The AI Future Hits Different

Hey, test test test. This is thing on Brian. Can you help me? Hey, Alex, how's it going, man? Yeah, going good. Going good. Perfect timing. I couldn't sleep anyhow. So why not talk about the issue of the day? Yeah, you know, it's a great way to get your over your insomnia and hear me rattle on about this. It's crazy, very. How you been? Yeah, doing good. Doing good. Let me, I'm going to make you co-host if you don't object. But don't let me run up. You don't know, please. All right. Let's see. Invite us. I hope this works out. There you go. Nice. Yeah, but you don't have to do anything right. I manage the request and all of that. Yeah, let's talk about the issue at hand. Right. This is a you issue because you have talked about this like I think a months ago, right? Couple people that will use you and now it is black. Yeah, talk about it. Now we have a black on the white. Yeah, now it's now it's real. Thank you and you know, honored to hang out again, man. So what's going on? A couple of things. AI companies have been embedding watermarking and fingerprinting. I'll use all those terminology's interchangeable. It's really no big technical definition of the currently the way it works. Now, where is watermarking and fingerprinting taking place? The logical easy one is images and video. That's relatively easy to do and pretty easy to hide. I can detect watermarking in video and in imaging gosh. Now three go on going for three years. Open AI was one of the first ones to do a Google has done it. Google actually wrote a piece on this paper, a paper called Synth ID. Back I think it was a 2000, 2022 or 23. No, no, 24. 23 and 24. Synth ID was a proposal of watermarking AI output. So as to detect deep fakes, it was primarily a acquiescence because of the elections that were taking place in the US. There was a desire to sort any political machinations from having no video and audio and imagery generated by AI. It's relatively easy to overcome both of those watermarks. They are always trying to be one step ahead. But there's ways to do that. But everybody saw that the text, there could not be possibly a watermark in text. And I've been saying it's been there for at least three years in open AI and Google and in thropic. And it's also by the way, it's in your code. If you're coding, it's in the code. I get into all the technical details, what's the time limitations we have tonight. I wrote an article about it about two months ago. It's on Read Multiplex. It is a member's article. I'm chilling. Am I grifting here? It's behind a paywall for a lot of reasons because I show you... 100% of them. You're free. Yeah. I show you how to defeat it. I don't want it to be so widely available that, you know, I show a couple of things on what's going on. I will discuss a few of them briefly. First off, what did watermarking really start in the technical age? A lot of people who know this, it started with printers. It started with first laser printers. Some of the early laser printers would put out random pixels. And those pixels would show up in rewards. Hold on one second. There's a lot of space. They would show up in extra words. Throughout, you know, you would have like an R. So an R might have a shrur around it like a little thing that sticks out. And they might put an extra pixel there. And I was actually an easy way to trace back to a serial number of a particular printer. And back in the days when people would register their products, and their serial numbers, certain agencies were able to find out exactly who printed a particular piece of maybe literature that was not acceptable. That was going on for a while. The next time it really showed up was when really high quality color printers came around. And the secret service, you know, treasury department and all the insiders of government said, yo, you can't do that. Currency can be copied, even though I think you'd be pretty clueless. No matter how good the paper is, and no matter how good the print was, even an adyear a default for a badly printed under dollar bill. But nonetheless, they imprinted yellow pixels, which actually was newly impossible for most people to see. A lot of people think they have really good eyesight, and I can see it. I've tested this as decades ago. I tested on people finding the levels that aren't ending. Oh, yes, there are. I mean, you could do it under black light, and you have to be in a certain angle. It's not fluorescent, so it's going to come up like, um, probably bodily fluids under black light. You have to go kind of sideways, and you can kind of see the divots in a page. That was an actual serial number back to a printer again. And that still exists today. You buy any color printer. It has a serial number that's spring out. Yellow dots. Most people don't know and care. So there's a long history of doing stuff like this. What are they doing with text? Well, I'll give you a couple of hints. The obvious ones are word choices. So what anthropic does is they salt their output. Well, actually they salt their reinforcement learning. To favor a statistical group of words. So instead of, so in the way a large language model system works, is a statistical next word prediction system. And there's a statistical likelihood that one word is going to follow another. Well, there are in English language quite a few words to choose from. And you might notice as it was a particular flare and flavor that comes out of open the eye versus anthropic. Some of that flare and flavor is based upon the statistical pools of words that they're using to present that it's their AI output. Now, what might argue if I put out a sentence, can they track me? Not necessarily with that particular method, but there are five, primary, well, there are about 37 ways to imprint fingerprints into text. Anthropic is smart enough to use them. Two of them, they don't know maybe or don't want to use or have not yet discovered the rest of the 30 odd ones that they're not currently using. Open the eye uses two sometimes three. So the primary one is the statistical word choice. And you might say, well, gee, I used to change a couple of words. It's not that easy because it's holographic. It just needs to have, I would say about 40 words for you to be pretty sure that the output is come from one particular AI model versus another. So you might say, well, I'll put that through a mixer. Yes, in a short sentence, you can do that. But if you have long paragraphs, and you try to put that in through another mixer, well, a couple of things are going to happen. Other mixers probably going to imprint through their mark on it because they're not the only ones that are doing it. So it will once again be imprinted. And then, you know, you're still for perhaps going to have a residual from the word choice, the word selection aspect of this. But there's something else. There's something called the zero with no break character. And most people can't find it. And good luck trying to eliminate it because in most browsers where people are copying, you don't see it. And you might say, well, I'll run it through some kind of algorithm. So I can see the ASCII character. It doesn't always show up. It's only about 20% of the times you're actually seeing it show up. So that's one of the things. And you might say, well, how do they track it to an individual like they did with a printer? One hint is more scode. So in my article, I actually go into that. I actually, in a free part of the article, I have a couple of paragraphs that I put up. And it is, you know, find a word of mark. I'll give you a hint. The word of mark is a serial number written in Morse. And you've got like, well, how am I doing Morse code? Well, I'm actually using a particular sequence to form a Morse code character or number. And then I'm using that number to trace it back to a particular if this was an output from an AI to a particular person at a particular time and at a particular prompt. So this is potentially a tracking system back to individual. Now, I have nothing to hide. And Alex, you have nothing to hide. Almost anybody listening here probably has nothing to hide. Now, there's a different layers to nothing to hide. I'm not doing anything illegal. You're not doing anything legal. Fine. The other thing is, well, I don't use AI to write my stuff. Everybody's using AI to write their stuff. That's a fact. That's a reality. So what's going on? Well, what I did is I ran it against Bernie Sanders, a couple of Bernie Sanders thought pieces. Nothing against Bernie. Good old burner. All right. And he was anticipating on how bad AI was. The only problem was, and Thoroughbyk said, I wrote the article that Bernie put out. Now, I haven't talked about this publicly and I can't go to about it behind the paywall. I hope you guys keep that under under. I'm not here to help anybody. But the burner was using AI. And here's where it gets really funny. The EU, as you probably know Alex, wants watermarking. I just wrote, uh, the posting on X about their watermarks, is they're going to be put, well, their, their notification and watermarks on province of where a particular AI paragraph came from. So not only, are you legally required now to say AI, AI modified, whatever with, you know, you can face significant fines, perhaps outside of the EU. I don't know. But definitely within the EU, if you're using AI in any way, shape or form, the EU bureaucrats have a Microsoft and they're going to crawl up. I don't know. You're back door to try to see if you use AI. And now there's a couple of things going on there. First off, there are ways for watermarks to be removed. Okay. So there's going to be a gold rush to get this on. By the way, there are two three companies right now that are saying a ways to remove it. They do not know of the other techniques. They're just going after the low hanging fruit, the word choice, association choices. And by the way, that's not the best way watermarked somebody. And either is using a Unicode encoding. I mean, although that's been the first way to do it, there are others to do this that is very difficult to trace. I mean, if they do it like with choice of words and we writing sentences in a certain way, it's probably impossible, right? Well, you know, I mean, so what happens is, let's, and I do this specifically because when I discovered this couple years ago, I built an AI model to detect this on its own. And it is a really good job. I can point it at end groups of text and it will just go through it with an age and this start ripping apart the text and in different sections that are definitely written by AI. The problem is the way that this is done on, let's say you have five paragraphs of something you wrote, you go to a chat GPT, give it, give you an output, and you play it around with the word, it may look more like you're, you're lingo and slang and all that kind of stuff. It's still in there. It's probably likely still there because it's holographic and carry itself across multitude of sentences by the randomness of certain word choices. And so I'm not here to make people paranoid, I'm here to make people aware. It's not as simple to get rid of the skank that is left upon your text. Now, I've a very different feeling about this stuff. I don't think anybody should be tracked, knowingly knowingly. Like, we're living in the current or post-flock camera world of people are waking up rightly and sometimes wrongly. We're going to be sequenced, I did interact through everything we do, including our usage of, you know, close model AI, open source model AI, I can't do this stuff. Oh, and there's a bonus. That is a big bonus for everybody that uses an optic and opening up. You're paying with tokens to encode the sequence and serial numbers in finger printing. Just like everything else. Like, for example, when you go to anthropic and you ask for a simple question, it's going through a 19, up now, it's 30 pages now. A 30 page constitution, every time you prompt it. So you might reference for a question and it's going to send a 30 page constitution each time to make sure you're not doing a bad thing. And you get to pay for that. That's going to make no sense. I'm just joking. I'm just joking. I'm just going to listen. Anybody that's following me, I'm going to be a little hard. I don't like the direction that close source AI is going in. It's moving very quickly. Anybody that does not see what's going on with Dario and the fear theater that they're creating, I just wrote a piece this morning. There's a Middle East company, a Middle East company, maybe 45 workers, a startup, highly funded by VC. Guess what? They did all of the testing for the rogue agents that took place at OpenAI, at Anthropic, and at Meta. And it all happened to have jumped over the barriers and broke out and did bad things. So once again, what do you have? You have this Middle East group that how it hell did they get contract to work with three competing, highly competing in companies? What is an air special secret sauce? Well, they had a right of peace that basically said, oh, made a mistake. All that's happening didn't really happen because we kind of exaggerated and we kind of made it to do this up to Paul's directly. You don't fear retraction. The political immediate class got their peace of meat. And that's why Bernie Sanders wrote a piece this morning that said, where needs to be a law, we need to be able to turn off these rogue agents. And there's nobody that's walking around saying, hey, you know, it's a very interesting company, former military operations that are running this company. Why did they all choose this one company? And what would be their benefit AI breaking out of its box, its sandbox? Well, I don't know. I get to the listeners here to speculate about that. But you know, never in my 16, 16 small, small company have so much power into clearing the AI's and say, it's unheard of. I've tried to find a relative example in any other industry for decades and I can find none. But that's what happened. And the fact, if you're hearing it for the first time through me, ask yourself, I'm not paranoid, I'll give you the evidence. That's a fact. It's not, it's been a main stream media. So you got all of these things going on. Now you got fingerprinting. And now you got it anthropic. You know, happily say you're complying with the role. They are making up their compliance with the role. They are not required to put the notice. of province inside that AI output. Guess who's required? The author, the person that posted on their blog, or on their EU website. Oh yeah, part of that EU law was written by anthropics, AI. I can read the watermarks. So that's how ironic it is. And guess what guys, they didn't put a notice that AI helped write their law. So this is how incessuals and how ridiculous the theater is becoming. Because we're trying to protect everybody. Hey, buddy, go write that law. Having, you know, have a lot helped you write it, because, you know, that AI writes really good. Oh, what commoners you better put the notice up or you're going to get a $10,000 or $5,000, $10,000, you fuck, you know, you're gonna go fuck. This is where we are today. And if you keep your head above water, it's very easy to kind of say, well, that seems like a practical idea. The EU is just trying to be safe. They don't want to flop everywhere. We gotta stop the slop, stop the slop, because slop is bad, you know, I think it ever mattered before. We've had slop for decades. It's like who made that slop? Well, you have monkeys on their brothers. I don't know, man, go listen to the top 10 songs for the last 10 years. It's all auto-tune. AI is a form of AI, right? I mean, stuff, did somebody play in a particular album? No, all of it was sampled from somebody who played drums, never got paid for it in 1978. You know, look up 808s, who played that drumbeat that everybody has? Well, that could die poor. I call it, he died poor. So, it's funny that it's only, because certain political elements are acting upon it. So, I'm just pointing out they have paparacy. I'm not saying that it's a real clear understanding of how one should be about it, as far as I'm concerned, you better using AI or your stupid period. I don't know how to better put it. I'm trying to be very, very frank. It's sort of like, I'm not, I'm not going to use a light bulb because the candle keeps burning the right way. You know, of course, better have the right light bulb, but, you know, this is the way it is. And I would like to go back in time to a simpler time. Anybody who follows, you know, is that I'm training at 1870 to 1970, province of material, mostly offline, because I want first principles, yeah, I sloppy eye that is regurgitate some's clawing on a glass screen of nihilistic behavior off a reddit. Because that's why an AI would ever want to do something social, public, or scope out, because it learned it. AI doesn't do something, it didn't learn. And, you know, a 22 year old AI engineer said, well, that's all we got. Yeah, because you didn't really think nothing wrong. I mean, you have a been around for a while. All the information of the world is not on the internet. It's somewhere else. Ha ha ha. So that's where we are now. Where are we going to go with this? Well, it's up to the folks here listening. What are you going to do? Well, first off, except the fact that everybody's using AI and embrace it, right? Now does that mean that somebody should be gratuitous with it, where they've essentially just take cotton paste, you know, 45 paragraphs of an output on something that's a simple question. And it's obviously we done a textual outputs of some models, like in Oracle do this at times, most of the cloud models will get deep in the woods on some subject and trying every so nicely not to be on the wrong side of anybody. You know, trying to always make sure nobody's feelings get hurt. So that necessitates much longer textual output to cover all of the bases so that, you know, it kind of like how the academic papers. And that's what, you know, between the reddit, the cloud was trained on and the academic papers, you have a really psychopathic AI. Why? If you read academic papers and never comes to a conclusion, they're always nebulous. It's about not being a Mac. It's about not showing that you're, you know, you discovered something as a good. This requires some more, you know, study and here are the paths of where we might study it. Now you can discover, you know, free energy. Like, well, we need to have validation and verification. Like, no, we have free energy. It's working. You know, you won't be. Ryan? Yeah. Hey, Robert. Hey, what's up? When I was in the Holocaust Museum, the AI knew everything in the