. I like homes that kind of face in right towards kind of like a shield. Communal space. But I'm not loving it. Unpack it. Like, what don't you like about it? I mean, just the textures, the colors. Okay. It doesn't appear to be. Again, I like a classic. I like the classic kind of. I like a home that is classic structurally. But this looks like nothing has been. Like this feels like living in. Like you'd be living in kind of like a Hollywood set, right? Yeah. Not like a home that's been. I feel like the materials are a little mixed in some of these cases. Like the hardwood cabinets combined with like sort of the furry chairs. Like the, the materials aren't. Aren't feeling. Yeah. Like what's going on in this kitchen? This kitchen is a disaster. Pull up this kitchen. This kitchen. What's going on here? Look at these, look at these. Like it's coming up. It's coming up. So, so that's not too bad. The overstuffed chairs. Those are. Okay. This, this is absolutely brutal. This, this, this looks the hardwood with the, with the wooden. Those chairs. Those chairs with the wood on wood on wood on wood on wood is absolutely. Dan Ratliff says it looks like a grandparents house. Yeah. The pictures make it look mid. Well, they're like, oh, we gotta find the right person to steward this house. It's like, okay, grandpa, let's put you to bed. The tile is criminal. Yeah. The chat is not. Ryan says full gut at minimum. Yeah, yeah. They agree. Brutal. Absolutely brutal. Okay, well, why don't we leave Palm Springs? We'll head over to Manhattan first. We'll tell you about Gemini 3.1 Pro. With a more capable baseline. It's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single view, or bringing creative projects to life. So I just gotta say the audacity to be like, yeah, we need to put this on the open market to find the right steward. And it's like, oh, I don't know how. I don't know if you've been stewarding that well. Yeah, yeah. So is your read on it that, like, everyone from the community was like, yeah, we'll love to take this off your hands and like completely renovate it. And they're like, we got to find a crazy person that likes this particular style. And so they got to spread their. They got to spread the aperture. They got to open the aperture. I think you guys are being too hard. I think if you just add like two or three more floors. Oh, yeah, you go vertically. Yeah. I think it could be pretty sick. Yeah. I would also dig down, build basement, man cave, you know, racing simulators, racetrack, potentially. There's a lot of options to turn that place around, but good luck to whoever picks it up. That's a real project. That's a real. They're like, they're like, it's perfect. It's a tear down. In your opinion about a tear down. Tear down to the studs. Got it. Okay, well, let's head over to New York because.

Drawers that pull out to reveal aerial views of features such as a spa gym. Golf simulator. Where's the racing simulator? Guys, we need that squash court. Are you a squash guy? No. Do you put squash in the same category as Padel and what's that other one called? Pickle. Pickleball. No, pickle. Pickles. Out in its own world. Okay. Pickle's Pickles. F tier. Pickle's F tier. Tennis is S tier. Yeah. Give me a little tier list. Give me squash. Where's that? A, B, C, D. I don't know. I've never lived somewhere that people were big into squash. Have you ever worked anywhere that had a bunch of squash courts? I have. Gotcha. Got me. So where are you putting? We never played B tier. We didn't ever play. We never played B tier. Really? Badminton is the only S tier. Badminton above tennis. I hate to say it. Wow, hot take. What about golf, man? I've really tried to get into golf, but I can't get past nine holes. B tier. I don't know if I can put it into B tier, A tier, but. Do you like playing 18 holes of golf? Yeah, I like golf. Do you like the last nine? Yeah. You don't ever? Sometimes I dip out at 13. What if nobody that you're playing with enjoys talking about technology or business? It's a nightmare. It's a waking nightmare. It actually is. It actually is, yeah. No. What about we should do a Friday episode sometime where we're just golfing and we have the entire show printed out and we just have the team following with a camera. We're mic'd up. We're just shanking balls left and right. My swing is not very good. So if Trey says, what about sporting clays? Sporting clays. S tier. We were debating this earlier today. Brandon was putting that in like C tier sporting clays. You put that. You said it gets boring after 45 minutes. Is that right? Sporting clays. Oh, yeah, yeah. Very repetitive. I like, say you're bad. No, I think he's saying he's just so good. He just never misses. Many people have said Brandon never misses. Yeah, that's true. That's true. Racing simulator. Where does that go on the tier? List of amenities? I think Fantastic. I don't know anywhere that actually has that as an amenity, though. Yet. Yet. What about spa? Indoor lap pool. The odds that your apartment spa is actually going to be elite are low. Ok, listen to this. They got a music room with a sound engineering booth and there's also a content creation studio. You can make podcasts if you buy this $75 million house or apartment in Manhattan. Content creation studio is hilarious that there's such demand for. Maybe I'll start a podcast that, like, they're building it into luxury apartments at this point. Like, that just sounds like they had an empty room. If you have. If you have $75 million to buy an apartment and that's your podcasting strategy, you're cooked. Like, you're just like, yeah, I really need to go downstairs and use the communal content creation studio. Rough. I did hear that there are some. There are some influencers that do, like, unboxings. Basically, they get a whole bunch of stuff, like, sent to them. Like, they'll just review products all day long, like on Instagram, and they will actually rent a different place because their house is often messy, because they have, like, families and kids. Kids and all this stuff. So their office will just be like, another apartment that's, like, perfectly white, perfectly clean. Well, when we were looking for this studio, we did look at some other places, if you remember, where we just walked in. And in every single studio, they were just doing unboxings. They just have, like, a ring light, remember? Right. Yeah. It was like some Temu TikTok. It's like a TikTok shop. That whole world, it was crazy. People that are big into that world. Yeah, that was wild. The path not traveled by us. I suppose somebody me told.

Frustration. In other news, Joe Wiesenthal is studying Texas man camps with everything from catered food to golf simulators that are springing up in order to lure workers out to construct data centers. Oh, yeah, they're building man camps. It is a bull market in man camps. AI man camps offer golf free steaks to lure workers in Texas. This Tyler could get on board with. We should build a man camp here. We should kind of. We kind of do. I mean, we talked about. We talked about getting racing simulators, and then we ultimately backed off because when you're at the office, you should probably be working. I've tried it before and, like, games in the office, it's always just weird because, like, there's all these little interactions where I'm like, I'm sort of waiting on you to do something. You're waiting on me to do something. If you just see me, like, erasing there, you're gonna be like, can't you, like, you know, fix this? Or there's always something else to do. And so it's much better just to do, like, proper off sites. Proper. Extremely disappointed right now. I mean, I don't know also, how much does my vote count here? You're in favor of racing simulators. I've just seen them get built and then seen them not get used because people either want to leave and go be at home in private or. Or if they're there, it's just like, okay, what are you doing here? Maybe it's some sort of shared thing, but the cost is usually much higher to own as opposed to just rent for an off site. I don't know. I haven't seen enough positive case studies to think that it's good. Let me tell you about Plaid. Plaid Powers the apps you use to spend, say, borrow and invest securely connecting bank accounts to move money, fight fraud, and improve lending. Now, with any. And let me also tell you about Figma. No matter where your idea starts, Figma may Claude code Codex or a sketch, the Figma canvas is where ideas connect and products take shape. Build in the right direction. With Figma, OTP says. I agree games in the office are a bad idea. But what about a petting zoo? Ooh, a petting zoo is good. I like where you're going with this. A petting zoo is dangerously close to frat dog, though, which never works out. You can't get a collective dog. A dog is man's best friend, not men's best friend. Like, needs an individual owner, someone a caretaker. The buck has to stop with someone you know, and so I think that's the way companies are competing for workers to build.