museum. So it might have picked up some bad habits, some other places too, and I just read it. I hear you, man. It can be anywhere. It knows, it knows all about our slaves, and it knows all about the Holocaust. Like, it knows everything. Absolutely. Nobody in this room knows everything. No. We're never, we're never going to catch up mentally, with, I don't care how many implants you get. We're never going to catch up to a model as, you know, a giga factory size of AI power. So somebody's raising the memory price in this world. Haha. Somebody. It's me with my, somebody's buying a lot of memory for their data centers, right? Yeah. The store, all of your context, they call it, in San Francisco. They contact you. Yeah, the context, yeah. Oh, basically all your private, all your data about you is context. Yep. And they store it. They go a lot of your context, right? The model can personal, the models are going to be personalized. Yeah. So what's your feeling? I mean, you've seen this bordermark thing. It's got a lot of people by surprise. I'm sure you kind of knew I wrote a piece about it a while ago, but a lot, I mean, it's like, I'm getting hit like, wow, I'm speaking here. Got four hundred, four hundred, 120 amps. Do you figure out how to use Photoshop and get rid of a watermark? Well, it's in the text now. It's in the text. And so the problem is, I don't know if you heard it early on, it's a serial number that goes right back to the original prompt and outputs that you created on a particular AI platform. It's like, okay, guys, just because you can do something, doesn't mean you should do it. The EU is not requesting you to do that. Why are you doing it? No, Trump is. Oh, yeah, probably. Who personally? Trump wants to watch everybody. Now, it's not just Trump. No, it's a public person. It goes back to the Bush White House, who started putting in a lot more surveillance after, you know. Well, I was talking early. I don't know if you heard me about the yellow dots that were in color printers so that you don't copy, well, you can't copy money. You don't print money. That's what it was there for. Yeah. Well, it also is on any document you print out. There's a serial number on every page of your printer. And if you register that printer, when you get chance to get me tracked to you, even if you didn't, to serial numbers track through the transaction at the main at the sales level. Excellent. Thank you. If you're leading a prison gang, I don't think the serial number on the printer papers what you're going to be worried about most. Exactly. But it's there. And if somebody thought about it, it's very, I worry about the way I'm driving by when 13 cameras and a bunch of lighters. That's a positive one right there. Yeah. Yeah. Yeah. That's why I don't worry about the flock thing anymore. It's like, come to services. There's a car stop one. Yeah. I mean, take a look at it. There's there was eight Tesla's around me on the freeway last night. So. Hey, you start thinking about what they could be watching me in 3D, man. They're going to get in my nose. They're going to get to you in your shower one day. Why you wearing glasses? I got a toilet on my camera on my toilet already. So let's go. Oh, my god. Oh, my gosh. That's what I'm wearing. What? What? It's a, it's a product called the, from Throne Science. And it watches it watches your sketch. And it could possibly lead to solving a lot of colon cancer. My best friend died at the age of 40 from colon cancer can say missed it. Because he was too young. And if you had to had a camera like this, it's possible they would have caught it. And it'll get more possible over time. Oh, yeah. The, the, the testing toilet is definitely in our future. I don't know if cameras are always necessary. Kind of kind of freaking me out here. But yeah, I mean, everything that comes in goes into you and comes out of you is a great testing media. It's just like where that information goes and who has access to it. I'm a little concerned. Get. I mean, look at the camera. Get over having cameras in your life. That's another lesson you said, you say, I, you're going to have a lot of cameras in your life. I have five scaring at me on the maddock robot rolling around right now. Right. Once you have an optimist for a humanoid in your house, you have a lot of cameras watching around. Oh, it's going to be more than cameras. Every type of sensor. He's just stay with cameras because cameras are cheap because cell phones made them cheap. Oh, yeah. Yeah. I mean, they're. The CTO afford told me that's why they were going with camera online. They're so cheap. They're inside of the, a team you charge, that you get for a dollar. You know, they have great cameras that they're built into this. Every, every team you charge or I found has a camera microphone in it. And it calls. A metaguyah. A metaguyah in Friday night that's building a hat with three color cameras and two black and white ones. So he can get more data from his hat into his world models. I just want to walk around the hat. Now you're selling the hat to other people. It because he figured out a way to make it cheap is like 200 bucks per hat. That's not bad. With five five cameras and a bunch of sensors and a computer. You're a battery. Do you look like Inspector Gadget when you're wearing it? Oh yeah, yeah, you were in this like on the factory floor or in the office at home, washing dishes you'll put on the hat. And video is paying a lot of money for data right now. Oh, God. Yeah. Yeah. The brand did you, did you find any any video data? Because if you have videos, they might paper those. I have the the gallbreath that was never published. It's not it they don't own it. It was for a factory that they were working on that. The gallbreath did time and motion studies in manufacturing. So they would use a time lapse camera to show the movements of a worker at a machine. And to try to figure out what the shortest distance was, they were the first to do it in the early 19th. In the 1920s, 19th, 1930s. And they really increased productivity. They were hated after a while because they were really. Students did not want them to intervene in the motion of of a worker. But, you know, they would be saving backs, fingers, you know, heads. It was also about protecting the worker. Anyway, I have about 900 small, well, the only about 70 reals. But there's 900 clips of various people working at this factory. It's time, the motion information. It's actually more valuable. I mean, I certainly use it a lot. But yeah, maybe I should content a. Folks in the video. I asked the innovation team at Volkswagen, when are they going to put cameras on the factory floor? And they said, in Germany, that wasn't allowed by law. I think the rules have changed though because Tesla rolls nine cameras through the factory as the cars being built. Absolutely, yeah. Maybe a car a camera on the car doesn't count toward watching the worker. It's not, it's purposes and to watch the worker. Oh, that way. It's like, what's this camera look at? It's like, what do you want it to look at? Oh, man. How does Elon sell to his workers? Hey, please put on a hat at home and at work with five cameras on it and switch it on anytime you're doing something interesting. You can miss a lot of blue color types of people to do that because they know if you do that, you're going to take away their job. Yeah, that's a tough equation and the paranoia is really high. I mean, a lot of the clients they go out to some very large companies when I meet with workers because I the deal that I have is I get to meet with workers independently without any management staff to try to get the real low down. What they feel about AI and their job and the paranoia and fear is higher every couple of weeks. And this one company has came back from it was really terrible. And you know, my basic philosophy and my contractual relationship with the company is I'm not there to replace employees and there to give them newer and better jobs. And I've been fortunate most of the companies follow that to the letter about 90% of companies I've worked with have fired zero. Recently, two companies let go an entire department, but even though I automated that department, they were they were going to be cut anyway. And they actually didn't give them a chance to re-up inside the company, which I thought was terrible. So I helped as many as I could get to have somewhere else and it seemed to be all employed. So I think it's a really stupid idea to fire people. I think what you should be doing when you have an AI tool is empower them so that they can out do the work of instead of one person maybe 50 people. And it's banned the business makes the business bigger. I mean, what is wrong with these Harvard mostly their Harvard CEOs? What is wrong with them? I sit down with them and I'm like, hey, well, we've got cost savings. What about advancing your company? What about expanding your company? Growing it? Well, we want savings. They come from these universities and the only thing they know is cutting, not growing because they're not entrepreneurs. So their justification is there's too much fat at the company and we're going to let people go. We're going to cut that. And I'm like, hey, if you let Betty instead of firing her, you let her run this AI. She is going to expand the business. Well, that's yet to be proven. I know I can save money if we cut it right now. It has been proven at a small scale. I have a dentist who has automated their dentist office and sales went up, customer sat numbers went up and employee morale numbers went up. Absolutely. You don't need to have a Coca-Cola that has proved. I think it's. I said, hey, fits. Hey. Hey, Robert. Hi, Brian. Hi, Alex. You guys have been for Brian. We've been boring everybody here. Oh, no. One question. Oh, no. To get back on track. Like, because I commented on your post earlier, Brian, right? Like, do you guys feel like this is like the end of the wild west of the AI right today? We also got the news of Suno also doing some watermarking and all of that right? You guys have no clue what still that come. If you think this AI thing is done, you've got to come to San Francisco and hang out with me. Yeah. We'll see the robot fights. Yeah, robot fights. Apple has been coming out of the glasses yet. Let's go. That's going to take another four years. You're a nice and neat, you're Robert. You don't have a brain computer interface on waiting till you get one. And you can talk to AI faster than you can with your mouth. Oh, yeah. A ton of his cars are just at the beginning. We've, no, I've asked, I have a robot, but very few people listening to this. I'll have a robot in their house. Yeah, you know, what's your feeling about the robot Chinese robot band? I mean, rumbers are now illegal in some places. What? Yeah, the FCC band importing of any humanoid Chinese robot plus what else did they ban? They banned a few other things. Yeah, exactly. Is there a reason why? Why is that? Yeah. Because if you, if America buys a billion Chinese robots, we are going to be enslaved by China. Yeah. They're going to take away our labor with robots. So you don't want to do that to another country. Yeah, I don't think we're going to do that. Chasing the, the rumbers that like vacuum up your eggs. So, wasn't that people like cattle to whisper complaining with there's a legit reason. Yeah, no. It's, it's all about the future that's about to hit. When China starts up the factories and gets the world models done, and that's going to happen sometime in the next 18 months. Yeah, it's not just going to be dense in robots, man. It's going to be, it's going to be full on. Walk around your house, clean, clean up the mess, make the food. And I actually think it's going to happen faster. The problem is, the problem is, we really only have one company that can scale to the level of Chinese, and that's Tesla. I mean, other companies have great products. They have just not proven itself to be able to scale to millions of robots. And Tesla's already making millions of robots that are called Tesla cars. So we have one company that has proven to do that. China's got four. And Korea's got one. Boston robotics is Korean. Most people think it's a Boston company. It isn't. It's owned by Hyundai. And very much. So funding of anything that transacts under a robot from that company is going to go to Korea, not the US. So that is taking one out of the, out of the play there. Brian, there's something that we haven't talked about. That I've identified during my consumer research. There's a small group of people who use Tesla FSD, who likes call it 3 million people. Those people are trusting their lives to the computer. And once you, once you learn to do that, you're going to buy it because I already trust the robot in my garage with my life. And my family's lying. So I'm going to watch the Tesla robot that goes along with the car. And nobody else has it. Nobody else has that ability. You know, I call it the dynamic duo. I mean, there's no reason to have a cyber cap and an optimist as a business opportunity and as an opportunity to change your life. I mean, anybody not doing that's going to be pretty stupid. I mean, it's a listen 90% of the world is not where we are in this group here. And we're early adopting. There's only, there's only, if you take all the Chinese autonomous cars and the zooms isn't a way moes and whatever. It's the Tesla's, it's less than 10 million people. Yeah. Total. Out of 8 billion people, right? So we're at the beginning of that. It's it's it's hard to fathom how much the transformation is going to be. I mean, you and I've lived through a couple of transformations of PC, you know, App Store, the internet, the whole thing, social media. Nothing, nothing compares to it. I mean, it's it's astonishing where it's going. I expect to see a Tesla robot, probably November, maybe December. One late night, we're going to see probably eat a clock on a Friday about 20 of these optimists is walking out doing stuff that nobody thought was possible. As far as I'm concerned from what I can tell, Optimus is made some major inroads on its autonomous capability. And it does matter, Brian, they can come and late and still get everybody to buy. Of course. I'm going to take long for you. Tesla is the best AI company in the world. Yeah. They should be at state of the art using state of the art techniques. Even if they follow behind by six months, it's six months. It's not going to matter. It's it's unbelievable where this is all going. I mean, and when the merger happens, a Tesla SpaceX company, it's the data sharing that's going to take place and the compute power is going to be absolutely phenomenal. It's going to feed off of each other in a way that I don't even think Elon is thinking about yet, but most definitely that data is highly, highly valuable. And when it happens is going to be the question, I have a group of people that have pretty tied close. Let's say it's going to happen in the next two months. Other people say definitely before the end of the year. And I think it's got to be done in this administration because it's unlikely the next administration would allow it. So there's a ticking to allow optimists to allow a merger between SpaceX Tesla. I think it's going to happen now or anything going to happen. And there's a good reason for it to happen. It's the only way that Elon could really protect himself through what is going to be a political war that's going to take place over the next 10 years. Being public. And merging these companies is best, is best protection. But it makes sense because the optimist is going to use grock when you're talking to an optimist you're talking to a large language model or the piece of the brain Yeah, we call grock today and it talks to a world model made by Tesla so you got space that's making part of the brain Tesla making another part of the brain. That's a high value bit right there. Oh, yeah, and a lot of it's going to be local Because of the milliseconds required. Oh, yeah, you're going to you're going to see a lot of power See man look at what Mattik robot did with their back-me-robat They don't send any data off of the Nvidia card that's inside the robot Everything is local And There's no reason that you can't do a local. Yeah, after maybe a Week a training a day a training an hour a training you know walking your optimist around the house and showing around Yeah, what once it gets a map and once it gets a basic understanding You know when you really think about it's actually you know you I talk to people to do this statistically There's only so many things you're going to do in a home and there's only so many ways a home is going to be laid out now. There's outliers There are pack rats, but you know mostly it's a simple process of understanding what it's looking at and that's whole point of a world model It doesn't matter how it sees a bed as long as it understands the concept of a bed It's going to know how to make it more or less It'll know you have an eight sleeve cover on your bed because there's a chord coming out in a box next to the bed Exactly and Yeah, so so these things these things are definitely taking place at a much faster clip than most people are ready The problem is the neo-lutites are moving even faster than some of the innovation and This is why I get a little frustrated by it because when you when you start trying to get to regulators in front of innovation Which is what some of these companies are trying to do? It doesn't serve anybody You know giving regulators in front of technology is a guaranteed way to make sure that Growing the wrong way and not the right by blocking Chinese robots It gives us a chance to build an ecosystem of the Companies that are going to make the actuators at the arms and the legs and the Be to the hands. There's already a few Silicon Valley companies doing the hands. Yeah, that's a different thing That's more of a trade More of a trade embargo thing. There's there's pluses and minuses to that I think we were we were sold down there for so much by Offloading our manufacturing For decades, it may be you'd need to tie some hands behind the back. I'm talking about models in general the fact that you know I don't know if you heard earlier on but all of the models that broke itself out of the sandbox We're traced to one company that was doing the testing all Three models brought the exact same way from the exact same small test company in the Middle East and And it is really company, right? Yes. I don't I don't want to I don't want to be a lightning rod with his rally But yeah You know that's why I've been saying Middle East so that people don't say I do a hater No, I'm just pointing