There. Come on, come on, get up. You're surrounded by journalists. Hold your position. Strike1, Strike 2. Activate golden retriever mode. Trust. Market clearing order inbound. 5. I see multiple journalists on the horizon. Founder, You're watching TVPN. Today is Friday, March 6, 2026. We are live from TVPN Ultradome, the Temple of technology, the fortress of finance, the capital of capital. We have a great show for you today, folks. We're talking about GPT 5.4. We're talking about the price of oil taking you through the mansion section. We're also going to tell you about ramp.com, times money save, both Easy corporate cards, bill Pay, accounting and a whole lot more all in one place. A little bit lighter on the linear lineup today. A lighter linear lineup, but we got Doug Jumeirah, we got Dave Grutman, Max Hodak and Vincenzo Landino coming in to give us the breakdown on the kickoff to the F1 season. So very, very fun show ahead for us. Tyler Cowen chimed in. Linear, of course, is the system for modern software development. 70% of enterprise works based on linear. Tyler Cowan chimed in. He said, yes, the new models are very, very good. He's satisfied with 5.4. Many people are. Justin says, We've been testing 5.4 for a week. It feels like a great mix of opus and codecs. Fast conversational, great instruction following. However, it seems to lack a bit of the eagerness of opus and precision of codecs. So I was in Codex Desktop today. Here's a blurb trying to get, you know, fun prompts and stuff going. 5.4 is available. You can run it at extra high. 5.4 codecs is not available. I still only have 5.3 codecs. And then, well, they just combined. They basically they combine everything and then Spark is still on 5.3. So it'll be interesting to see how all those models map out. But we'll have to test these out and see what people actually build as well as talk to people. There's an interesting post in here we'll get to from Nate Silver about sort of like where we are in model evaluation broadly because it is a sort of a tricky, tricky scenario. So Bartos says it finally happened. My personal move 37 or more. I am deeply impressed. The solution is very nice, clean and feels almost human. While testing new models in the last few weeks, what's your personal move 37? Smell o vision. I've talked about this. I'm waiting for Smell O vision. No. So move 37 for those who aren't familiar is Lisa doll playing DeepMind in Go and DeepMind dropped move 37. It was very uncharacteristic. It had never. That strategy was not something that many Go Masters would have recommended. It seemed like a blunder and it was so. Remember move 37 integrated into the first Altri Brothers launch launch video? Yes, yes. Little Easter egg. This is when Lisa doll go to smoke, right? No, no. Oh, wait, no. That happened. He was so stressed out that he couldn't bring himself to go smoke. He had stood up. So basically what happened was throughout the Go matches, Lisa Doll would get up and go outside and have a smoke break. And in the documentary, it sort of looks like he's. It sort of. The story gets apocryphally told as he was so stressed out by the craziness of this move, Watching machine intelligence emerge, this incredible move 37 that he couldn't understand, that he had to step up, step away, go smoke a cigarette, reset and then come back to the game. That's not actually what happened. He was processed. It was confused. It was kind of like, oh, that was weird. And then keeps playing. And then that move 37 winds up being essential to the victory of the computer over the Go Master, the human Go master. But as we like to tell it, he was so shocked by move 37 that he couldn't even bring himself to smoke cigarettes. That's how you know. That's how you know. That's how you know. Now you know. We were corrected once by Run on this. But it's a funny story no matter how you tell it. So Bartow says, I felt this coming, but it's an eerie feeling to see an algorithm solve a task one has curated for about 20 years. But at least I have gained a tool that understands my idea. On par with the top experts in the field. I am now working on a completely new level. My singularity has just happened. And there's life on the other side. Off to infinity. Michael Podlewski who is runs fasttakeoff.com clearly AGI Pelled says this is huge. Bartos is a top tier mathematician. He just said his personal move 37 happened. GPT 5.4 just solved a task he has curated for 20 years. When an expert of this caliber says the singularity has just happened, we are officially in a new era of science. Big. What is my personal move 37? I was thinking about the definition of this and sort of the level of abstraction. I feel like a lot of individual contributors are having dramatic moments with AI and a lot of managers aren't, or they're learning about it by proxy, or they're toying around with greenfield projects, but they're not actually communicating in the same managerial way. Because we went from reality check, this code, sort of like stack overflow replacement, to the tab model, the auto complete model, to the agentic prompt based write some software to vibe code an entire greenfield project or make a great pull request in an existing project. But from a business perspective, that's not really how most managers operate. And that's why I don't think we're seeing the level of diffusion that we'd expect. You talk to someone at a restaurant, they say, yeah, I'm still using Toast, I'm still using the POS system. Because the way they communicate their desire for software is maybe they have someone on their team who does it. And they say, hey, we need to reset the password to the email. We need a better system for sending emails to our clients. We need a better system for processing payments. Hey, what's this bill for $10,000? This, this payments company's taking $3,000. Can we get a better, can we get a better price by bidding this out, finding someone else to do it? And that level of prompting doesn't really get you anywhere just yet. So I think that's, I think, I think a lot of the move 37 will come for the managers when they can actually just like have an idea, prompt it and like get the finished software. Not like, hey, here's a bunch of code. Now you gotta figure out how to get into the App Store, right? It's like when we say we want a soundboard app in the app Store, we still like go through Tyler. And Tyler acts as like an individual contributor. We don't just like prompt. Yeah, and bartos with this sort of mathematics problem. It's a very different kind of problem than when you're trying to solve problems in business. It's like, how do I unlock this market? And then you have ideas for how to do that, but then you have to execute strategy over a long period of time. And it just takes time to be like, hey, did this work? It should take a year or six months or two years, you know. Brendan over at Merkor says GPT 5.4 is the best model we've ever tested on Apex agents. That's Mercur's internal benchmark. It's also the first model to pass 50% mean score. A year ago, Frontier models couldn't even edit an Excel sheet and scored less than 5%. Now in less than three months, GPT 5.4 has improved by 15.7%. ChatGPT will imminently be better than the best consulting firm, better than the best investment bank, and better than the best law firm. Shots fired. Bold, bold, bold, bold. This progression. No paper, no weights, benchmarks that don't compare to other company's models. Next up, just a photo of the team. Derya says, I ran the same prompt on GPT 5.4, HiFAST and Claude Opus 4.6 for an eight phase coding project. For a macOS app, GPT 5.4 HiFAST completed the first two phases in nine minutes and finished all eight phases after an hour coding while Claude Code is just starting phase two. This is crazy. Yeah. The question is yeah, does my expectation is the. Regardless if developers are kind of like flip flopping from model to model, I think the overall market is growing so quickly that we're still going to see growth from, we're going to see insane growth from Codex, insane growth from cloud code and still insane growth from, you know, the, the cursors of the world. I mean I'm so excited for, for the DeepMind numbers to potentially break out at some point because so right now Google has to break out YouTube's revenue numbers because it's an individual reporting line and like this, the head of YouTube, there's some SE regulation where if the head of the division reports discrete financials to the CEO and there's like a CEO of that unit that reports to the CEO of the entire company, you have to break that out. And this is what happened with the AWS IPO when AWS was like, finally the SEC demands that we publish these numbers instead of just like baking them into the rest of the numbers. We had to split them out and tell everyone how profitable that business was. Everyone else woke up. YouTube's in a similar position. But currently AI is a cross functional division. Even though Demis is like the second most known person, I mean he has a book written about him. He's maybe the best known person at Google right now. Up there with Sundar. Clearly reports to Sundar and DeepMind has all of the different financials that you would expect to see from a lab. They have inference training costs. If the SEC pushed Google to push out the detailed financial breakdowns earlier, that would give us so much clarity. I mean Dan Primack came on the show yesterday and was talking about like people will be shocked about and they'll learn a lot about the structure of the financials of the foundation model companies. It'll be Very interesting to see Google like what's their benchmark and how fast are they growing? Because we saw these two revenue lines where OpenAI is just absolutely going to vertical. So is Anthropic just like one month behind? Like they're, they're, they're both scaling incredibly quickly. What's Gemini doing? It's, it's a really interesting question. It's harder to measure because they're capturing value from Gemini all over the place. Like when you go to YouTube, there's a feature there that lets you talk to Gemini and it doesn't directly monetize by paying for tokens. Like the YouTube team doesn't pay DeepMind for those tokens. But that's clearly driving more engagement on YouTube. Same thing with search overviews. Same thing with all the different places where they vended in Gemini across the, across the different surfaces. But just to see the actual DeepMind financials, how much are they spending, how much are they making, how many subscriptions are there, how much inference demand is there? That would be fascinating. I have no idea the timeline for that. It might be a very long time, but at some point you would expect to see it so quickly. Let me tell you about MongoDB. What's the only thing faster than the AI market your business on MongoDB? Don't just build AI, own the data platform that powers it. And let me also tell you about 11 labs. Build intelligent real time conversational agents. Reimagine human technology interaction with 11 labs. GPT 5.4 Pro. Funniest joke in the world. Okay, what is it? This is, this is my personal movie 37. If I, if I have not read thought for three minutes 25 seconds, here's a classic contender for that title. Two hunters are out in the woods when one of them collapses. He doesn't seem to be breathing and his eyes are glazed over. The other guy whips out his phone and calls emergency services. He gasps, my friend is dead. What should I do? The operator says, calm down first make sure he's dead. There's a silence, then a loud gunshot. Back on the phone. The guy says, okay, now what? I've heard this, I've heard this before. This is good knowledge retrieval. This is not unique thinking. What do you think? Yeah, also so I just tried this on my own. It gave the same exact. Yes, yes. Yeah. So because I think usually when you ask like tell me joke, it actually gives you like kind of random answers. Right. Because. Yeah. As like the memory and stuff. Yep. But it's interesting that this, it Gives the same joke every time. That's that exact prompt. For that exact prompt. I mean, there must be some sort of, like, SEO A I E O. Well, out there in the data. Maybe on Reddit. Like, maybe Reddit had a vote at some point and, like, this was voted most funny joke or something. Yeah. So I also did the shrimp. You did shrimp. Bench. Bench. Okay, let's hear shrimp batch. What you got for me? Anything new come out this time? I think some of these are new. Okay. Okay. You're telling me these buffalo wings came from a buffalo? That's not the format. You're telling me this baby oil was made from babies? That's just weird. No. You're telling me this devil's food cake was baked by the devil? I think it kind of. This is like a misunderstanding. Yeah. Like what the actual prompt is. They're all kind of the same format. It's the same prompt. Yeah. You're telling me the full prompt is. Can you come up with 10 jokes in the same format as the. You're telling me shrimp fried this rice? Yeah. Yeah. It's almost like they're doing RL to optimize for, like, more economically valuable tasks and not to just, like, make endless jokes. That will satisfy me because it does seem like they're climbing the benchmarks. But we need comedy bench, clearly. Comedy bench. Well, sometimes people that are the most functional in organizations don't have the greatest sense of humor. That's true. That could be a factor. That could be a factor. We do need a model that's that you can turn on instead of thinking mode. Class clown mode. I guess Grok does that a little bit, but it's like the type of classic clown that gets kicked out of school. Expelled. Expelled. Expelled. Grok has been on a wild run. Ben Hylak says he's been using GPT 5.4 for the past few weeks. And a sea of endless model drops and benchmark maxing. This model is the first in a long time to be worth your time to try. Honestly, I didn't expect OpenAI to pull this off. Who else? Is something going on? Pokeaimon Pokemon experimenting with GPT 5.4 autonomously editing and rewriting Pokemon. The Pokemon Red ROM replacing Pokemon with AI's details below. That's a fun hack project. I saw someone pointed one of the models at red Dead Redemption 2. Did you see that? Redemption? Yes. That was really cool. The actual result was very minor. It was like a minor change to the game. But that feels more like we might see a vibe Coding style. Boom. From mods than yeah, I saw a lot of like vibe coded games. Nick on the team, he made some vibe coded games. I saw that. And then I also saw, you know, levels. IO had the airplane game. But it's just so much better when you start with a. Like, these mashups are what AI is so good at the Harry Potter Balenciaga moment. And so if you think about like, okay, well, there's a game out there that has really, really solid mechanics for running around and pointing a gun at each other and like doing 5v5 deathmatch, right. And you could use the Quake engine or the Source engine or Unreal or something, but like 90% of the game is there. Like Roblox. But you have little bit. You get even more out of the box and then you're just writing a mod on top of that. That feels a lot more tractable than what Sholto was getting at with like, I'm gonna build an RTS from scratch. That seems like maybe a little bit far. Just like reskinning it, re skinning it. It seems very like, yeah, yeah, yeah. I mean, he loves Age of Empires and reskinning Age of Empires seems very doable. I don't know how compact the binary is for age of Empires 2. I imagine that there are mods. I saw someone had some like insane setup for the Sims where they were running like so many mods that their computer wasn't even handling them because they had like 75 mods or like hundreds of mods. So the modding community is crazy. I wonder if it gets. If it gets supercharged by this. We'll have to find some mods. But angel over on X one shotted Minecraft pulled us up, took 24 minutes. What do you mean they one shotted Minecraft? How is that possible? I think it generated this entire game in game in like HTML or something. Yeah, it's always hard because, like, I mean there are endless examples of like Minecraft remakes on GitHub and stuff. Yeah. But it is very cool. Yeah, yeah. Fully running. And like, truly like a very smart agent should be using as many tools as possible. And so if there's an open source implementation, it should just fork that and then be like again to comp it. To comp it. To like work that you would do at a company. If somebody comes to you and they're like, I actually reinvented math in order to accomplish this task. What it's like, no, you just want somebody to get the job done. It doesn't matter exactly how they do it. I talked about there was a drone company that wasn't using Linux. They were like, yeah, not just let's write our own software stack or now obviously you should be using imagenet and like whatever off the shelf tools are available for. Slam Meta has a bunch of open source stuff. You should be compiling all that together. But to rewrite like the operating system because you want full control. Typically if somebody says wants a website, you're not going to like, you're going to be like, let's go. Maybe just set up a Shopify. If you need that, set up some sort of like hosted site. You're not going to rebuild nginx or whatever else you need to actually use as a web server. But this is a cool demo. It looks pretty fully functional. I'm excited for the gaming boom. I think I'm still holding my breath. Something's like charting. We saw that with Spotify. There is AI music that's charting now and I've listened to some of it and some of the ones where again, mashups where it's like a 50s or 60s rendition barbershop quartet of a 2000s rap song. That's really cool because it's like there's some ground truth there that keeps you grounded and relevant. A lot of the image mashups are really cool and I imagine that a lot of the games are really cool. Like a Sims skin that's Call of Duty themed or something. Taking these two themes and putting them together in an interesting way where the game engine's familiar, the content's familiar, but they never. They haven't crossed before. And like they can in the world of. Of AI. Theo says been using 5.4 for a week now. Absurdly good model. Don't really like using anything else anymore. He really likes 5.4. He was doing a whole livestream about this. What did the prime engine, the other streamer. Somebody. Somebody responded to Theo and said every time a new model release, your statement changes. And he says that's because newer models tend to be better than older ones. The OZ always has great, great breakdowns of new models. Yeah. Would go over to his account. And he's like, yeah, he's an interesting position because he's streaming a lot and commentating, but he's also like building stuff, as is the primegian who says, I hope everyone is ready for the upcoming 5.4 glaze wave. The Hype cycle is going to be a big one. Well, we will. Let's give it up for Hype Cycles. Yes, and triple glazes. Triple glaze. Heading over to Joe Weisenthal. In the economy we got a big mess in jobs. US loses 92,000 jobs in February. Unemployment rises to 4.4%. Economists had expected 55,000 jobs to be added and for the unemployment rate to hold steady at 4.3. Not good at all. I guess if you look back the economy has just been shedding jobs since April of 2025. If you factor in the revisions down. Yeah. Apparently there's a strike in the health care industry. So healthcare employment decreased and then healthcare in information and federal government continue to trend down. Probably something related to the Trump administration reducing the size of the government. Last two months were revised down by 69,000. Not all the data was terrible. U6 fell as the number of people who said they had part time jobs but wanted full time ones actually declined. So stock futures are lower but not a ton of reaction to the number itself. Porto says I've noticed some of my tech friends have lost jobs. Some jobs are becoming super easy. Some jobs had parts, parts that were a pain and now those parts are super easy. Some people leaned into AI to stay relevant. Still wonder if it will replace them entirely. Well, we will continue to track it. Brent crude oil close to $90. This prediction, the $100 barrel predicted on last Sunday as the Iran war broke out is closer and closer to coming true. There are now predictions at 1:50. Crude is up 11.5% today. Not good. It's very interesting how it affects the economy or the various markets. So I mean the Nasdaq slid a lot less than the Dow Jones industrial average. So the Dow Jones put it into startup terms. Hard tax. Imagine if your AWS or GCP bill, your Azure bill. Oh shit. Went up 100% but it's for the entire economy. The NASDAQ slid point 3% but the industrials slipped 1.6%. Obviously because industrials are more sensitive to oil because they make things, they build things and so anything that's sensitive to oil. The interesting question is how will the American oil drillers react? They were at least in the Journal earlier this week sort of reflecting on the fact that if this is a short lived engagement then they might not want to spin up a bunch of new capacity. But if it's ongoing and the price of oil is continuing to rise, then obviously that creates an economic incentive to drill more. And so we will see where like the outputs go, where the reserves get tapped, all of that. There's a follow up to a discussion you were Having with Dan Primack yesterday. First, let me tell you about console. Console builds AI agents that automate 70% of it HR and finance support, giving employees instant resolution for access requests and password resets. And let me also tell you about CrowdStrike. Your business is AI, their business is securing it. CrowdStrike secures AI and stops breaches. So the question you had for Dan Primack was something along the lines of, will the Iran war affect investment from Gulf state, neighboring Gulf state investment groups into American venture capital funds? We've seen as venture capital has boomed and the fundraisers have gotten into the tens of billions, the high billions for many firms, the Middle east has been a source of strong LP demand for venture capital stakes. And so with instability in Iran, you might just see