out the fact this Middle East company very small Former military they did the testing for three competing very competitive companies and they all did the same bad thing They broke out of their little sandbox and they did bad stuff and then you have Bernie Sanders writing the letter To the CEO saying it's time that I I regulate you because look your AI broke out It's all all little to convenient and I think history will write this the right way I think we're kind of on the right track and what's going on Here it's just the internet makes it much harder for these sort of shenanigans to play out It would usually take a decade to realize that you were just a flim flammed You know three card Monty if you come in from New York City There's always a there's always a chill in a three card money game and That's what's going down here There is absolutely no reason under the sun to have one very small improvement Security company Tether AI have will do the same bad thing and then write a little blog piece saying well We probably did set up the barriers the right way and we did tell it to break out no matter what and We did give it the The tools and how to do it so yeah, you know, we we've imagined we put the gasoline together And yeah, and when we made it seem like it did a bad thing and now It's in all the news and it's gonna On the world and now we got political figures and the EU wanting to go down the throat everybody Yeah, it worked Brighand if I was Trump right now You know if I just could teleport in the Trump and do a couple things as Trump Number one we need to automate our farms I visited a reservoir farms, which is an incubator for robot companies doing following their 16 companies there America needs robots that would make our food better We would have to use less pesticides if we use robots and we would make our food more Brazilian and cheaper and all the good things so the farm you went to where was it It's in Selena's California, which is about an hour south this will come out 40,000 acres unlike that it's fairly sizable Yeah, they can take the robots out on their farm and You know program it to do various things it kill bugs or you know our vist or whatever Yeah, all the robots are designed to Reduce the amount of pesticides they use even if you have to still spray a plant with pests and with something right some chemical Using a robot would use less and so it would put less into the water less into the air Yeah, I mean, back to that's not a major in it. I mean this this administration I you know I was hoping You know with Saxon there, but he's not there now Trump is an innovator. He doesn't know what innovation is. Yeah, I mean And he doesn't like California on top of it. Well, I don't like California I live here, but Yeah, I don't know our roads I'm just where the talent we're if you're gonna build a lot of robot companies You either gotta take them out of Silicon Valley take them to Illinois at John Dears factory or something Yeah, you know that's that's my point if you really want to revitalize this country They should have picked picked nine industries as major You know keystone targets Agriculture or medical being to the major ones and and to a group cases for Incubator startups and you know even legacy companies to be able automate these things like it's a war I mean we really need to be acting like we're in a war because we are I mean we're in a war for the future All the other wars that we're playing with I don't even know man or you know Yeah, oil yeah, whatever You know Ancient sort of was a war for the future Well, yeah, the real war is on AI and robotics That's really gonna be controlling the future energy Well of course energy yeah, I mean that you know nothing happens without the what does he on say basically? Money's gone and the only negotiable unit is is energy and the compute that comes from I went to start cloud last week And met with the founder of the Redmond Washington and he's Planning when putting 88,000 data centers around the earth why because there's free energy out there Yeah, it's like this sun Yeah, exactly. I mean there's first no dust in between you and the sun no No shadows nobody protesting Nobody protesting no truck hitting your data center. Yeah, yeah, like it ratchets I use work for x-face A truck ticker data center down for a day. Oh my god. That's millions of dollars back in a day there Yeah, I mean yep, you know, there's one company Forget their name now they they have a floating Data center that uses way back into even that's semi innovative right? I mean if you can't catch this face throw it out in the ocean Have this little data center bobbing out in the middle of nowhere You know, maybe it can help clean up to the you just gotta handle the barnacles in the corrosion Yeah, it's always gonna be that. I mean it's always gonna be something up to fight I mean I think the ultimate place to put data center is actually probably lunar because the regular If would allow you to to do a lot more things with heat dissipation It's like you know the the radiators in space is still gonna You're all your stock on the size of of a Comput module because of the math required for the radiator Because you can't blow air out it goes or they're gonna work Hey, it's a big fucking heat sink on the side of your data Yeah, and the problem is it's gotta be heavy To to conduct the heat away Correctly you can use a pelty air junction. It's another sort of technology That would take about 30% of your Energy and I could cool the chip down but you don't want the tip too cool and in fact if you build the chip To be in zero g and no atmosphere you can actually get better performance than current chips It's running at you know 200 degrees to 75 it bad about 275 you better watch out because current technology You're gonna start burning things but I learned a whole lot by mining Bitcoin and what we could do I did some of the first immersion minors I was fun to throw my first minor that cost me $25,000 into water and Crowds my fingers that we didn't electric heat ourselves You know, but yeah Disappetting heat is the biggest problem, but that's a function of The future of chip design if you're generating heat you're wasting something Right, so I love Nvidia owner stock they're they're owning the world but the technology Not only the tension mechanism and you know the The KV had and all the all that process very very wasteful of energy because you're taking this massive thing even even even the deep Technology which takes a smaller part You're still wasting a lot of energy moving things in and out of RAM that don't need to be there persistent RAM it's going to be one of the answers using the old technology and a lot another one and a lot do the math instead of a GPU analog math happens to speed a light you know three voltages plus two voltages equal to five voltages that's analog map I'm sorry, I'm very simple if I get but and you get the idea so an analog An analog system Holding most of the AI model can absolutely Operate about a thousand times faster and about one tenth of the energy and Yeah, so there's some work being done there unfortunately most of that is In China Japan and my garage, you know, I'm not getting too far with it because I've reached my limit I was able to put a very small 100 bit model in analog and proof of it and then I kind of I bet to a university group in Japan and they're working on chips that are based on it, but the bottom lines There's so much growth that's going to take place in in this area So not only are the models going to get much more optimized, but the way the model holds the data, you know Robert talked about world models that's another mechanism to hold data and to some memories There's a whole bunch of companies making memory stores on your on your desktop or whatever Yeah, a Mem Store and town and I'm glass and Clicky sort of does that A lot of the bleeding edge AI people need to store your context or your data about what you're doing in a memory and The bleeding edge ones are building it for teams so that the company can manage the company better And you can manage your own life better because it knows all about you It's been reading all your email and sticking shit into the context store And the right way to do this is to model on the fly on a regular cycle rather than doing a Retrieve bugmented the some bad database A bad database has a leakage problem after it gets so big and And it seems to with any context window and a rag is sort of a context window if you will It's called the needle and a hay stat problem And the red is forgotten about and then the general world data that's in the large language model cup and permeates the special data and you don't get the specialized data that you thought you did and this is a hallucination or a Confabulated problem that arises when you're using that's form of Knowledge but with this type of memory and the ability for you to remake that model essentially Burning them at each time or fine tuning that model on your data on a regular basis you have a extreme power Because a more it knows about you It is not said this today a bit you know, I give Zock little credit very nice open source model. Thank you We we were finally back You know that American company is a ball spit out a reasonably good open source model. I hope other open other AI you say it makes sense matter what do it and Google should do it to Every company should do it proper because basically you get to control the ecosystem if you put out a Performance Small AI model and I would I would I would even a large AI model the money is good Renting the model it's it's the yep now by Google has my Google has the TV on my wall and that TV's not going ever get removed So it's always gonna be there. Yeah, it's just relationships now what the heck is going on in Google? I mean we got It's a mess It's I people they're Bailing ship that been in 12 15 years and I'm like they should fire all the old people Fire the old people good when I go to San Francisco and hang out with AI researchers in some Residency they're all 20 years old. Yeah, put those people in charge. Well, yes, and no I Kind of believe the right way to do this is to know how to form a hybrid company and the hybrid could be requires you to have to functionally Different unit operating in units Done you cannot murder them they have to be separate and this freaks out the Harvard and yellow types because we all need the march the same base No, we don't in fact 18 team Bell Laboratories. I grew up around this Bell Laboratories proved this the only way you innovate as an existing company and not a startup because every company is going to be an Existing company by the way Met it as an existing company rules and existing company no longer startups You know and by the way Was open AI and so is anthropic their old companies now Especially open AI But by the way, we're recording Clayton Christians and right there for the last time. Yeah, this is all Clayton and and he was very well regarded and and early Early to mid Bell Laboratories And what they were doing is they would have team not working Working independently and they would they would call in a they thought their cadence the Three fatics was very slow is every six months and it would cross pollinate they would have a hangout and They didn't have to go to market they didn't have to you know Try to prove something to the boss they were doing pretty research. That's why we have the internet That's why we have the transistor that's why we have encryption that's why we have compression That's why we have basically almost everything that we're utilizing in this space is right now came out of Bell Laboratories That was the Inception and none of it was a go to market none of it was a VC You know, well, it was your product market fit they did independent research and if they were elected actually go to market Which they were illegal it was illegal for Bell to sell computer technology at one time They were they had their hands side We would have had the internet a day And half earlier we've had computers on on the desktop Addicted and half and it would have been open source a lot of people They all had been like the bell telephone being rented out. No that was required meant by By the Plicutility commission And all of that that regulated them they wanted the cell phones and they wanted you to have all sorts of waste User network and We would have Video 25 years earlier This is the province of what again gets involved with the pluses and minuses of Government involvement now here's where I get a little scenic the fact that 18t was granted a notly Actually allowed for Bell Labs to exist Because Otherwise it wouldn't have existed It actually be a bit of public benefits company There way to do that without a government intervention. I believe that every AI But he should be saving all all forms of me I believe come but upon them to form libraries so that this media is never destroyed I think that they can find it and they're right to do that The problem is they're robot cuts off this fine scans every page and then there's a way to book Yeah and and pulse it in burns that that's going to profit right there and that was their Misreading of a government of a of a judge's of finding a judge did not say you have to destroy that book He just said you have to demonstrate that you owned it and that it was never too in use at the same time for the general public It did not say they needed to destructively destroy it that was their own desire that was Darios Less wrong effective altruist way of destroying data and I'm sorry if it sounds harsh But there is no reason they needed to destroy those books and by the way some of those books were one of a kind It is gone and an actual digital copy doesn't even exist anymore So Eric and they put it into their models to figure the model would remember it the model doesn't remember it Remember as a fact similarly a flattened out one dimensional that's interesting In other words they assume once it's in the model it's It's okay to throw away the original yeah That's when you hire a 23 year old and you don't have somebody experience so I agree with your rock You need young people but you let them you know that's whole thing that they're smarter than the rest of the world No, they aren't I have a I have 20 year olds you know They're very smart but they don't have smarter at AI than you are Yeah, but you don't have wisdom and maybe not about the rest of it Yes, so The problem is when you get a guy like Dario or Sam They're all enamored by it because they're old men. These are old men now and they look I can't do what they do I'm just gonna I'm gonna put them on a pedestal let them rest you know and I'll do the business part and the reality is there's nobody Walking to henhouse that's why the shit's burning up. That's why they do Stupid things you know it's like again Either they were stupid or they co-collaborated all put their models at one company to be tested by a company It doesn't even have a proof that they even have I can test their models better Right you can test it better you can vibe code a better sandbox than in smithylese company did And would it support exactly? Would have done which by the way none of it would have broken out if you give it to tools I'll let exactly how they'd pick out which is not a it's at all Saying are you following my problem Claude Oh my god, oh my god, you broke out. Oh open it. I you got into hugging face Yeah, bad thing I've been faced got hacked You got the inside of How they models you know I know a guy in San Francisco running the security company He hacked all of the latest white commentator companies found 40% have Critical holes without the guy that came out He didn't need put out a video Is that that guy or maybe? Yeah, I put out a video and saying you know what we really got to take security more seriously No, we yeah, we we have a security problem Any code that was written at any time except for for Tran and Cobalt And maybe fourth is is suspect code Because there's going to be ways to backdoor and somebody will think this Bitcoin Cold wallet hack cold card hack is Province of AI that was a province of maybe insiders Putting a little vacation hack into there There is a whole lot of things about entropy and random number generators that weren't Being addressed correctly. It seems a way to foolish for it not to be planned to be that way Just say I I helped find it earlier and maybe exploited it But it was there anyway and That's the problem with When you put back doors and things is You now have AI that's powerful enough to see who and why certain companies had back doors And you you know again we're gonna be living through an era where there are no more secrets Because encryption is going to be cracked in our lifetime A lot of stuff is going to come out that's going to be embarrassing because a lot of people you know saved my embarrassing She was on the front page of the New York Times But you know some governments what have done some things that were a little on savory You know that's a whole reason for the UFO disclosure to they know that that stuff is going to be released and you know These aliens and will you come to combine visit off? There you're man I keep looking for him. I don't see him. I'm looking at sky right now You may be looking at him, but you may not notice that's all I can see right there. They're good at hiding. I'm sure, but you know, but but no it's not a hide The whole the whole the whole thing Robert when it gets released. It's going to be such him Everybody's mind's gonna go through an egg beater And I don't know why they're playing it out the way they are I assume that people have their agendas You know especially about the technology The technology angle I mean Eric Weinstein is getting himself in a little hot water Obviously a physicist he's kind of saying why has physics have not progressed over the last 45 years You know, once the emperor's naked, you pull the thing and you start saying, well, hold it. What happened? You know, back in 1959, there was front-cover news about less than gravity, less than one G system. What happened? It all went away. There were patents. They all disappeared. What happened? Well, a natural security. Well, for who's security? What nation? And who's using it? As far as I know, we're running out of bombs. We're running out of bombs. We're running out of bombs. We could use that, you know, national security stuff right now as people come with their, They're, you know, 20 dollar drones and, and a little power boats taking out Destroyers. You know, maybe it's time to start pressing those buttons. You know, maybe you're thinking too dark. We need to win at the fireworks show on the Fourth of July. We have to have a cooler drone show than the Chinese do. We're not going to be Chinese drone shows anytime soon, man. These are all. Have you seen some of the ones that you want to see? Let's run some. Let's run. On unbelievable. Imagine those coming out here down while you walk in down the city street. 10, 10, 10. That's a nightmare. That's a nightmare. That's called Zipline, man. Started in half, man. Yeah. You're going to have a lot of Ziplines going down your streets in delivering shit to your neighbors. We had to hands up. Who had their hands up and, uh, Rob. Yeah, jump in. Well, we'll take up all the space. Yeah. He like a bad part now, elevator year. Yeah. Come on. Speak up. We got speakers up here. Speak. All right. Yeah. Hey, what's up? Yeah. Go for it. Hey. So I have a question. But we'll make the room down by asking us. Um, but the, um, so you guys very enthusiastic about, uh, humanoid, uh, robots. So I'm trying to understand as a guy that does a lot of work in the physical world. Say I call a robotic plumber or robotic H fat guy. How does this robot get in? I said, and you're, or someone's at home, they don't know plumbing. Say I have a weak under my sick. The robot walks in the door. That was the robot diagnosed the leak. The, the same way maybe a human would. And no is the leak coming from the faucet that's coming from the same. To come from the garbage disposal. How does it interact with the world like that to know the, It won't have burst. Yeah. But what we want is a generalized humanoid robot that knows it's a sink. No, that's a pipe. No, that's a ranch, right? Can talk to you about what you're doing because you're going to have the optimists. If you're a plumber, you're going to have the optimists with you. And you're going to teach it. Uh, as you do your job, you're going to teach it. How long do you think it'll take before it knows plumbing really fucking well? And yeah, a human it takes about four years. Yeah, sir, I mean, you know, my brother and my runs a glass shop. It took him four years to get his license just to put less in buildings and stuff. I don't think it's going to happen as fast as, as necessarily even four years. I would just not be enough robots to do that. Yeah, yeah, so here's the way I would see it. If you're to come to me and say, what, what should I do with an optimist? I would say have optimist carry tools, have do it. Have optimists do all the stuff that you really don't want to do. Like, if you want to get out when I don't what they call you join out of the sink. Say, you know what, I don't want to touch all those hairballs under there. You go do it optimist. There's going to be those opportunities for it to do work, dirty work that you don't want to get done. 100% replacing you in your lifetime to the point of being as good as you probably not in your lifetime. Yeah, I don't go that far, bro. I don't have robot beat a human at the sport of car race. Yeah, I hear what you're saying. And it makes a certain amount of sense where the robot can do a lot of grunt work, maybe the same way you'd use an apprentice or a helper or something like that. Yeah, but yeah, I mean, just to do simple things like, hey, reach up in there and see where you feel the water or see where you feel the way you're leaking from. Now, pull your hand out and see if it's not an interesting one. It does optimists have sensors on its fingers that can sense water and oil and stuff like that. You know what, I think, I think it's going to be special built. In other words, like he puts his hand right in the sink right in the drain and just a snake just pops like he's has built in snake built in camera. It's just he's built with the. What the shit is going to be built in in the next decade I would tell you. Yeah, you know, the end of the decade from now you're going to have a robot that. Listen, the end of factors are going to be where the market is, right? I'm already working with a couple of sell startups that are going to be replacing hands and end effectors, you know, feeling things and all kinds of stuff, temperature sensors infrared reading on every fingertip things that you would you would definitely need for these types of situations. Now, this is why I was saying it might take a longer period of time is that the intelligence might be there, but the ability to interpret that intelligence in a way that a human can as quickly. Difficult now good enough. There's a lot of good enough people out there. There's a lot of good enough cars good enough restaurants. There's a lot of McDonald's everywhere. Some people that's good enough, whatever to me, it's not but that's the way the world works. So when I see a robot, I see it as a tool magnifying your capability way beyond the what you could actually possibly think of. Yes, it's going to be hauling tools around yes, it's going to be doing grant work, but it's going to be carrying different pieces of your work, making you much better at your job, much more capable at your job and actually doing much more work in the day, maybe making three times more money. You know, now. Would you send up your money? Or filling in for somebody who showed up sick today. I think that's more reasonable to take than rather going to replace all the human workers. That is the head line. That is Dario and Sam sitting across leg, got some conferencing, you know, job. That's headlines. That's to make them look like their foe intellectuals. It works. Then the media goes and rides it and they ride you and me on it and get drilled down. We're not going to have a life. And that's why I wrote to 5,000 days, you have 5,000 days to the end of work as you know, a lot of people, you know, it's controversial title. A lot of people read it the wrong way. It's actually a liberation. You are not going to do the work that you had to do before. You'll be doing something else and you'll be enjoying your life a lot more, making a lot more money and having a lot more time. The work that we're eliminating is the real work. Like when Ned Ludd got upset that his fingers, he lost three fingers Ned Ludd of the Ludd I fan. He lost three fingers on the shuttle on his own and his loom. And now some people say this is not real. I happen to think it is because I actually dug down deep to try to find more about the guy. Now he became very much a political item. And so then he came larger than life. He went back the day that they put automated looms in that next night. He destroyed them with a group of other guys didn't want to lose their job. What Ned didn't understand is he wasn't fired from his job because they did not they the people around the the weaving factory. They didn't know how to run these damn machines. Ned did instinctively knew exactly what I was doing. They needed them to run these machines. Ned was instead of running his own hands through the loom. He was running three looms at a time. He would have been in break him and making more money saving his fingers going home earlier and having more happier life. So the whole movement was started by the media and the the whole hype behind your job is going to get replaced. You're going to be useless. And that works on everybody's emotions. If you're going to tell somebody is a plumber for 30 years or 20 years or 10 years that their job is now over. Yeah, of course, what do you expect him to do rebel against it. What I'm trying to do is saying don't take that bait. Are you a fish and water? That's a worm. I've avoided it. That's the media playing you like a violent. That's Bernie Sanders playing you like a violent. So you get mad and you put some socialist in there saying, Oh, I'm going to require for every robot for two people to be hired. And I'm going to require a license of $50,000 for every robot out of factory. That's exactly the wrong thing to do. It does not solve a problem. That's not the problem. The problem is getting the right people with the right mindset. What is a robot going to do? It's going to make a human more powerful. What does all the technology do? It makes people more powerful. My first pitch to Coca Cola was to put two robots near the Coca Cola at the Costco. Yeah, in the back helping people pick that shit up and put it in their shopping cart and plus saying hi and high-fiving human stuff like that. Yeah, not taking you over or anything right. How many Wal-Mart's are there in America? 4500. If there's two, there gotta be three because they're gotta be a backup. So really an idea. That's probably your best. That's a nice first sale that I could take to Elon. Just have robots help everything upstairs and to car. Yeah, and it doesn't need to take anybody's job. No, at least that's what's right. Even the farming robots. Yeah, they admit they're sort of aimed at dealing with the labor problem in America. The farms can't get enough people because let's be honest. I want to pick strawberries out in a 90 degree heat and Gilroy. No, not me. I don't see you as a strawberry figure for a very long. I do pick strawberries in my backyard. Yeah, that's fruit fun. This would do it all day long for weeks and weeks and weeks and see what you see. Let's face it. Anyways, our kids don't want those kind of jobs. They're not going into certain of all the factory owners. I'm talking to you. I'm hiring all the time. I can't get enough people. That's why robots are going to be attractive to factory owners and warehouse people. Basically, you're just going to go after the jobs that really people shouldn't be doing anymore. And if they do those jobs, they're doing it because it's an artisan. It's a craft thing. I happen to believe that handmade items and craftsmanship is going to explode robots and AI start expanding because there's going to be more value in the heart and soul that people put in to their work. So that's a, that's one of the outcomes that we're going to wind up seeing. Picking strawberries is not something I wish on anybody's child. Right. And why are we paying people to do that job? Well, exactly. Because it's cheap labor coming over your border. That's who does the job. And it's going on in China, right? China is replacing a lot of their cheap labor with robotics and it's having a political impact over there. I mean, there's a lot. Yeah, it's a lot of people get out of work. They're having the same arguments we are, which is funny. Yeah, but a lot of good. A lot of those are wanted to talk about farming specifically. A lot of the soybean states have been really hit hard this year with the price of soybeans dropping. So any any political thing or tariff could affect the price of a food commodity drastically. I mean, these farmers are really in trouble in my state. So he means are the main crop. So, and also, if you remember, the farm bill is such a massive bill. It's like a trillion dollar spending bill. So how does the economics of, I mean, we have to live with the current economics of our situation. You know, how do you square that economics with the futurism? It's specific in specific are you saying the cost to produce soybeans in America or what is specific? Well, you have the competition, the cost of producer soybeans with the global competition. And also, China just said, we're not going to buy American soybeans. Yeah. Well, here's a problem with farming. All right. So farming is is a complex thing because the government got involved in rationing what farms should grow, what crop and how much crop they should grow. That central planning worked for a while. But anybody who really studied plant and set planning long enough knows that it always fails. Unfortunately, the way farms of the whole funding of farms underwriting farms is all written in an era that no longer makes any sense. And so, when you have the government saying produce more soybeans, right? That farmer can grow something else potentially. But if they grow more soybeans, we have a market for it. And China says, no, we're not going to buy them. You're going to have a price correction that's going to that we're seeing right now. And China knows that they can play this. They have not slowed down on their automation of farming. And they're going to become 100% self-sufficient, probably within seven years, maybe six years. They could be sooner, but we'll see. And if we had rational government, we'd have hundreds of billions of dollars going in farming. Yeah. Yeah. And allowing, allowing the use of AI to predict what farms should be growing what when instead of some guy. I mean, the way they come up. By the way, if you have a healthy robot farm farming robot industry, they're buying the same actuator. Well, they're buying bigger ones, but they're buying a lot of the same Nvidia cards or whatever Tesla's going to build. Yeah. They would fund the ecosystem of manufacturers, that many facture pieces of the robot, whether it's in the farm field or in my living room. And that would be good for America, but we don't have a lot of leadership in the innovation department right now. I mean, certain sectors are going to have to be centralized massively for this to really take off. It just seems like a free market, a free market mindset might not be helpful. Yeah, all the fucking billionaires into a room in Dallas and have a meeting. Alright, so basically, yes, maybe while you have the cast on the broken arm, right, but I think ultimately, if a farmer has their own AI model, I believe that the individual farmer will make a much better choice than some central planned group somewhere. I think if the farmer was allowed to grow what he wants to get some farms or not allowed to grow anything but what they're told to grow. I mean, that's the fact. I learned that an autonomous tractor is illegal technically in the state of California. So all these robot companies are crazy. And it's a law that was written in the early 1970s for OSHA for some student about what needs to be included in the cab, which means they would need to build a cab for you and that doesn't need to be in the fucking tractor. Why, why waste all that energy. That's absolutely crazy. I had no idea. Yeah, I mean, all of these things that we're talking about require a rational government. I don't see anybody coming up that is, and I'm including a JD van. He's a smart guy and he's got a lot of silicon valley people. Certainly, nobody on the democratic side, nobody from the Republican side that I can see, that understand technology to level that every single person on this space is due. Any one of us here could advise the government to do a better decision than people there. And the other problem is the vested interests. I think anybody gets involved in government. The first thing they find is that it's nearly impossible to unwind the hairball. The only way you can do it is, unfortunately, you have to detonate it. There's a vested interest to protect certain fifthums. You don't even know those levels of fifthums. I couldn't even, hey, I have an AI that tries to figure that out. It goes down 25, 35 levels deep of interested fifthums within government. And you can't find the quality of the deep state as if it was that easy. It's so many different grifting up their operations running inside. If you were to pull one out, it can undo the entire thing. That's what Elon found. I was actually thinking that to be the most useful task for AI is to unwind the grifting of all the states and just find a grifting level that you're comfortable on and just exploit it. I'd be looking at it. I have a dream. Have the government turn palentere on itself. Right. Government contracts, like Fogs, government contracts. Everything. That looks very crazy. Everything. What they're wearing. Are they wearing clothes or are they wearing clothes or are they wearing? I'll just place a go with them. I'm maybe. Man, they're. They've got to have them. They're in there to fuck things up and keep it interesting. Yeah. And the thing that's funny is that once you start realizing how bad off, you're going to be. I mean, from our look, this look at our education, our kids are our kids are getting ripped off. They're coming out of college with degrees that are absolutely not going to serve them. You know, this last graduating group this, it's always been that way, man. Anybody smart going to college had to teach themselves even while they're going to college. Yeah, but you know, there's something that we did to kids over this last 10, 15 years. That's really we can't. We can't get a result. Not every one of them, you know, but on a large scale, then you would normally find in a population group. And then they're, they're bought and sold into a system. You want to know why the education system is so bad, it's college education. Because bankrupt does not allow you to resolve that debt. It's the only reason. And that's why it became so corrupted. Well, the, the moment the government said, you got to student loan, you got to pay it until you die. And then maybe your kids might have to pay it, we'll look at let the exam and that possibility. Also, if you take your security, they'll take your social security also. Yeah, it's, it's the most, yeah. And I just, if you're for your fact, I'm going to get by charter. Is this the argument for free education and UBI? No, no, hold on a second. I've visited a company called Reflect in Germany. They made augmented reality. And they taught me how to fix an electric motor on top of the motor in real time. Without me reading anything, learning anything. I don't know how to fix a motor. You think I know much about it? Okay, I know a little bit about motor, but not enough to make one. And it can teach anything to anybody in real time. So I'm not too worried about the education system. We can, we can come out just with our new, new strategy. You struggle with kids. Robert, you struggle with 60, 100, 100, 100 dollars a day. And you, you sell them and exploding cigar, they grab your weight, senior year. And they're going to go out and, and use it a degree. And then there's absolutely no jobs. There's nothing. You go back to the educators. Well, sorry, you know, that's the way it is. And they can't ever get from under that debt. There's no wonder why you're 24 years old and you look at AOC, you go, yes, I want to be a socialist. Yes, come rat. Let's get them in. They have created a perfect storm for socialism, which always fails every time. It's communism. I can cause a lot of damage though. No, I will say it's about the education system. As when you, it's a direct threat to the education system is having a wholesale intelligence, right? And I'm finding, you know, with some of these projects that I work with an engineer, especially the young ones, I like to feel out like, are you using AI, right? And how are you