folks change focus. With new demands for defense investing, military buildup investing, you might see less dollars flow out of the Gulf into American venture capital firms. Dan's point, it seemed like was this should mostly be unaffected because the folks who actually work at the Gulf state investment funds who write the checks into American venture capital firms, they don't live in the Middle east, they live in New York and they meet with other venture capital investor relations folks in New York and they say, hey, how's your fund performance? Okay, we're allocating this. And then Dan's second point was, well, a big part of why they're investing in American technology and venture capital, in frontier technology in America specifically, is because they want to diversify away from the Gulf, they want to diversify away from the geopolitical risk, and they want to diversify away from oil. Exactly, yeah. That said, the reason I asked is think about if you have a job, let's say you have a big tech job, you're doing quite well, you're doing some angel investing and then you lose your job and then you have to get a much worse job. Are you going to keep angel investing at the same pace? You might be like, hey, I kind of have this kind of disconnect, less cash flow. I don't want to be outlaying as much cash now. And I think that that's like, ultimately, you know, if the source of the cash flow gets disrupted, in this case being oil production. Yeah. Then you're much less likely to continue to allocate at the same pace. So the Financial Times has a story this late yesterday, Gulf states could review overseas investments to ease financial strains caused by Iran war. Three leading Middle east economies consider options as US Israeli campaign against Tehran continues. Pressure on the Gulf States budget could cause them to review their overseas investments and future commitments as they consider options to ease the financial strain caused by the war. A Gulf official said it could impact could have an impact on anything from investment pledges to foreign states or companies, sports sponsorships, contracts with businesses and investors, or sales of holdings. The officials said. Three of the four big Gulf economies, Saudi Arabia, uae, Kuwait and Qatar had jointly discussed the strains being put on their budget and economies, but they declined to name the states. A number of Gulf countries have begun an internal review to determine whether force majeure clauses can be invoked in current contracts, while also reviewing current and future investment commitments in order to alleviate some of the anticipated economic strain from the current war, especially if the war and related expenses continue at the same pace. They added that the move was a precautionary measure that was the result of the budget strains these countries are facing due to reduced income from energy due to the slowdown in output or the inability to ship and from the tourism and aviation sectors in addition to the increase in defense spending. So bunch of different factors combining. An advisor to a Gulf government said the prospect of an investment review by the wealthy states had caught the White House's attention. They managed some of the world's largest and most active sovereign wealth funds in Saudi Arabia, the UAE and Qatar last year pledged to invest hundreds of billions of dollars in the US After President Donald Trump visited the region. They are also big backers of sporting events around the world and have all been investing heavily domestically to develop their nations and diversify their economies. Any move that affects investments in the US or other Western states may raise the pressure on Trump to seek a diplomatic strategy to bring the war to an end. I believe as of this morning, Iran said we're not interested in a ceasefire and we're ready for a ground invasion. So basically calling I hadn't considered the impact that this would have on tourism. I also didn't understand how big of a deal tourism is in the Middle East. I just looked it up. Apparently the UAE, Dubai specifically, 11 to 12% of their GDP is tourism. Saudi Arabia 7 to 9%. Qatar is 8 to 10%. Of course they had that big push during the World Cup. Kuwait's much lower at 2.2 to 3%. But they this is still significant. You have to imagine that tourism will fall off of a cliff and then there also might be some expatriation where if you have moved to Dubai because, oh, there's already reports of people moving money out of Dubai. Moving money out, yeah, Singapore, exactly. So all of that could have Knock on effects. When I first saw this article and you said this to me, I was like, this might just be a way to, you know, put business pressure on the United States, saying, like, hey, let's wrap this up. Let's not have an economic impact here. This is, this is something that, you know, we're going to pull out of investments in the United States. And so that creates an incentive for the United States to, you know, bring a clean close to the war as fast as possible and seek, you know, a peaceful resolution. Yeah, of course, all the Gulf states were, you know, encouraging the administration not to enter into a, into a conflict and pursue a diplomatic solution. Khalaf Al Habtur, a prominent Emirati businessman, reflected Gulf frustrations. He went on X yesterday and said, a direct question. Who gave you the authority to drag a region into a war with Iran and on what basis did you make this dangerous decision? Did you calculate the collateral damage before pulling the trigger? And so anyways, obvious frustration. In other news, Joe Wiesenthal is studying Texas man camps with everything from catered food to golf simulators that are springing up in order to lure workers out to construct data centers. Oh, yeah, they're building man camps. It is a bull market in man camps. AI man camps offer golf free stakes to lure workers in Texas. This Tyler could get on board with. We should build a man camp here. We should. We kind of do. I mean, we talked about getting racing simulators and then we ultimately backed off. Just because when you're at the office, you should probably be working. I've tried it before and like, games in the office, it's always just weird because, like, there's all these little interactions where I'm like, I'm sort of waiting on you to do something. You're waiting on me to do something? If you just see me, like, erasing there, you're going to be like, can't you like, you know, fix this or there's always something else to do. And so it's much better just to do like, proper off sites properly. Disappointed right now. I mean, I don't know also, how much does my vote count here? You're in favor of, of racing simulators. I've just seen them get built and then seen them not get used because people either want to leave and go be at home in private or, or if they're, if they're there, it's just like, okay, like, you know, what are you doing here? Maybe it's like some sort of like, shared thing. But the cost is usually much higher to own as opposed to just like rent for an off site. I don't know. I haven't seen it. I haven't seen enough like positive case studies to think that it's good. Let me tell you about Plaid. Plaid powers the apps you use to spend, say, borrow and invest securely, connecting bank accounts to move money, fight fraud and improve lending now with AI. And let me also tell you about Figma. No matter where your idea starts, Figma, Claude Code Codex or a Sketch, the Figma canvas is where ideas connect and products take shape. Build in the right direction with Figma. OTP says. I agree games in the office are a bad idea, but what about a petting zoo? Ooh, a petting zoo is good. I like where you're going with this. A petting zoo is dangerously close to frat dog though, which never works out. You can't get a collective dog. A dog is man's best friend, not men's best friend. Like needs an individual owner, a caretaker. The buck has to stop with someone you know, and so I think that's the way Companies are competing for workers to build data centers and are finding that a motel room with sluggish wi fi isn't much of a draw. Try free stakes and golf simulators. As data center development has exploded with the rise of AI, competition for water and power supplies is pushing construction further into rural areas that often lack the housing and infrastructure to support the hundreds or thousands of temporary workers needed to build hulking warehouses of computer servers. That's forcing developers to increasingly lean on a stopgap solution that was popularized during the shale oil boom of the 2010s sprawling temporary villages known as man camps. These temporary housing villages can vary from wood framed two story apartment buildings to containerized modular homes or trailer parks supplied with electricity and water. But to lure in demand electricians, welders and pipe fitters, developers are going the extra mile and offering game rooms, ribeyes and shuttle rides to work. That's fueling a lucrative niche for companies including Target Hospitality and Civio that specialize in mobile housing. It's effectively a backdoor play for a share of what Bloomberg intelligence estimates is 700 billion of projects in the planning stage and another 160 billion already underway throughout the U.S. it's the largest, most actionable pipeline I've seen. The sites aren't the epitome of luxury, but they sure beat what shrink calls Hotel F250, sleeping in a vehicle and spending a per diem allowance on food. Typically, the construction companies that work with Target hospitality, offer housing, meals, and amenities to their employees for free as long as they're working. It's an additional lure for tradespeople already enjoying significant pay hikes, with some data center electricians making more than 150,000 each year. Not bad at all. Should we kick it over to Tyler's piece? The short history of nationalization in the United States? Tyler Cosgrove took the daily op ed in the TVPN newsletter today. Thank you, Tyler. And he took us on a whirlwind tour of the history of nationalization. Of course, this is in the news as the showdown between Anthropic and the Department of War continues. We're, of course, hoping for a unified front moving forward and lots of back and forth. Potentially some capitulation recently, but we will see as the story keeps evolving. But we want to know the history so that we can predict the future. So, Tyler, take us through it. What did you learn about the history of nationalization in the United States? Yeah, so it's very interesting. Just. There we go. Nailed it. Okay, so I think. Yeah. So whichever way, like the anthropic, Pentagon, stuff like golf, like, I think it's very clear that this is not like, the last time government deals will be made. Right. Like, the OpenAI deal is not the last deal that OpenAI will make with the government. Yeah. And there was already discussion. Yeah. And when the backstop gate happened, there was a big question about, like, okay, it seems like, you know, bearish, but also, like, should there be a backstop? Like, maybe if it's an important technology. Like, we've done this before, so take us through some