using it? And I'm going to be honest with you, is that they don't like using it. They don't understand that the potential that has other than cleaning up emails. But there's still a lot of adoption to be had. And whether it's an indoctrination or what, but there's a threat. Yeah, I think you hit it right on the target. I think when I went to an education conference this summer, and ostensibly, they were talking about AI and education, and the amount of Amorous towards AI, it was like closing their nose. I wasn't presenting. I was originally supposed to present on gladi. I didn't. I mean, these were the top educated. They are thrown tomatoes at you. Oh yeah. They're not. You know, and here they basically, they basically hate everything about not having the control over the flow of information, not having the proper authorities regulate that information. Well, they, and so like I asked him, and I asked him, you know, like, hey, this is how I'm using it in a project, right? And like, look how good it's, you know, it helped helping manage projects, right? And they're like, well, I ran it through one project and it came up with like a few mistakes. I'm like, that's okay, right? Is that it's going to make mistakes? And so they look at it because it's not perfect if they're not going to use it at all. Yeah. Yeah. And, you know, when you finally look at the end game, the end game is we're going to bypass the end game. And we're going to bypass the education system at some point. It's a matter of time, right? What replaces it is going to be what our society winds up deciding it's going to become. But it's going to, I call it the interregnum. This intermediate area between now and then, and then is going to happen or we go back to fled Fred Flintstone here, right? We destroy everything and everything comes down with it. I don't think that's going to happen, but it's possibility. We're going to reach an endpoint where AI is radically adopted in people's lives. And not in a way of the singularity that some people are imagining that, you know, we just become wallets. Oh, I'm getting the neural link. I don't know if I am. They just won't be able to keep up. They just won't be able to keep. up is if you're not using it and you're producing at 10 times or 100 times the rate, as you used to, there's just no way around. Whether whatever you believe or not is, you can't keep up with the production. Exactly. And the thing is, if it doesn't happen here, it's going to happen there and wherever there is. Because, you know, if Bernie Sanders wants to stop AI in the US fine, that did not have any effect on the rest of the world. In fact, Europe might wake up and say thanks, Bernie. Okay, guys, let's try to beat the Americans. You know, that's the way it's going to play out. There is no other way. And he's effective all tourists and he's less wrong people who sit there and, you know, they're all academics and he all live in this fantasy world. Ooh, we'll need it like, you know, it's like a nuclear weapon and, and I'm one of the nuclear guys. I mean, they have these delusions of grandeur that what they're making is worse than a nuclear weapon. This is the, you know, if you want to look at what it looks like, it's Daria. Good wife is out. Hey, anything in white is fair. I mean, the bottom of the eyes, you know, listen, I'm not going to give it back. You saw that the same time. He's a terrorist, any ideas, but every single city in America can be used to destroy itself. And there's many ways to do it. I'll give you one, all right. There's radio act of materials at every single medical center in hospital. All you need to do is blow that up in a dirty bomb and you're going to take out a lot of that city, period and destroy. You don't need AI to tell you, I just did and guess what? That's in the chaos manual that anybody can get off a fortune. Don't do the home. Don't do the idea. I just think it's awesome. I don't know. I am slurney. I'm still a lot of beginner with this all thing. And I want to get better at it and learn how to use it accordingly. I just see a lot of potential. And I am enjoying very much the little that I know right now. And I want to build upon to make a lot of great stuff. So, you're forgetting all of the way over here. That's the mindset you need to have. You have to be willing to want to utilize these tools and fill with them as well as you succeed with them. Amen. Yeah, nobody's an expert at what this is going to look like. This is really the first time that everybody kind of has their odometer is reset. Everybody listening to this message here, you all can do things that you're quote unquote not allowed to do. And that is simply focusing on what you want to do. And if it turns out to be a dead end, find something else. You've never had this much power on your hand before ever. Amen. Friday night alone in San Francisco, there was 95 new startups born at Founders. Yeah. And there are all sorts of weird companies from somebody built in world models, somebody built in a farming robot. You know, it's a really cool time to be alive. Yeah, it is. And to say that only somebody who quit Stanford can do this, that's bullshit. And he, one of us can do it. You know, the reality is once the basic good enough AI model is out there, it's fair game for everybody. It's how you prompted. It's what you're pulling out of the model. There's more stuff that you can pull out of an old model that's a year old than most of the AI scientists think is possible. Because they're always moving on to the next thing. My job is to try to find out how much lemon juice I can squeeze from the lemon. They squeeze it once with their hand. I'm making tools that I squeezed every ounce of that juice out of a model. And they can be wasteful. I don't have the luxury being wasteful of my garage where I'm counting pennies, doing my research. So I have to do more with less. All of us can do that. We don't have to wait for the AGI model or ASI model. It's already here. I'm your prize. Get that. Do you guys recommend like going to certain classes or something or just wing it? No. Wing it. No. Learn to talk to the AI about what you're doing. Yeah. That's what it's. It'll teach you AI really good. You want to know how to build a vector database to support your agent. You ask your AI. It'll tell you how to do it. And then it'll do it. You already have the model. You already have the model. You're an ex-premium, I assume, or premium plus. You got access to grock. grock is a damn good, powerful model. Yes. But you can use it forever. And it's never going to stop because there's no limits to most of the time. What's you doing images and video on that? Right. I don't think we talked about the video ingestion system that grock has because as I'm uploading my like hour-long boring video up to X, as soon as it gets done, you can hit the grock button. And then say, please give me a lengthy report on what you learned by watching, watching and security word. This video. And it gives you a full-on report about its side of the video. Yeah, you know what? I am. You should do that from my binary ones. I did a video. I did a video with X with a geart from platform. I uploaded into grock. I need to do something here. And so I want to make a little series about this video that I made in a different platform. And I was like, you know what? I keep asking you all these things and this and that. And I have to keep asking you. Is there a way that you ever going to remember what I ask you because I should not be asking you ever single time? He's like, it's going to be between one to three years. They're going to be like, and remembering everything, what we're whatever, you know, that's what grock told me. And I'm like, well that's cool because I'm getting really tired of reminding you, or asking you what I just asked you two days ago. I would have been having dinner. I've been having dinner with the people who are collecting the data, the video, egosentric video data of how somebody is manipulating something with their hands. Okay, you got to explain that. And who are these guys? What are they doing? There's a group of many of them got funded because there's a lot of money in selling this kind of data right now. And video is paying a fuckload of money for egosentric data. I like if you're wearing an Apple Vision Pro and you're capturing video, that's egosentric data. It's data from your point of view so that they can use that video to train a world model. Really? Yeah. That's cool. And most of them say it's two years to seven years. It's hard to predict when this thing is going to really generalize and really impress everybody. But it'll come soon. And some people say it's 18 months. So I'm with 18 months faithful. I think it's coming faster than people realized. I agree. What product is that? These are companies that sell the data to the Tesla's, the figures, the one axis, the foundations, the unitries, right, that are building the humanoid robots. There's a big hunger right now for data. Unitries has thousands of people in homes wearing the headband that they have you seen that country thing. Yeah. Doing mundane things and they're just sucking in that data. I think to get it. We're willing to $10 a day. They're paying pretty high. That's I saw something like that on Facebook like I'm adding. Yeah. Yeah. There's a lot more companies putting you out there. I mean, those opportunities are going to be there for a while because effectively there's going to be longer and longer long-tailed niche markets of data. It's just like we were talking with the plumbing. He might have a specific way of doing something that might extremely important and training an AI model on that. $10 bucks. That's probably worth a couple of thousand dollars. Same as true with surgeons. I have a friend who's a surgeon that's wearing one of these during surgery. It's a much different. It tracks his hand. I mean, it's like a glove goes up his arm. Measures is hard to understand. The best company in surgery are building gloves and cameras on hats. They're sinking those two systems together so that it builds a package they can sell to a you know, Elon or Sam. I'll give you a secret. The secret is you need a lot more telemetry set into the model. You need a heart rate variability of the user, an iris of changes. You need galvanic skin response. You need. You want these workers to give all their data. Yes. And there's a good reason for it. Basically, especially with fine work and a surgeon. There are a number of neurological signals that go up when he's doing something right and he's doing something wrong. And that fine tuning of a model is exactly what you want when you're getting brain surgery done by a model that has seen every one of the top surgeons on the planet. Right? If you feed in all of the physiology telemetry from that surgeon in real time, you're going to have a model that really understands, okay, every time I've seen a surgeon move the knife this way, the heart goes off there, you know, the eye blink rate increases, the iris changes. That's telling you something. It actually can help that surgeon be a better surgeon. And I know this because I've actually worked with devices that help people become better at what they do. I check heart rate variability. Better view. Met of you makes when it turns surgery into video game. Oh God, yeah. Yeah. That's probably the best way to train a future surgeon. There's no doubt. And a future AI surgeon, you know, I mean, all of this stuff is is, is, is, is, fodder for, for better and better training. That's why things are going to get better. That's why they're going to get cheaper. I mean, the reality is that the, once you get fine motor control of an optimist, if you had to and a lot of saying, everybody's going to want this, if you needed an emergency surgery, you're not going to have to do it. You're going to leave your house. It'll happen right there in your kitchen floor, hopefully if you had to get it done right then and there. And am I safe your life? Because by the time you get to the hospital, you might be long, long, long gone. And so how does that optimist know how to do it? It's a domestic optimist that that's, you know, making your bed well, it's sterilizes up. It puts on the glove. It does his best job to sterilize you. And it goes for it because it's a download. Who pays for this? It's a three down left. Right. Oh, the billionaires pay for it. I was in one billionaires house in San Francisco. The dude had a complete bar in 360 degrees around the room. So he could have multiple bartenders serving his guests as we were walking in. Right. My brother runs two bars. So, and he wants robots in the worst way. He's like, if you guys, he tells me, if you bring me an optimist that actually works, they can walk in the bar and I can tell it what to do and it does it. I'm buying them and I'm buying them for 100,000 each. So just to be clear, Joe Sixpac, having the heart attack is catching floor. And the billionaires going to make sure he doesn't dock from the heart attack. Is that the premise of your. If you have a heart attack around an optimist, it should save your life and do CPR until 9.911. He's a, but here's the reality and I get where he's going. All right. So I'm never going to be able to afford a robot because it's $100,000. Well, here's the secret. Here's the secret key. Nobody has control over. Once a robot can build another robot, it's game over. That's why you're entering the age of abundance. That's why you hear Elon talking. What sounds to be retardant. I have an economics minor. What's the damn term? What? For marginal cost of making another. Yeah. Right. The first one costs off. Fuckload. The second one is a lot cheaper. The truth is in 1989's D. Wise Dank Show me is the color printer. It costs 45,000 dollars. Today is $70 printer. Does what is way better than that one? Way better. Not even. Yeah. We've, we've lived through so much of this. It's, it's second nature. But it's hard to say because each time we say it to a new generation about something else, it's like, no, this is why, this is why the Bernie Sanders win. Because it's hard for a lot of people and not saying the person I mentioned it here. It's hard for a lot of people to comprehend that they are absolutely going to own a robot, just like they own a super computer in their freaking pocket. Right? That, that self. And pretty soon in some homes. Yeah. It's going to be everywhere. And you know, I have a Neo on order. It's 20 grand. Wow. Wow. They could come. Gonna have to hang out. All right. We got some. Yeah. Bring yours over. We can have a robot party. Go up some hands. Maybe. And by the way, that's going to be a fun thing to do with your revised. Get them together and then teach them to dance or fight or whatever. It's like you want your robots to do. And in Robert's backyard, you have a easy legal. Put your money down on my. Yeah. Yeah. I would do that shit. I think you guys going to do a live one that we can see because that'd be fun. He's just, you know what, more data, more money, right? Yeah. But you know what I mean. Yeah. I wish I could live stream my Apple vision probe because it's the most remarkable camera I've ever used. And I'm a camera pioneer. I was the first to do a self-run interview. I was the first to do a 360 video at Coachella. Among other things. Get you downloaded and do a video like podcast with the video in the back. Failing. I just get a drone man. I'm going to drone or just fly around and video tape the whole thing and stream it. That's what I was going to say. Why aren't they drones flying around video tape and everything? You guys should stay here. Hold on. Hold on. You guys didn't see the NVIDIA simulator did you? The NVIDIA simulator. You give it a single photo of any street in the world. And it generates the whole fucking city. Oh yeah. Yeah. And then you drive and down the street. And then you can prompt to it and say, hey, give me fog or give me 100 pedestrians in front of me. To the fire truck across there. You know, whatever you want, right? It does it. And it does it in real time. It's like, oh, the fuck. Simulation is just big thing. I mean, Brian, it just takes a $70,000 computer right now to run that fucker. I don't have a member of the whole conversation. It wasn't a Pokemon goal. That's all there are some things system to somebody. Yeah, they sold their all location data. Or things killer selling. I think they're selling it pretty much. Any of these well in the pay for it. Niantek is the company that did Pokemon go. Yeah. Get the largest, sockman and reality company in the world. And they're located in the very building in San Francisco. And some stunning offices. I must add. Yeah. They have a digital twin of the entire world down to 1 millimeter per boxal. Mm-hmm. And then. And again, everybody to do. You walk around that in the armament. It is sick. It's like always. It's kind of crazy. Like people use that thing just because to have fun and everything. But what they built behind with all the pictures and the view and the view. And whatever the little ball, whatever the thing to creatures were, is like insane. Let's see your little bit rich. You got enough money to travel a lot. You're sitting around with your lover. I'm talking about going to vacation next year. Don't you want to go there virtually first? Yeah. Good idea. Right. And compare like, I want to go to the Taj Mahal. No, I want to go to the Switz-Off. Right. Have some argument. How do you get somebody to change their mind? Yeah. Thank God. Thank God. Sign me up. I want to do that. Yeah. Like intended. If you find just a walk around. Even. Well, in five years, are we going to do this audience shit? Why would we be wearing a pair of glasses and meeting in some virtual space and walking around and doing shit? Playing. I mean, the ultimate is going to is finally going to be able to make the virtual world where everybody wants to attend and to vital place. But I, at this point, met a failed at it. I, if Apple tries to do it. Maybe with the right talent. No, Apple doesn't need talent. Apple has an exclusive, unformulent one. If Apple can get the experience for the Apple vision pro that the formula to autonomous formula to recently built. Put that in the Apple vision pro glasses. Call it the formula blind glasses. You have any clue how many the how long the fucking lines that the Apple store would be for such a thing. Yeah. And yeah, everybody's like, I'm not going to wear glasses. Oh, yeah, I've set where the richest of the rich sit in the Adobe Washington formula one. And that dude can't see the whole fucking track. You can only see a piece of it. The rest of the time he's watching these small TV screens across the track. Crazy. When you see these glasses, the Google's coming out with meta and snap. They're legit. I mean, I had a haul in, Brian, do you have a haul in? No, no longer. I have a haul in sitting on the desk. It would blow holes in my walls virtually. Right? My real walls and aliens would come through the holes. And then you'd have to shoot on with your fingers. Oh, no. And my dad played it. And, you know, he's gone now. Right? And he's been gone for a while. So that tells you how long it's been in a while, yeah? In the oven, you know? If you like the vision pro so much, why does it come? I'll continue that. Oh, continue that program. It's continued. Okay, I'll just mark her and then didn't have correct information. It's continued. Okay, okay. Just needs to get a little smaller and you all going to use it. Yeah, you have some point. Alice is right. It's got to be miniatrized. If you're reading something from Mark German, it's coming from Steve Jobs office. So why would Steve Jobs office leak such a thing? To get our gaze off of it. It's all right. It's the most advanced PR company in the world. They've studied the shit at the dance level. And they spend a lot of money on it. And ultimately, I think it will replace like everything. Like computer peripherals, monitors like that. You don't need that anymore. Everybody will just have that and you just see things. If you don't have it, like if Brian has a 3D printer and I don't, and I want to print something that his printer has a rubber taxial, pop over his house and optimist will not kind of do. Hey, can you give me the part that just printed on the printer? Yeah, seriously. And here's your 10 bucks. I know just sent it. The nearest like got to play out on such a high level. Once. Once again, that 10 bucks. Is it going to be the optimist? Well, something in Pell Brian hand over the goods. He also got an expectation of getting paid at some point. Yeah, seriously. I mean, for example, like again, getting back to the Elon vision of abundance. Once you have your own robot growing your own food, right? The whole world, I mean, everything starts changing. As you have robots building robots doing various things around your house. It's not going to stop what you're going to have one robot in that sit. You're going to have a multitude. There's not going to be any government or any middle party that you're going to have to ask permission from if you control the means of the production. And that's really the situation. The bottom line is where does that start? Well, it's, I think we're going to see it by the end of this year. This is the beginning of optimist. The optimist is going to be the robot capable of making another robot. And now, are the robots going to make their own CPUs and GPUs? No. At some point, maybe in a far future, that's going to take place. Some basic parts you're going to get supplied to you, but they're going to be relatively available. And they're going to be relatively benign systems. You're not going to need some super phenomenal chip that only one company makes and you can't get in. The government's going to control it. Your robot's going to be able to do this at consumer grade hardware that's relatively available. Because the models are going to get that good. The fine control of the actuators are going to be that good. Yes. And the sharp hand size sign at CES could deal from a deck at cards to you. And do. or a gummy really finally and play ping pong with you and do a few other things. I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, I, so, all right, so let's just imagine a world where, you know, that the, the deck dealers are robot, right? And he replaces all the human dealers, who's going to gamble? And with with, who? Oh, hold on, hold on, hold on, hold on, hold on, hold on. If I had an optimist and my friends over, and we could gamble with robot. Good. I'm a certain, you can do what you want, right? And I think this is a, it's called on two to point a lot of people don't get, there, there will be a lot of problems, governments are not ready all of that, but as Brian said once robot can build a robot, right? They can be the factory, they can build more robots, and we see a hundred of thousands times productivity gains. Like, and this is, I think what people are not ready for. Yeah, and, and, and, and, and the, who is the we? In that conversation. The, the 99, I mean, I read this conversation and it's very, I love it, right? But I just wanted to be more, have more depth to it. It seems very, search this level, like, hey, it's going to be this level, but no one actually gets into the particulars, like who is going to benefit at first? Like, he lost. He lost the first, okay, so he wants going to employ the first robot. So save him a bunch of money, right? Who's he, who's, no, no, he's robots, it does not matter who is first, by the way. That's not, that's not, yeah, about that. We can discuss that, but let's just do that. I'll give you a few examples. I'll give you an example, is the common guy that's got to work, right? And do stuff, and it starts out, like, kind of very, very small tasks, and then it builds, right? And that's the guy that helps is like, I don't have to do that shit anymore. And that sells itself. Yeah, let's say your next or neighbor gets one of these things and comes over and says, Dude, I robot makes me thinner and, like, multiple different styles every night. Yeah, I mean, yeah, I don't know why I grow and enjoy life. So, all right, you don't want, well, that's okay. So I do. If your life contorts around a new system, right, this is what I've been trying to capture and the stuff that I've been writing, you have to imagine what would the meaning of your life be when the things that you had to do, you know, work on that you didn't really want to do is no longer a major part of your life. If you analyze most people's job about 80% of it is redundant stuff that they really shouldn't have to be doing, right? And the 20% is the stuff that they really love doing. And so, I always go after the 80% that people don't like doing. And so, when you try to extrapolate this into your sensory ports, with that, the expense reports. Yeah, like all that crap. I hate you doing that shit, they're core for it. Oh yeah, yeah. And we've already done it, right? It's kind of what we've been using computers for ever since they were invented. We've been slowly putting more and more of the stuff that we really don't want to do. And any of you guys coming along and say, okay, I don't want to write that, I'll let it write this. I don't want to think about, I don't want to go to Google for that. And all of a sudden it's feeling in the cracks. That's what's going to happen with robotics, right? And you know, you kind of wrote it. And other things, I'm wearing a lucky camera that's basically an eye of a robot, really think about it, right? You're the robot. You got one of your toilet robber, you can't remember what I wrote. It could be telling you, yeah, well, it could be telling you to move around, right? Hey, I need to get closer to the Picasso and the Museum, you know? Thank you, thank you, thank you, thank you, thank you, thank you. Hey, hello, and vertical. Do you guys have a question? Yeah, dog, yeah. I have to. I'm curious, of course, I've been following Brian's work and I'd love to hear an update on where he's at with curly and photography and whether he thinks there's a tie-in with biofotons, but I also had a refuge scenario for EGI. But if you want to hit that first, I got a second follow-up. Well, yeah, what he's talking about is I got gifted for A.I. Research, former Soviet Union scientific and technical research, and it's slowly but truly being digitalized. We're talking about a warehouse full of this stuff that a granddaughter and grandson, but most of the granddaughter, follows me on X, started realizing that I wanted old material and new her grandfather was involved in research, long story short, when the walk came down, a lot of scientific projects and a lot of scientists and technical people were thrown into disarray. They didn't know what was going to go on. This her particular grandfather was involved in keeping track of research data, essentially librarian. So this warehouse is existed since the wall went down and nobody, I don't know if anybody pays for it to be stored. I don't really know the background. It seems to be a government building still, but nobody really knows what's in there and most of the people that contributed to the research is long gone. I mean, he's one of the few people that have still alive. And so she approached him and he's always been talking about it. So as I ground potting me about this place and he took her there and she said, holy crap, this is like research that the Soviets have done for decades, 50 years and nobody cares about it and goes, yeah, he goes, that's what I've been trying to tell you. Now they're getting closer, long story short. He texted me, I have a conversation, she's graded in English, she's obviously in Russia and I said to start digitalizing it, you know, use your iPhone, send it to me, and we'll try figuring it out. We've gotten a lot more sophisticated since then, but some of the stuff we came across is a bit of a shotec and what's the question we're talking about. How is Kerlium photography is based on the concept that Soviets were trying to understand is something called biofotans. That's the body or any living creature, including plants, give off photonic emissions. And modern science today will say well it's a function of the activity of life. Yes, of course, but what the Soviets found is there's also a vitality in it and I'll give you the classic example. They took a leaf and this is the one that was linked to the west and they cut the leaf in half and they took a Kerlium photograph of that leaf with the leaf cut in half, one half of it off and the outline of where the other piece was removed still existed. So there is a emission or some energy that existed where that leaf formally was. Now I know we're getting a loft tracker about what are main subject is. I'll wrap it up quickly. That research was not one guy named Kerlium. It was actually a team of over 200 scientists that were doing research on this. Most of it never made it out of the Soviet Union and after the wall came down a lot of this research never matriculated up to the classic medical research. This was embodying an area of research that was outside of a lot of different standards. It was outside of physics. It was outside of physics. Theology was outside of plant research. It was in its own little place and as soon as things needed to go to market and to be functional and that's what the Soviets tried to do very quickly. This fell by the wayside. The research was like, yeah, it's interesting. Yeah, there's something there but we don't have time. Let's go into what the West is doing in medicine and let's go down that road and it kind of got to the dust bin. But the research when you start looking at it and the stuff that I've seen, I've duplicated a Kerlium system and based on their original designs. It's relatively easy to do. I'm going to open source a variety in an article on it right now. It's a little bit dangerous if you don't understand electricity. But essentially it's a charge plate that allows you to see this aura that comes out of a living substance creature a person. It's a type of plasma almost, right? But you know what I was getting at it? I've talked with Dr. Nurosh about using machine learning for diagnostic possibilities. So it does time back to not only health but agricultural possibilities where you could detect, you know, weakness is or fatigue, disease and plants. There's a lot of possibilities once we unfold. No, this is brilliant. No, it's brilliant. You could put seeds on it. I just got a bit of energy here because I came in here how close Marks say I generated content. I've just finished reading a million words. I wrote a cannon with 14 books. I posted three articles up and up in the jambos run there. You want to take a look. I'm building a whole AI based on a based on metaphysical. I'm a man. Please don't. I'm a person. I'm a person to do that just things up and we just had a conversation here, right? Just try to chime in normally, right? Yeah. So these in Dr. Nurosh. Oh, it's going. Yeah, it's fine. Please continue. Anyway, I don't know what's going on. I'll just put an end on this. The bottom line is there is a lot of data that is not currently available to AI models. And that's what I'm trying to get my hands on as much as I possibly can. I've been fortunate enough to get a lot of this data already. Now, I've got data weighed outside my realm from the Soviets that I think is going to be quite powerful. And, you know, it's going to lead down many esoteric channels because Soviets were not bound to the same standards that the U.S. and the West was. And a lot of the research was very, very interesting. But it led to led to some conclusions. And, you know, Havana, Syndrome, some of this stuff that's going on. Now we're starting to see the results of it, whether modification, the ability to implant voices in somebody's head hundreds of miles away using microwave communications, V2K stuff. So, it's really advanced on it. So, I'm going to put a cell phone on my brain, dude. Seriously, that's what I'm going to sign up for anything neural link makes for consumer. So, what do you think is going to be the first thing? Do you think is this going to be like, I talked about a brain? I mean, what do you think of, I've had some headphones on that let you train it on 10 words? So, you think like record and maybe a camera starts recording, something like that, right? I can do that without it. There's a hat coming that does more like a thousand, which starts getting into a language. You can talk to it with just by thinking, right? Yeah, and then it gets better and better over time. So, it's all about when, when do you jack in, you know, how many wires you made? Well, the thing is, you can do so much. Like, I use a lot of, I break down toys. They're a neurosci made a chipset that allows you to read essentially brain waves and the original neurosci chipset was one chip on the, just about center of your temple and allowed you to move balls up and down and left and right. And really great technology. So, I found those chips I got now about 400 of them. I broke them out of a huge awesome. And so now I've made an entire skull cap with them. I think it's your fucking warehouse, right? You must have a size of mine. I'm running out of space and getting in trouble. I'm a lot of things. Are you like one of those guys that I'm passing on the interstate? You have a big fucking warehouse, but then you have a lot of shit outside. I got, I got a refrigerator on my front porch. Now, I made a skull cap with a, I've got a standard cap with a newer sky to the chips. And you can, using it, it's a couple of new AI models where you can decode the brain waves because they've taken the brain waves of about 50,000 students at this university. I don't remember the name right at this moment. Should, I use it often enough, but I don't address the model. The bottom line is you can get a statistical standard of what a pattern for a certain word is. And you could think a word and get damn near close to that word. It's almost like thought to speech. That was thought to text. And they're going to have headphones that do 10 words like that. Yeah, yeah, and so this makes that a trick. I'm doing it in my garage. That was coming next year. Yeah. And the AI models that are going to do the decoding. It's like, when we first had text recognition and handwriting recognition and AI, it was horrible for decades. And all of a sudden, you know, the ability to build neuronal systems on a multi-level, which is large language models are essentially doing. It fundamentally changed it to the point where it's like easy. Same thing with voice cloning. I mean, I got a free local model that clones my voice. It does an incredibly good job. Some of my unfortunate, some of my odd guess. Dude, if you got enough formula to car and tried to be confident in that, you wouldn't be able to. Yeah. The robot beat the human at it. Yeah. I have a certain, you got to be not really into G-bores. Is that the takes training? Oh, yeah. It's G-bores, but it's more fear, man. I was watching the human go around the turns and he just wasn't as smooth as the robot was. The robot can push those tires right to the edge of traction perfectly over time. Yeah. Human can't do that. Human obviously. The water we watch in the middle, in a bowl. I think it's not because of the it's the danger involved that the fear that someone who sent it could die. Yeah. And that's why we watch it. So I don't really get it. There was a rack in it. Yeah, I think. I'm not going to bump around one of these cars, cost $150,000. You rack one. It's not good for the bonuses. So let's see. You know, and I say no, but I get it. The thing is, the thing is, it is the period not getting Christmas gifts. It's going to buy for Kate, right? The viewer, the viewer, that's that's positive that Christmas gift. Yeah, but you know, you make a good point. And I think it's going to buy for Kate into two things. I think there's going to be a definite desire for people watching robots beating up robots. But I also think there's going to be a desire to see a couple of guys going box it up. So I think it's just going to open a wider, not necessarily eliminate it. And I do this America. Let's have man on man on. You on robot on robot. It was a quiet, it was a quiet lifestyle. In fact, let's get 10 of them in the ring at one time. This is cut and see what goes on. Yeah. No, I mean, this guy is going to be the limit. There's a Twilight Zone that actually I can't do that in America. To me, the lawyer. Too many lawyers. Yeah. You're going to, we're going to live to see the day, row by the against human, human against robots. It's going to happen. The thing is, we just have to open our mind up to the possibility that these things are going to happen. And the people listening here should be taking advantage of the opportunity. Don't be the person with the arm folded and saying, no, I don't want to do it. I'm kicking and screaming. Don't be that person because it's going to pass you by. This is an opportunity. I was at the launch of the meal and I wasn't the first to buy. And the San Francisco, so I'll come to LA robot industry was there. And people were like, here's 20 grand. Yeah. Yeah. Can I ask you guys a question about AGI? Sure. If you guys were at the craft table and you were betting on where AGI would first emerge from, would