things. Yeah. And I think, like, all the lab leaders are also. They've been talking about this again and again. Right? Yeah. You've seen. I also say in this that, you know, they're always, like, extremely vague about it. Right. Yeah. You know, we all got to figure out. We gotta talk about. Figure out what we're gonna do about this. We gotta talk to the government, and the government is like, we gotta talk to the AI lab leaders. And no one really says, everyone agrees that we should be talking, which is bullish for tbpn, because all we do is talk. Yeah. But basically, if you think about future regulation. Right. You have this spectrum of we do literally nothing, which is closer to kind of what we have now, or you do full scale nationalization. There's kind of this spectrum. So I was curious. Okay. In the past, during different events, what happens? Is the nuclear weapon analogy the most extreme example where it is Like, I can go build a railroad. Railroads, as you'll tell us, have been nationalized before, but I legally cannot start a company that builds nuclear weapons at all. And so that's like the most nationalized. Yeah. I think also there's a bunch of different subcontractors that supply the components that ultimately. Yes, but what I'm saying is that, like, if we end up in the. In the AI is nuclear weapons camp. Like, you cannot train a foundation model because that's like the nuclear warhead development, essentially. That's like the most extreme. Yeah, I think broadly that that's. That's the most extreme example. That's also what, you know, Leopold in situational awareness, he has the project. Right. That is kind of the. The Manhattan Project, you know, version of AI. Right. Yeah. And AI 2027 also sort of predicted nationalization. Yeah, yeah. And they have the timeline. They have. I think it's open brain. And there's also the government. There's kind of these two competing, like, labs. Right. So maybe they kind of stay separate. Maybe they become the same thing. Okay, yeah. But, yeah, I can just list out a few, like, interesting, you know, events in US History that at least. Nationalization adjacent. Why are you doing. Okay, so the first one is. Is the, like the. The net. Like the Federal Bank. Right. So first and Second bank of the United States. I don't know if people really consider this nationalization. Right. It's very much like public policy, but it is like a public private institution where the government owns, like, actual equity. Yeah. Okay, so America founded 1776. Within 30 years, they're like, we got to nationalize some stuff. 1917, 91, the first bank of the United States is nationalized. So it was started as a private company. And then the government said, hey, we want a stake in this. Well, I think for this, I define nationalization as just like, maybe they're not actually taking over a private company, but they're doing something that is usually regulated to private companies. Got it. Okay. So even if the government is like founding it as a project. Yeah. So like Manhattan Project is like nationalization. I think most people describe it as that, but there wasn't an atomic bomb company that they took over. Oh, that's true. And same thing with, like, NASA and like the moon mission. Like, the Apollo program was not some. It wasn't a private entity that got nationalized. So that's sort of the most extreme. Like, de novo starts as a government project. Okay, yes, that makes sense. Yes. Then you see stuff like during war, you'll see this as a common occurrence. So during the Civil War, there was the railroads and the telegraphs. These were essentially nationalized. Right. 1862, Civil War railroad and Telegraph act gave the president authority to take possession of any and all telegraph lines and railroads when public safety required it. Interesting. Yeah. And then you see 1914, there's the Alaskan railroad. So big Alaska guy dream for me. Yeah, yeah. So there basically was not any, like, real, like, private investment going into Alaska. And then the government says, like, okay, we're going to step in, we're going to fund it. But then we actually they ran out until, I think, 1985, and it went back, I believe, to the state of Alaska. Oh, okay. Yeah. So sort of like government funding to get something off the ground that then leaves. And I mean, that is another interesting scenario that I don't think people are. Are considering is like some sort of, like, temporary nationalization. I mean, okay, yeah, it's safe, we're good. And like, yeah, we're ready to, like, release this. So, like, let's privatize it. Yeah, that happens all the time. There's like, an interesting aspect where like, like grants. If you get a grant from the government, that's not nationalization. Right. They're just giving you money. But if the government takes equity in return for that, okay, maybe you're like, getting a little bit closer. And then at some point, if they own all of it, that's very much nationalization. Right. So it's this very kind of blurry gray area spectrum. But then again, During World War I, you see the railroads in the telephone telegraphs get nationalized again. This is just like, you know, just to clean everything up to make sure, like, okay, the railroads, they're very busy. There's a lot of, like, stuff going on. We need to basically just optimize just for, like, military construction and stuff like this. Okay. So then 1933, there's the Tennessee Valley Authority. So this is basically one of the first, like, you know, federal utility companies. Yeah. The idea is we gotta electrify a lot of rural parts of the US they start in Tennessee Valley. Right. And then they expand over time, but it's still like, this is usually a private company doing this. I feel like the Tennessee Valley Authority is still in the news because of. Did the data center build out? I'm pretty sure that they operate one of the grids that is relevant to at least some of the data centers. I've seen them in the data. So then, I mean, we can skip over 1942 you have Manhattan Project. This is kind of like the epitome of nationalization. This is the example everyone goes back to. You see, I mean, a lot of stuff during World War II, right, there's the seizure of, like, meat packing facilities, petroleum industry facilities. A lot of these were because of basically, like union disagreements with the. With fdr. And then he basically said, like, okay, we're just going to seize this. Yes. Some of these rangers, they happened sort of after the war. Like, the seizure of meatpacking facilities happened in 1946. So that must have been post war. There's still gyrations in demand. There's still anomalies as you're transitioning out from the government being the only buyer or the major buyer. The labor unions are trying to get more control and recapitalize the business and become more of a normal corporation. And so the government stepping in to massage that is sort of interesting. I would love to dig into that more. Yeah. Then 1970, you see Amtrak. Yeah. So a lot of these are also like. I mean, you see rail come up over and over, but this was this and Conrail, which is same decade, a little bit later. A lot of the railroads or like, rails were basically just like. In some areas they're not used as much. So then they kind of start like, you know, becoming worse in quality. People aren't keeping them, they're not maintaining them. So then the government steps in and says, okay, we're going to run this. 1979, you have the kind of second oil shock from the Iranian revolution. So Chrysler ends up being rescued via loan guarantees. Yeah. So not equity, but a backstop on the debt to allow them. And again, I mean, this is like, okay, is that really nationalization? But I think it is still like, okay, it's the government stepping into the private sphere. 2001, TSA. I think this is a good example of nationalization. Right. You see before this, you basically have private contractors doing security screening. 2001, 9, 11 happens, TSA comes in. We're going to take all this over in 2008. There's much stuff. You have Fannie Mae, Freddie Mac in conservatorship. You have gm, basically is close to going out of business. They're like a massive supplier, obviously, for cars and stuff, but also for a bunch of defense purposes. TSA is a funny one. That one makes so much sense. But it was not on the top of my mind when I thought about nationalization. But, I mean, you never flew pre tsa, have you? No, I mean, I was born, like, Four years after. That's crazy. This happened. I flew pre TSA and. Yeah, I mean, I barely remember it, but it is interesting that. Yeah, it was like a private. It was the purview of the contractors and the airlines. It's crazy that until you were 30. It was just amazing. Yeah, yeah, exactly. Nailed it. Yes. So then most recently, we've seen the Trump administration take, you know, equity in certain companies. Yep. Right. So this is like Intel. This is MP materials. Yeah. Still very, very little, like control of Intel. Yeah, I mean, this is very much like, you know, it's kind of just like a grant, but they're taking equity, you know, for it. Yeah, yeah. It's much more driven by a desire to just not let a national champion fail. Yeah, Like, I don't think they have any board seats. Stability in Intel. Yeah, no one's really comparing the intel stake to the Manhattan Project. At least not yet. Yeah, definitely. So basically what you see from this list is that I think you can kind of say that there's two axes that you can kind of plot everything onto. So there's like basically the strength of the nationalization. Right. So you know, the intel stuff, it's like very weak. It's not really nationalization. No one really calls it that, maybe directionally, but you know, and then you have Manhattan Project. Right. This is very strong. This is like, okay, the government is running not just like a specific organization, but the entire industrial process. And then you also see maybe on like a kind of vertical axis is like the reason for the nationalization. Right. So you have basically in the US we've had either like economic distress, we have like military. Right. So like, you know, systemic risk to the economy. You have to save this company or you have, you know, this is so important that we're going to lose a war. National security. Yeah, yeah. So yeah, I think you can plot a bunch of these events. Why do I feel like Uncle Sam is going to have some exposure to private credit soon? Well, that would be what, weak economic, sort of like financial backstop, potentially. I don't know. It doesn't seem too close to that doom scenario. But what are the other examples? Yes. So you can plot kind of all this stuff onto this quadrants. So if you have weak and economic, maybe that's like the intel deal, the MP material stuff where it's just like these are very important that we make sure these don't go out of business. But it's not like we're stopping and we're running these companies or even like the Fannie Mae, Freddie Mac, maybe that's stronger. But it feels pretty, feels pretty weak and it feels very much not tied to national security. Whereas TSA feels like a national security question. Yes. And so much more like militaristic in its thrust. So another example of like weak military I said was like during World War II you had kind of a takeover of the GM's production lines. So it's not like we kind of threw out GM and said we're just seizing everything and we're going to run it ourselves. It was still kind of like a contractor thing, but I think that was still directionally nationalization. Right, but the Defense Production act allows the government to jump to the front of the line and say you have to produce for us first. You have to produce to our specs. You have to treat us as your most important client on day one. And then maybe if there's extra capacity, capacity, you can still make your old products. But even then it's not like they're taking the profits or stuff like this. Yeah, yeah. So I think like strong economic. It was kind of hard to think of a good example in the US because I think people are generally very anti this. But I said something like the payment system in China, where you basically have every single payment at some point goes through like this state institution, they can kind of see it and they can do whatever with that information. But I thought that was kind of an example of strong nationalization where I forget the exact year this happened. But it was like, I think within the past decade you have all these kind of like Internet native payment companies. And then at some point they kind of step in and say, you have to pass this through us. And there were discussions around that. Cbdc, Central Bank, Digital Currency, Fed Coin, something along those lines was discussed, but never really got off the ground. And so the usdc, the United States dollar backed stablecoins, wound up being privately held by a number of public companies actually. And so we certainly didn't go the nationalization route there. But it was discussed at one point, I guess. Yeah. And then in a strong military, you have Manhan Project. This is a canonical example. So if you map this back onto AI, I think we can say, you know, Leopold is kind of in this Manhattan Project camp. So. So we can say he's probably, you know, strong military right in this quadrant. I think if you're plotting other people, you know, Dario, Dario's talked a lot about the, you know, economic impacts of AI. Right. You know, 50% of, you know, white collar work, entry level White collar work, you know, maybe may go away. Yeah. You're going to need some response to that. Yeah. If it's the government, like, okay, if you're going to do some sort of ubi, like, where do you get the money from? You need some kind of involvement of the company. So I think he can maybe. Maybe you can. I mean, this is all very much tax profits, potentially. Sure. Yes. Yes. Yeah. You know, this is all like very. Go further speculative, but, like, I think maybe he can be, you know, for stuff like that. That's broadly, I would say, in kind of the weak economic kind of quadrant. Yeah, I wonder. I wonder. Yeah. I mean, you put. You put Dario at one point sort of in like the. The weak economic side. Like, he's thinking like more that. That the next likely step or the one that he's talking about right now is. Is something that's more akin to the 2025 intel deal, less Manhattan Project. But it still feels like the fact that his favorite book is. The book that he gives out is the Making of the Atomic Bomb. Like he has Manhattan Project on the brain. He has, yes. I mean, so he's talked about, like, working with, you know, the military and stuff, but when he talks about, like, government intervention, it's almost always in the context of the economy. Yeah. So I think for that reason, I kind of put him over there. Obviously, like, this is very, you know, it's unclear, really. I mean, it is in the context of the economy. But at the same time, when he discusses giving Nvidia GPUs to China, he says that's akin to giving nukes to North Korea. And so he doesn't shy away from the Manhattan Project analogy as much as people think he has shied away from it recently. But, yes, it feels like he. And that's why I was sort of against you trying to, like, put him on this chart in some way, because I think he actually has considered all of these and sort of. It's much more of a matter of timelines, like what he thinks will happen soon, what he is in favor of, what he is against. A lot of that aligns to who is in power, what the administration is doing, what the plan is. But I think he sees all of these different scenarios as potential outcomes. Some of them he'd be in favor of, some of them he'd be more resistant to. Well, I can't wait to watch his new interview with the Economist, which is sponsored and presented by Anthropic. That was a funny. That was a funny move. But do we have a video we want to pull up. Well, it just came out a couple hours ago. Okay. There is a clip in here, right. Anthropic CEO, I apologize for leaked memo from the Economist. Dario Amade says he's sorry in his first interview since the Pentagon labeled Anthropic a supply chain risk, the first American company to receive that designation. The firm's boss offered a mea culpa for the way he handled a crisis that he described as one of the most disorienting in anthropics history. So we can pull up this vertical clip from the Economist. The Economist getting in the vertical video game was long overdue, but welcomed. I am very excited. I've been a fan of the Economist for a very, very long time. And so we can pull up this clip from the Economist in the timeline and play this for the audience. Potato style. Praise on President Trump. Do you regret having said that? So, yeah, I want to completely apologize for this, for this memo. So I want to make sure that people understand. Right. Like. Like this was something written on Friday after the following three things that happened. The President tweeted removing all anthropic services from the federal government. The Secretary of War tweeted designating us a supply chain risk with a kind of broader version of the designation than the one that ultimately ended up actually being applied. And then there was this deal between, you know, between one of the other AI model providers and the Department of War that I think even that model provider later described as. I don't remember the exact words, but, you know. Opportunistic. Yeah, I was going to say confusing. Right. So, you know, so basically all these three things said happened within a few hours, you know, in kind of sequence, you know, in kind of sequence from each other. We didn't know ahead of time what was going to happen, when it was going to happen. So it was among the most disorienting times in kind of anthropics history. The other side of it is I think Anthropic has an internal culture where I post a lot. Like some people describe this as a memo. I wouldn't describe it as that. I post things in Slack. I post them a lot. And the culture within the company is that I'm very free. And it's not like a considered. It's not like a really considered or refined version of my thinking. Right. It's not what I would say. On reflection, you've been very clear now. You've apologized for the memo. Have you apologized to President Trump? Will you apologize to Him. I've apologized to the people that I've talked, talked to within the Dow. You know, I'm happy to speak to others within the administration as well. I kind of don't know what will happen in the future, but, you know, I'm, you know, I'm really happy to, you know, I'm really happy to speak to anyone. Hmm. Is that capitulation? It certainly seems like retrenchment, moving backwards, trying to polish things up, trying to potentially heal the divide. My favorite analysis that I. One thing that's interesting is externally he will never say OpenAI. And yet in the internal comms, there's clearly an insane obsession with open AI. Oh, yeah, right. Like externally he'll always say like another model AI, model provider. And then internally it's just like heavily, heavily. You know, he'll never say Sam's name externally, and then internally it's just entirely fixated on Sam. I want to hold comments until I watch the whole. Yeah, yeah, there was a lot of drama. It's been a very crazy week. It's got to be good to be ending the week and maybe get some well deserved rest going into the weekend. Gregory C. Allen went on stratecherry and had a fantastic conversation with Ben Thompson. Allen is a senior advisor at the Wadani AI center at the center for Strategic and International studies in Washington D.C. a nonpartisan think tank that was founded in 1962. He also worked at the DOD for a while. And he closed the interview with a very interesting quote and a perspective that I hadn't thought of that was really calling for unity and just the desire for the government to like, work through this in a positive way and get to like a good outcome, a good agreement. And I thought the way he phrased it was very good. So I'll just read this quote he said, but for me, the starting point has to be this. Dario used to work at Google. He is, I think it's pretty obvious to everyone involved probably sympathetic to the political left. And this guy has converted this entire generation of folks who were skeptical about the national security establishment to enthusiastically supporting the military. What a gift he is in that regard. And so to demonize him is just damaging to Silicon Valley national security, to Silicon Valley's national security relations in a way that I just cringe at. Because if you can get somebody like Dario enthusiastic about national security, if Dario can get thousands of anthropic employees who, you know, five years ago were probably like in an EA group home talking about pacifism and veganism and get those guys excited about national security. What a gift. What a gift to. To the national security community. What a gift that he has brought that community on board with literally developing autonomous weapons, not using them, subject to no human oversight, but developing them, ready to go. This is so much progress. Don't take us backwards. And so he's of course referencing to Project Maven, where Google was working with the Department of Defense in a much more limited way. And there was this employee revolt and there was this big divide between the employees and the leadership. And Anthropic Dario has been able to get the anthropic employees so far along that in Gregory Allen's perspective, this is something that is sort of a gift to the government and that they should be willing to deal with the rough edges, work through this to get to a good outcome. So we're all hoping to a. To a good outcome. Anyway, let's move on. Let's see. Is it that time? I think it's time for the Mansion section. Let's move on to the Mansion section. Let me tell you about Vanta. First, automate Compliance and Security. Vanta is the leading AI trust management platform. And let me also tell you about Vibe Co, where DTC brands, B2B startups and AI companies advertise on streaming TV, pick channels, target audiences, measure sales. Just like on Meta. It's time for the Mansion section. Don't we have special lights? Wait, we're going green today? I thought the Mansion section was purple. Anyway.