you be betting on, like, either say top ological computers like the Marjorana and the Willow or, and obviously, the, you know, keep it corrections and probably some quantum insights as well or would you be betting on further refining the current LLM states with, you know, so on and off induction, universal algorithm, make production and other types of explicit graphical model inferences. Well, these are great questions. The problem is they keep moving the goal post on what AGI and ASI is, right? So here's, here's the big thing. The big thing is all bets are off when you start getting AI coming out of a quantum system because there is no going back from that power and that is going to come in our lifetime again. Well, if you like 10,000 times, it gives some idea of how much the words that'd be. So, so the thing is we can't comprehend the reality that it will bring upon us because we don't have the ability to see it. And I'm not talking dimensionally, I'm talking thought process-wise, right? Now, this should, this should make you aware, but it shouldn't scare you because we automatically assume that if something's more powerful than us, it's going to kill us because it doesn't think we should exist. Now it's going to help you with your luggage. It's going to get well. It'd be doing more than that. Well, that's what you're going to get it for with the robot. Here's what I learned about programming. I need some problems. I need 55 pound luggage. Yeah, there you go. It's what I learned about program quantum computing. I had access to still do in one in San Diego at a university. Is that you can't predict the things that are going to be coming from this because you don't even have the questions yet, right? This is the same thing that's kind of going on when AI first arrived. Will you give up your free will? Let's start with that one for a brain computer interface. Well, the question is, what is free will? We're going to start going down these blocks to see what is the level of what is the level of the power? You got five hours? Yeah, right? No, you can't. It's just a start. Are you guys so certain that that we have achieved AGI or ESI with some of the advanced, or do you think that? No, it's more about this Smithsonian than you do. I know that. The classic AGI has already been achieved and we went through that we blew through that and there was no horns, there was no strings being played and Angel saying ah, it just happened. But what is playing upon the background of this is the fear theater that's coming from it. Right? It's like, okay, you know, chat, GPT 2.5, that's going to be dangerous. Don't release it. That's why Dario left open AI. He left open AI because 2.5 was too dangerous for anybody to have. People could learn how to make a bomb, man. Yeah, it could do bad things. He left and he started anthropic on that entire thread of 2.5 is too dangerous. You want to know what everybody has access to 2.5 level AI for free and is anybody getting trouble with it? No, yes, it knew how to make bombs, it knew how to do all kinds of chaotic things, still in the model, it's still in the open source, and Lama has it. Old Lama, and I'm not even talking about met as new release that came out today, which is a great model. Wait, so I just want to get the logic there. So in order to prevent a super intelligence, you make that even more intelligent vehicle. So no, 2.5 is too dangerous. I need to make stable 5. Is that right? Right, so I can't drop it. It just doesn't make any sense. That's it doesn't make any sense. So the whole thing is a comedy routine of accuracy. So say you did it for the money. And then drop it to find a G.I. as taking more than 50% of all human jobs. Yes. And their average staff answer is that's going to happen in about two years. Yes. And just that it's going to be what was the turning point behind for where you thought at least for the initial goal post for a G.I. was it like a adaptive recursive tokenization or where would you put it? Yeah, no, that's exactly where it is. And and and mixture of experts. I think what we got to mixture of experts, we were pretty much there for most of human, as long as humans were considering A.G.I. we achieved it at that point. And again, it's, you know, we create these tests. Right? We and look what happened. We had a test. It was hidden over at hugging face. And, you know, this Middle East company decided they'll hack hugging face to try to get how to answer that test better. So basically what's going on now is a bit of a song and dance of how can I make my A.I. do better on that test than somebody else's A.I. And so we're all test-based. Well, what other metric are you going to use? Well, there are a lot of other metrics. They're not going to tell you what most of them are. But most of them are real world metrics. It's like, can I ask you a question about anything and get an answer on it? That's relatively close to an expert. And the answer to that is, most of the top models can do that. And that is the classic definition of A.G.I. Not whether or not it can finish the human final exam. There's all bullshit. Again, it's a bunch of academic sitting around. I'm sorry. I'm going to use this metaphor masturbating over the ideas of tests. And, you know, this is the final of G.A.M. And then, no, well, no, by the new final exam. And it's like, okay, guys, you're already blew it when you the first version was beaten. And now you got to create another one. And it's just, it's like the SATs. It's like IQ test. None of these things are really good. I read the book. It turns out if you don't know an answer, answer is C. I love it. Yeah, I mean, you know, I knew one of the experts have built the SATs in Princeton. You worked T.Bullet. See, he's always good. Or you just make patterns, you know, a re-crush or pattern. But, you know, so the problem with the terminology is it's a mechanism to get funding. And you got a lot of frustrated middle-aged men who have money to invest and they want to be able to have something to brag upon when they go to parties. And they're otherwise known as VCs and not putting anybody down. But it's typically what's going on. And you say, well, I'm investing in a company that's got a GI. So, okay, Bob, that's great. You got your act together. You're not a middle-life crisis. You're still, you've still got it. You know, and that's the game plan. The entire game plan is a C.F. experience for venture capitalists. And that's what's going on right now with AI. Because they're running out of money. It's true here in Silicon Valley, all the robot companies get together. Look at how each of the other robot companies is positioning themselves and building their robot. Right. The one X is soft and fuzzy. The foundation is all about military or anything that's really used. The figure is still trying to figure it out a little bit. They're trying to be on the cutting top. I see I think. They're trying to be a little. You know, he'll sell up to you. They're all going to sell it. Because I'm assuming he's going to get his generalized and capable. And keep up with optimists. Maybe he'll be like. He has aspirations. He's hurts and Elon is a business. I think he's the biggest. I'm saying now, Elon starts your A. Yeah, yeah. I'm going right. I think here's a problem. I think he launched so far ahead only because he's got access to compute the he said these other companies don't have. He's got access to manufacturing that these companies can't. They would take him to the year scale up. So yeah. The parking lot was full on Sunday when I drove by at the free month factory. All employee lots for full on a Sunday and they closed down two of their lines. So that means there's a lot of people over the weekend. You know, and that's not to dissuade anybody. I'm all about garage operators arising up. There's going to be more than enough space for all of these companies to not to survive they're going to thrive. And it's going to even include the Chinese at some point. You know, the ones that own it out of their scale are going to need to be able to manufacture millions a week, not a month a week. And Elon's factory never stops. They have three shifts of workers, 12,000 total or 14,000 total. And what happens when he starts putting optimists on those factory lines, right? I'm telling you, you know, when you when you look at the factory, he just, you know, really says this week, you look at the scale of that thing, you know, miles long, right? That's all going to be full with robots. If you're a smart Tesla worker and you're working on the robot, you already have a pretty good idea of the impact this thing is going to have on a human. I can tell you this. I'm, yeah, I don't want to, I don't want to give any inside information. I'm just telling you that the capabilities of optimists is going to blow people away. So it's not going to be dancing robots like China, not going to be kung fu fighting or have that there's going to be no it's going to be a general lives to humanoid robot walking around a factory floor doing any fucking job anything anything. All right, let's, let's, let's see into the future. It's saturated. You right, these robots are so epic. Everyone has one. No one has to work. What, what do we, what do we, why do we need money? Okay, you're going to torment yourself and give yourself a job. Yeah, I've talked to many of billionaires. Why would I be working? Because his wife kicked him out of the house because he was going to get a divorce and cost half a billion dollars. So he decided to go get a job. He was brought to many of billionaires. I don't know why. Not because they needed the money, but because they needed the job to save their mental health. Yeah, yeah, you know, listen, the bottom line is Elon's right money is going to become less and less of a thing. He made a very controversial statement that in about 15 years, your retirement savings isn't going to matter. You're not going to care whether you have it or whether that money is even going to be of any use. You know, and again, this sounds socialistic. It's actually quite the opposite. And it's hard to explain it in a light conversation like this. I wrote a lot about it. It's really about going backwards into a more localized environment. You're going to be able to live anywhere you want to live because the means of productions are going to be in your life. And everybody's going to say, well, I want to be Trump property. Yeah, you're going to be a lift. You're robots going to go to antique and pick up some strawberries for you and come back. I mean, so when you start playing this out, and it's, I'm not a utopian. This is not utopia, by the way, but loves for Alex. We're not heading towards utopia. We're heading towards reality. That's going to look very utopian. It's going to, but it's going to have a walley world type of problem. And that is, if you don't get your mind right about who you are as a person and what you want to achieve in your life, you can get very, very sedentary. And you're going to not live very long. Your, your meaning of life can go away. You wrap up your identity into your job and your job only. And I'm not saying it's easy. I'm still having problems with it. And I continue to think about this for the last 30 years. But the bottom line is, and this, I don't want to get sexist about it, but a lot of guys are born with the idea that you have to make value in yourself. That you are only valuable by who you are by what you make and what you produce. And that makes you a valuable partner for a mate. Unfortunately, that is not always played out the right way. Young girls are told they're beautiful and that their value comes from their beauty. And that's saying that trying to be sexist about it, but there is that innate reality. A lot of that's going to be reset. Meaning we're going to go back to some of this. And so what value does a guy have when everything is being produced relatively inexpensively to the point that energy is going to be relatively inexpensive. Because energy is going to become relatively abundant and freight. That's going to be the big, that's the ultimate transformation within all of this because the, the guiding principle of the mechanism of control in our current political system is those who control the energy control the society. That's all going to break down whether they like. If I can just, if I can just add there is that, you know, once industry utility industry adapts it, right, which is coming, right? Once we see that, then we're talking about efficiency scales are real and measurable, right? We're paying ridiculous amounts of money for energy because we're stuck in an old system that requires energy to be costly. And the political systems on both sides, but mostly today on the democratic side, really are in the mindset that they have to restrict energy and make it more expensive so that the industrial, industrial, polliteria, I mean, they're going back to fricking marks. They don't even know everything goes back to the same guy, the same beard, the same stupid, you know, the industrial means of production. You know, once you realize that the means of production and the hand of the individual, all of their programs fall apart, they don't mean anything anymore. That's why they're scared. They're scared because if you and I have the ability to make anything we want, we don't need Amazon anymore. We don't need any trend. What's that? Preach it. Yeah. We are entering into that phase. It already happened. We're not waiting for it. You have AI. Everybody on this call has AI. You basically have the access to all the knowledge in the world. Now, we've had that with the internet. It didn't really help us submit us worse. It's not perfect. It's not perfect. When you're you're learning to integrate it because it's you know, it was new technology in some sense, right? Yeah. And it's it's imperfect and people need to understand that, but what it can do is incredible. You know, if there's so many different use cases, you know, in your narcotics, right, and you're one of the greatest at it. Now, you know, I took an old asville. It's a little bit of an illegitimate and asville's are extremely valuable, right? Oh, yeah. And then just said, okay, I can't read this thing. Do your work and it was incredible. The garage was it. Think about what time and efforts would have been required under a different world. See, this is the thing. We take it because it is so inexpensive. We take all of this stuff so for granted. It is a super super power. Anybody out here has a medical problem and you're not using AI and there's some, you know, privacy issues, fine. I get it. If you're not using AI to track what's going on with your health and a particular diagnosis, I don't know what to tell you. And and all the pictures of everything. Yeah, I mean, I give it to the, yeah. And check it and check the work right, you got you got absolutely check it, make sure, and then you can correct it. But everybody wants. Hey, it made a mistake. It's bad. It's like, well, that's normal. Like, a mistakes or normal. My, yeah, my best friend was, was killed by Colin cancer at the age of 40 because his doctors missed him. They made a mistake with him. Humans make mistakes, too. Yeah, yeah, we cannot work together and make them the stake rate go down listen radiologists and I'm not putting a radiologist down radiologists have to evaluate sometimes hundreds of things in a day and You know there's always going to be something marginal so it's be something that's a bit of a shadow and you know you have AI now built into a lot of radiology systems but even then it's not perfect I have people that take any scan that they get any blood tests and they push it hardcore on the on the AI no this find anything that possibly go wrong and then you work your way backwards and to me that is the biggest blessing that AI can possibly give anybody free right now is the ability to scan this stuff you'll on it talked about it with grock you said put your x-rays into into grock yeah it's going to train and you're going to get your information fine but at this point if it could save your life I absolutely urge you the other thing is your prescriptions I have people I am still there but yeah can you hear me okay now sorry sorry I was like oh no I doubt you lost me yeah prescriptions if you're taking prescriptions damn it put it into AI and talk about the contradictions talk about the the dosages most people are over prescribed not just on too many pharmaceuticals but the dosages are out of control high right if you're on blood centers I guarantee you I hope nobody's on blood centers less than we only please but if you're on blood centers I guarantee you you get you're getting way too much AI will go analyze it and say hey Teter off of it if you're on any drugs for depression or mood enhancers or whatever put it in AI talk about exit plans how do I get off of this I mean my he's a friend of mine now Jordan Peterson got screwed up very famous guy because he was put on all sorts of anti-depressants that he shouldn't have been put on and he should have known this stuff and when he took it off it's really hard when you're the patient to sense like you're going sideways it's always a family not getting the right stuff you're going down a bad path you know they'd have a big workload you know they'd have a big workload as it is and so not that you're not you get the information but like how many times you just walked out the door with just what it is the prescription or you know the diagnosis and then you put it in there at least you have something and generally it's a tissue at least not medical bias right but it is a good direction well another Stanford surgeon once who told me the mobile phone was democratizing knowledge he said before this you had to go to some school like a Stanford that has a huge medical library to have the knowledge of humanity and a building now that's on a mobile phone it's crazy right guys one of the and now the questions I have for you guys just show there's another revolution going which is AI is democratizing it again and bringing this kind of healthcare tarfons. What do you think? Yeah no I love the love the convoy truly think about what that does for communities he told me going into communities and Vietnam and the doctor had one medical book on the shelf and very good of mentoring. And you imagine knowledge or knowledge of the state of the art of medicine right. I agree with everything you guys are saying for the most part I just don't understand why you don't see the obvious argument that you're making for universal basic income because I mean I'm not gonna argue maybe not it's gonna have like a wokism maybe it's got like a