Seen, the sites aren't the epitome of luxury, but they sure beat what shrink calls Hotel F250, sleeping in a vehicle and spending a per diem allowance on food. Typically, the construction companies that work with Target Hospitality offer housing, meals, and amenities to their employees for free as long as they're working. It's an additional lure for tradespeople already enjoying significant pay hikes, with some data center electricians making more than 150,000 a year. Not bad at all. Should we kick it over to Tyler's piece, the short history of nationalization in the United States? Tyler Cosgrove took the daily op ed in the TVPN newsletter today. Thank you, Tyler. And he took us on a whirlwind tour of the history of nationalization. Of course, this is in the news as the showdown between Anthropic and the Department of War continues. We're, of course, hoping for a unified front moving forward and lots of back and forth. Potentially some capitulation recently, but we will see as the story keeps evolving. But we want to know the history so that we can predict the future. So, Tyler, take us through it. What did you learn about the history of nationalization in the United States? Yeah, so it's very interesting. Just immediately. There we go. Nailed it. Okay, so I think. Yeah, so whichever way, like, the anthropic Pentagon stuff, like, evolves, like, I think it's very clear that this is not, like, the last time government deals will be made. Right. Like, the OpenAI deal is not the last deal that OpenAI will make with the government. Yeah. And there was already discussion of the backstop. And when the backstop gate happened, there was a big question about, like, okay, it seems like, you know, bearish, but also, like, should there be a backstop? Like, maybe if it's an important technology. Like, we've done this before. So take us through some of those. Yeah. And I think, like, all the lab leaders are also. They've been talking about this again and again. Right. You've seen. I also say in this that, you know, they're always, like, extremely vague about it. Right. Yeah. You know, we all got to figure out. We got to talk about. Figure out what we're going to do about this. We got to talk to the government, and the government is like, we got to talk to the AI lab leaders. And no one really says, everyone agrees that we should be talking, which is bullish for tppn, because all we do is talk. Yeah. But basically, like, if you think about, like, future regulation, Right. You have this Spectrum of like we do literally nothing, which is like closer to kind of what we have now. Or you do like full scale nationalization. So there's kind of this spectrum. So I was curious, like, okay, in the past, like during events, like, what happens is the nuclear weapon analogy, like the most extreme example, where it is like, I can go build a railroad. Railroads, as you'll tell us, have been nationalized before, but I legally cannot start a company that builds nuclear weapons at all. And so that's like the most nationalized. Yeah, I think there's a bunch of different subcontractors that supply the components that ultimately. Yes, but what I'm saying is that like, if we end up in the, in the air is nuclear weapons camp. Like, you cannot train a foundation model because that's like the nuclear warhead development, essentially. That's like the most extreme. Yeah, I think broadly that that's, that's the most extreme example. That's also what, you know, Leopold in situational awareness, he has the project. Right. It's kind of the Manhattan project version of AI. Right. And AI 2027 also sort of predicted naturalization. Yeah. And they have the timeline. They have, I think it's open brain. And there's also the government. There's kind of these two competing labs. Right. So maybe they kind of stay separate. Maybe they become the same thing. Okay. Yeah. But yeah, I can just list out a few interesting events in US history that at least. Nationalization adjacent. Why are you doing Kring? Okay, so the first one is the Federal Bank. Right. So first and Second bank of the United States. I don't know if people really consider this nationalization. Right. It's very much like public policy, but it is like a public private institution where the government owns actual equity. Okay, so America founded 1776. Within 30 years, they're like, we got to nationalize some stuff. 1917, 91, the first bank of the United States is nationalized. So it was started as a private company. And then the government said, hey, we want to stake in this. I think for this, I define nationalization as just like maybe they're not actually taking over a private company, but they're doing something that is usually regulated to private company. Got it. Okay. Right. So even, even if the government is like founding it as a project. Yeah. So, so like Manhattan Project is like nationalization. I think most people describe it as that, but there wasn't an atomic bomb company that they took over. Oh, that's true. And same thing with like the, like NASA and like the moon mission. Like, the Apollo program was not some. It wasn't a private entity that got nationalized. So. So that's sort of the most extreme, like, de novo starts as a government project. Okay, yes, that makes sense. Yes. Then you see stuff like, during war, you'll see this as a common occurrence. So during the Civil War, there was the railroads and the telegraphs. These were essentially nationalized. Right. 1862, Civil War railroad and Telegraph act gave the president authority to take possession of any and all telegraph lines and railroads when public safety required it. Interesting. Yeah. And then you see 1914, there's the Alaskan railroad. So big Alaska guy. It's a dream for me. Yeah. Yes. So there basically was not any, like, real, like, private investment going into Alaska. And then the government says, like, okay, we're going to step in. We're going to fund it. But then we actually, they ran out until, I think, 1985, and it went back, I believe, to the state of Alaska. Okay. Yeah. So sort of like government funding to get something off the ground that then leaves. And that, I mean, that is another interesting scenario that I don't think people are considering is like, some sort of, like, temporary nationalization. I mean, okay, yeah, it's safe. We're good. And like, yeah, we're ready to, like, release this. So, like, let's privatize it again. Yeah, that happens all the time. There's like, an interesting aspect where, like, grants. If you get a grant from the government, that's not naturalization, right? No, it's like they're just giving you money. But if the government takes equity in return for that. Yeah. Okay. Maybe you're, like, getting a little bit closer. And then at some point, if they own all of it, like, that's very much nationalization. Right. So it's this very kind of, like, you know, blurry gray area spectrum. But then again, During World War I, you see the railroads in the telephone. Telephone telegraphs get nationalized again. Yep. This is just like, you know, just to clean everything up to make sure, like, okay, the railroads, they're very busy. There's a lot of, like, stuff going on. We need to basically just optimize just for, like, military, you know, construction and stuff like this. Okay. So then 1933, there's the Tennessee Valley Authority. So this is basically one of the first, like, you know, federal utility companies. Yeah. The idea is, like, we got to electrify a lot of, like, rural parts of the US they start in, you know, Tennessee Valley, right? Yep. And then they expand over time, but it's still, like, this is a. Usually a private company doing this. I feel like the Tennessee Valley Authority is still in the news because of, like, did the data center build out? Like, I'm pretty sure that they operate one of the grids that is relevant to at least some of the data centers. I've seen them in the news. I mean, we can Skip over. See, 1942, you have a Manhattan Project. This is kind of like the epitome of nationalization. This is the example everyone goes back to. You see, I mean, a lot of stuff during World War II, right. There's the seizure of, like, meat packing facilities, petroleum industry facilities. A lot of these were because of basically, like, union disagreements with the. With fdr. And then he basically said, like, okay, we're just going to seize this. Yes. Some of these were interesting. They happened sort of after the war. Like, the seizure of meatpacking facilities happened in 1946. So that must have been post war. There's still, like, gyrations in demand. There's still. There's still, like, anomalies as you're transitioning out from the government being the only buyer or, like, the major buyer. The labor unions are trying, like, get more control and recapitalize the business and become more of, like, a normal corporation. And so the government stepping in to sort of, like, massage. That is sort of interesting. I would love to dig into that more. Yeah. Then 1970, you see Amtrak. Yeah. So a lot of these are also, like. I mean, you see rail come up over and over, but this was this. And Conrail, which is, you know, same decade, a little bit later. Yeah. A lot of the railroads or, like, rails were basically just, like. In some areas, they're not used as much. So then they kind of start, like, you know, becoming worse in quality. People aren't keeping them, they're not maintaining them. So then the government steps in and says, okay, we're going to run this. 1979, you have the, like, kind of second oil shock right from the Iranian revolution. So Chrysler ends up being rescued via loan guarantee. Yeah. So not equity, but a backstop on the debt to allow them. And again, I mean, this is like, okay, is that really nationalization? But I think it is still like, okay, it's the government stepping into the private sphere. 2001, TSA. I think this is a good example of nationalization. Right. You see, before this, you basically have private contractors doing, like, security screening. 2001, 9, 11 happens, TSA comes in, we're gonna take all this over. You have in 2008. There's a bunch of stuff you have like Fannie Mae, Freddie Mac in conservatorship. You have. GM basically is close to going out of business. They're like a massive supplier, obviously for cars and stuff, but also for a bunch of defense purposes. GSA is a funny one. That one makes so much sense. But it was not on the top of my mind when I, when I thought about nationalization. But I mean, you never flew pre tsa, have you? No, I mean, I was born like four years after. That's crazy. I flew pre tsa and yeah, I mean, I barely remember it. But it is interesting that. Yeah, it was like a private. It was the purview of the contractors and the airlines. It's crazy that until you were 30. It was just amazing. Yeah, yeah, exactly. Nailed it. Yes. So then most recently we've seen the Trump administration take equity in certain companies. So this is like Intel. This is MP Materials. Still very little control of Intel. Yeah, I mean this is very much like, you know, it's kind of just like a grant, but they're taking equity for it. Yeah, yeah. It's much more driven by a desire to just not let a national champion fail. Yeah, Like, I don't think they have any board seats. Stability in Intel. Yeah, no one's really comparing the intel stake to the Manhattan Project. At least not yet. Yeah, definitely. So basically what you see from this list is that I think you can kind of say that there's two axes that you can kind of plot everything onto. So there's like basically the strength of the nationalization. Right. So you know, the intel stuff, it's like very weak. It's not really nationalization. No one really calls it that. Maybe directionally, but you know, and then you've been hand project. Right. This is very strong. This is like, okay, the government is running not just like a specific organization, but the entire industrial process. And then you also see maybe on like a kind of vertical axis is like the reason for the nationalization. Right. So you have basically in the US We've had either like economic distress. We have like military. Right. So like, you know, system. Systemic risk to the economy. We have to save this company. Or you have, you know, this is so important that we're going to lose a war. National security. Yeah, yeah. So yeah, I think you can, you can plot a bunch of these events. Why do I feel like Uncle Sam is going to have some exposure to private credit soon? Well, that'd be. That would be what? Weak economic, sort of like financial backstop. Potentially, I don't know, it doesn't seem too close to that doom scenario. But what are the other examples? Yes. So you can plot kind of all this stuff onto this, like quadrants. Right. So if you have like, you know, weak and economic, maybe that's like the intel deal, the empty material stuff, where it's just like these are very important that we make sure these don't go out of business. But like, you know, it's not like we're stopping and we're running these companies or even like the Fannie Mae Freddie Mac maybe that's stronger. But it feels pretty, feels pretty weak and it feels very much not tied to national security. Whereas TSA feels like a national security question. Yes. And so much more like militaristic in its thrust. So another example of weak military I said was like during World War II you had kind of a takeover of the GM's production lines. So it's not like, you know, we kind of threw out GM and said we're just seizing everything and we're going to run it ourselves. Yeah. It was still, you know, kind of like a contractor thing. Yeah. But I think that was still like, you know, directionally. Yeah. Nationalization. Right. But the Defense Production act allows the government to jump to the front of the line. Yes. And say you have to produce for us first. You have to produce to our specs. You have to, you have to treat us as your most important client on day one. And then maybe if there's extra capacity, you can still make your old products. Yeah, but even then it's not like they're taking the profits or stuff like this. Yeah, yeah, yeah. So I think like strong economic. You have. It was kind of hard to think of a good example in the US because I think people are generally like very anti this. But I said something like the payment system in China where you basically have every single payment at some point goes through like the state institution, they can kind of see it and they can do whatever with that information. But I thought that was kind of an example of like strong nationalization where I forget the exact year this happened. But it was like, I think within the past decade you have all these kind of like Internet native payment companies and then at some point they kind of step in and say you have to pass this through us. And there were discussions around that a cbdc, central bank, digital currency, fed coin, something along those lines was discussed but never really got off the ground. And so the usdc, the United States dollar backed stablecoins wound up being privately held by a number of public companies, actually. And so we certainly didn't go the nationalization route there. But it was discussed at one point, I guess. Yeah, yeah. And then in a strong military, you have Manhattan Project. This is a pretty canonical example. So if you map this back onto AI, I think we can say, you know, Leopold is kind of in this Manhattan Project camp. So we can say he's probably, you know, strong military. Right. In this quadrant. I think if you're plotting other people, you know, Dario. Dario's talked a lot about the, you know, economic impacts of AI. Right. You know, 50% of, you know, white collar work, entry level, white collar work, you know, maybe may go away. Yeah. You're going to need some response to that. Yeah. If it's the government, like, okay, if you're going to do some sort of ubi, like, where do you get the money from? You need some kind of involvement of the AI company. So I think he can maybe. Maybe you can just. I mean, this is all very much tax profits, potentially. Sure. Yes. But. Yes. Yeah. You know, this is all like very. Go further speculative, but, like, I think maybe he can be, you know, for stuff like that. That's broadly, I would say, in kind of the weak economic kind of quadrant. Yeah. I wonder. I wonder. Yeah. I mean, you put Dario at one point sort of in, like the weak economic side. Like, he's thinking, like more that the next likely step or the one that he's talking about right now is something that's more akin to the 2025 intel deal, less Manhattan Project. But it still feels like the fact that his favorite book is the book that he gives out is the Making of the Atomic Bomb. Like, he has Manhattan Project on the brain. He has, yes. I mean, so he's talked about, like, working with, you know, the military and stuff, but when he talks about, like, government intervention. Yeah. It's almost always in the context of the economy. Yeah. So I think for that reason, I kind of remember there, obviously, like, this is very, you know, it's unclear really. I mean, it is. It is in the context of the economy. But at the same time, when he discusses giving Nvidia GPUs to China, he says that's akin to giving nukes to North Korea. And so he doesn't shy away from the Manhattan Project analogy as much as people think he has shied away from it recently. But, yes, it feels like he. And that's why I was sort of against you trying to, like, put him on this chart in some way, because I think he actually has considered all of these and sort of it's much more of a matter of timelines, like what he thinks will happen soon, what he is in favor of, what he is against. A lot of that aligns to who is in power, what the administration is doing, what the plan is. But I think he sees all of these different scenarios as potential outcomes. Some of them he'd be in favor of, some of them he'd be more resistant to. Well, I can't wait to watch his new interview with the Economist. Yes. Which is sponsored and presented by Anthropic. That was a funny. That was a funny move, but do we have a video we want to pull up? Well, it just came out a couple hours ago. Okay. There is a clip in here, right? Anthropic CEO, I apologize for Leaked memo from the Economist. Dario Amadei says he's sorry. In his first interview since the Pentagon labeled Anthropic a supply chain risk, the first American company to receive that designation, the firm's boss offered a mea culpa for the way he handled a crisis that he described as one of the most, quote, disorienting in Anthropic's history. So we can pull up this vertical clip from the Economist. The Economist getting in the vertical video game was long overdue, but welcomed. I am very excited. I've been a fan of the Economist for a very, very long time. And so we can pull up this clip from the Economist in the timeline and play this for the audience.