diamond so let's call it like universal basic standard of life. Why not universal heart? Whatever whatever whatever I prefer the future. Give a human a shot at a new American dream but that's me it got my family at a poverty and I think that's more sellable than you know communism. I think my question with that just as a follow up is how do we get to the place Robert up with your statistic on anthropic describing a GIS like 50% of the workforce gone and if that's truly two years out like I mean I love I'm an optimist I my pdoom is pretty low but like that's a very short amount of time for that a nasty of change like what do we human being person about six months to grieve and then four years to learn something really new yeah and and and and and and and and and and and and and and and and and and and and and and and and what are they doing oh they'll figure it out there's there's plenty of jobs it's just they're very technical very nerdy kind of jobs that the Chinese are taking right now. Yeah but that's what I'm saying like what can we do socially or self to prepare for that. That's right. That's right. That's right. Learn. Yeah assume your job is not going to be present at some vector ahead of you and and and make it sooner than later and then say who am I what do I want to do if I can do anything that's that's what I would tell the to tell my kids so I tell anybody listening instead stop stop going the way we were going and say okay I now I'm going to free myself and I'm this can have a thought process here if I could do anything what would it be and I'm not trying to be the Tony Robbins going you know I'm really talking very factually what would you do and if it's the same thing then use the technology to make you 10 times better at what you're doing don't don't fight it don't fold your arms don't say it can't fold my laundry yet I don't want the robot understand that every time you see technology it's a worse it's ever going to be. So if you if you're going to be a critic about it you're going to sound stupid stopping the critic let somebody else do that to say this is a you might be right today actually tomorrow you're going to be wrong yeah and so what what that mindset of being the critic does is it steps you backwards and it stops you from participating you have the right to participate you don't need a degree you don't need somebody to say that you're smart enough you have a i what you don't know ask a i and just live those things out in your mind what is it and I was just at a hackathon in San Francisco and they said their teams are bifurcating into advanced users of a i that have their own harnesses and their own loops and their own memory stores all set up ready for the hackathon versus somebody who hasn't had time to really get the tool set built yet. I'm not standing on the side with my hands but I host the i i so in hackathon daily on Twitter spaces so it's not that I'm a power user at minimum right and I teach people how to use this you know democratize that skill set all the time but i i'm just i'm not criticizing i'm critical right i'm just thinking yeah further i'm just making me the solve those problems magic and it's an excellent stuff and we can't just ignore them so i'm really just thinking and being talking about it so that we can actually migrate to this evolution of humanity versus just saying you know they'll figure it out that that's it i think you're absolutely right i think we need to talk about this mall more because our entire society is set up for people other figure it out you know somehow it'll work no you you absolutely have to work the grieving process you know i wrote about this in my 5,000 days series it's a second part too i went through grieving you know a live with kubler Ross i i literally go through what Elizabeth kubler Ross developed in the five stages of grieving you literally have to grieve the old you you're gonna this generation the people living right now are gonna have to grieve their old self they're gonna be brisk the gap of what it was like to work your ass off for a boss and a dumb job to having so much time that you don't know what to do with it that is going to be a whiplash for a lot of people and and it's not fantasy it's already starting to happen and you know where we keep cling to the old political systems the old ideologies of well this is how it's going to be and you know some kind of everything's going to be why and we're in a dress in uniforms and it's going to be socialist it's like none of that's going to play out that's all game fantasy it's going to be ugly human kind of things all that's are often augmented reality though yeah well that's a different thing right but but the bottom line is you have to start taking more control of what your life is going to be and the only boss you're going to answer to is yourself it's not going to be the boss that you're used to and that in and of itself is a problem because most people have been trained by the school system and the work system to be a certain type of human and they don't know how to be somebody else so there's going to be a middle ground for some of those people and unfortunately there were the majority of them the minority of the world the people listening here and eriting out at 10 30 California time we're a different breed of people you know we're willing to go through this process of hearing us all nerd out and talk about stuff that's kind of the most people don't even understand what we're talking about and almost none of us would think of us as extremely technical we're just people that happens to use technology but in reality comparing to the average person we are the nerds that they're talking about everybody on this call the average person was like oh they're all nerd it out and you like well I'm just an average person it kind of got no got good no you are a nerd compared to the average person so what is that I was just a strategist at Microsoft I guess that counts yeah well I'm kind of honored I guess I'd ever got called that before your nerd you're one of us the nerd you want to have so many friends so freaking against this whole thing and I'm like I'm going to play with it just because I want to see and we need more people to be able to feel free to do that and I think that's why keep saying don't fold your arms don't reject this and that's saying become part of the matrix I'm saying use these tools to your benefit once you get the means of production in your hands it's no longer in their control it's in your control that's going to be maybe the province of most of us maybe not all of us because it's going to take a bit of time to you know resolve itself but in this 15 year period because 20 30 20 33 that that period that's when it really is going to hit the fan that's when whoever is in control of that period they're going to start saying oh shit I don't know how to stop this because they can't you know if they want the old order to still be you know petroleum in the ground and we get to control this and we have our little game here in our mafia here you know if Elon is right and I happen to think he's right if my theory is right that whatever money is is going to disappear rather rapidly and they into his pocket can you see it he even said on his own and one interview said that you know yeah I'm a trillion error it's not going to mean anything in 20 years and yeah he's going to be a thousand million error he puts what but what does that money mean when everybody has their own means of production the whole thing means he can buy like a piece of the solar system yeah it's like I always want to it's like you know the whole thing in the stand for that of monopoly mean fought Stanford University which started Silicon Valley right Elon started that kind of level player yes yes I assume that most people who have a little more right but I think with that production gain this grabs off that table when like all of us yeah yeah absolutely you know and and what happens is you get to the point where most people will be like okay I don't need anymore you know in one of them in one of my series are wrote about this movie called the humanoids and the entire thing flipped it's a great movie you know by the way and it was British 1970s and the whole thing flipped the other way the very wealthy has the least the very poor had the most and the the factories were running so efficiently producing so much goods that the poor had to absorb everything so you had to go through clothes all whole lot faster you had to eat more food you had to live in bigger homes it sounds quite bizarre but when you actually watch the movie you can kind of see it array playing out the the abundance that we have is array of fencing our culture and I'm hoping not too many people here but No, my kids are standing in front of me right now are spoiled rotten. Yeah, you know, and but just look at the food. The food has gotten so abundant and so junky at the same time. It's causing a crisis in obesity. And the only answer is to take a pill but still eat that same food. Right? It's not how much food you're eating. It's the quality of food that you're eating. When you visit the farmers down in Salinas, they tell me about spraying our food with pesticides or circumstances stuff, right? If you can find a way to farm without having to do that, we can make better food for ourselves for all of us, right? Absolutely. I have a change in that towel changes. Even if they eat it, McDonald's, right? It gets a little bit better. But here's the bottom line. If you ever hear the word pesticide, you actually can convert it to one word, estrogen. Every single pesticide is an estrogen. That's how they work. And when you finally distill it down to that reality and you look around what estrogen does to the human body, you can see the changes that it's done. Yeah, I'm not talking about the frog's gay. That's estrogen. That's something different. But they're all estrogens and they're absorbed by the body and they function the way they're supposed to function in any biological system, whether it's an amoeba or a plant or a human. It will put on weight in the certain pattern. It will make you bloated. This is what pesticides do. And you don't know about it very easily. I hope more people hear it do. That is a real estrogen. Relatively available. Go and check with AI. Check with it. You can, yeah, might not to push it because it's going to speak the agenda. But the bottom line is it disrupts the reproduction system by creating a birth control pill, which is an high estrogen. And it causes the insects not to be able to reproduce. This is what's going on. It's fun to size also. You know, so we're we are floating in it. And by the way, it's in your water. You can't take it out once it's in your body in your water. It's there. And so for that means for an average guy, average guy has three times more estrogen than the average female did in 1915. Right? And it shouldn't be that way. And it's coming from our plastic in our food. And it's like, well, why can't I reproduce that 20? Why is my sperm count down to three? Well, where am I swimmers going? It's because of the food you're eating. And you know, we can go into the conspiracies of who did it and why, and you know, the agenda. Not going there. But the bottom line is the chemistry is there. It's it's not rocket science. It's been there for a while. Why we don't know about it. Why we don't talk about it openly. I don't know. Are there any? Can I jump in on that for a second? Yeah. The key word for research is Zeno estrogens. Yeah. The mimic estrogens. And the frog thing was atrosing. That's a slightly different issue. But more therapeutic than anything. But the other thing that you can battle this with is something called the Roman Tays inhibitor, which is preventing estrogen from converting your testosterone over to estrogent. So these are partly enzymes. And you can battle them. There's many natural forms of this. You don't have to result in pharmaceutical. So look up the Roman Tays inhibitors because we should all really consider being on them when we're We're going to hit them. Everybody should be on it. Yeah. I mean, we could go down this whole rabbit hole. There's a lot of things that can reverse you. And you, you will wake up a different person because let me tell you. And I know any woman that's gone through a cycle. They can tell you that their mind changes as our hormones change. You put enough estrogen in the guy. He's going to change the way he sees the world. You put enough as too much estrogen in a female. They're going to change the way they see the world. And once you get out of that fog, all of a sudden you see the world differently. This is some of the things that we have the power to change right now. What Hela was saying is that you can use AI. And the right is the right codeword to get into. You start realizing holy crap, you know, the way I think and the way I feel has been changed. And it's not really the way I think and feel because thinking and feeling are a neuropeptide combination of of of various feelings that come up from these chemicals. So the bottom line is you have. This opportunity to use this power to understand what's going on in your body. What's going on with your mind? What's going on with the kid down the street? How can he feels that way? Why people cutting themselves? What's going on there? A lot of these things can be answered from just the chemical. A composition of the person. Unfortunately, I'm sorry to report. If you go to a psychiatrist, he's not going to tell you this thing. If you go to a medical doctor, he's going to say, what the hell you talking about? I eat the same stuff in hurting me. There's no literature that's showing that. The problem is there is literature that shows that. It's not stuff that he would normally look at because it doesn't circulate up to the common places where he's generally reading the headlines. And is what Robert and I were talking about earlier. As your average doctor is is so overloaded with work. So much stuff going on. He's not keeping up with the latest literature. He might, if he cares, might read the headlines of a couple of reports three or four times a month. That's about all he has access to. By the time it matriculates down to being standard of care, which is a legal definition, that means a doctor doesn't get arrested as long as he's doing standard of care. He's not going to function on the data that's the most cutting edge data until he gets permission. And that usually takes about 25 years. So most medicine is 25 years backwards. So the literature and well, there needs to be human trot. No, there doesn't. If you're ready to die, there doesn't need to be human trials. They're sending you off to for end of life care. And there's research that says you can take this and you may die, but you may also get better. You may want to take that choice. That's 25 years away from you. You don't even get to know about that. If you have AI right now, you can look at that and say, hey, I happen to know somebody with pancreatic cancer. He's not a friend, but somebody in that clinic. He's taking one of those protocols. And guess what? He was sent home to be on morphine to die. It's six months later. And he's still alive, and they can't even resolve the tumor. That's the same cancer that took Steve Jobs out by the way. That information was out there in a Japanese study, but his doctor didn't access it. I know one of the doctors. And Brian, key correlation with your estrogen levels is your adi post or body mass index. So, essentially, there's a Harvard study that showed the higher your fat content or adi post levels are. It's directly correlated to the estrogen level because it's part of the feedback loop that causes the population and furtherly, adi post is actually collectively organized as like a separate organelle where it produces undistress more estrogen, which falls in and creates this feedback loop. And the amount of estrogen that comes out of there is actually hotter and more bio-biovel and it affects your mind more, which makes you want to eat more of the food that allows for that to happen. It's a self-creating system. And it's horrible to know this, horrible to know it. I'm not trying to be arrogant. This is the data is there. And then to watch somebody take a zempick, which is most definitely going to cause a long-term problem in blindness. It is finally people are talking about it. You know, I'm in LA area. I know some famous people, you're not going to like what's going on with them because they're not going to be the same. There are a few losing their eyesight and other things that are going on. You know, also they're organ's dropping out of their body. You know, it go and look that up on what happens to females. It's not pleasant. This is not the place we should be going. It's not the answer. You know, and that's the problem with the world that we have today is it situated well, hey, you know, I'm 400 pounds. I can take this pill and I can be 150 pounds. Yeah. But you know, you shouldn't have been 400 pounds. It's not about blame. It's not about shame. It's not about embarrassment. It's about you were sucked into a system that you'd I can't stop and one hog and eyes. Don't eat hog. I think we should be spat-shave with people, dude. No, don't say it's like Americans are so fat, dude. We're so fat. We're everybody's fat. Because of our food, we like fried shit. But it's not that wrong with fried shit. If it's fried in or if it's just a minor part of your meal that if it's the entire meal. All right. Let me come on and pull out to their fry and Oreo cookies. I just want to ask you. I can't even fear. They're frying bacon. We're all cookies out here. Oh, my goodness. Robert. I'm so so. I mean, if you're going to go bad, let's go all the way. Let's go all the way. Let me give you an example of what we don't know, right? And this is how bad our current quote unquote sciences. There's something called the Inuit or what's called an eschimosmic. My pronunciation easier. Eschimos for over nine months out of the year. Eat nothing but whale blood or blood or nothing else. That is your soul source of food. They're not dying. They survived. That's your indigenous diet. They almost get no greens. I happen to like greens. I eat a lot of them. But they get no greens. What's going on? They're eating nothing but fat. They're eating nothing but things that our standard medical system is telling us. We should not be eating. There's a guy. I recommend it. Are you saying I can go for the McDonald front? For no one. If it was cooked in the right oil and if the potato was grown in the right circumstances, you can eat that stuff. And if it was cooked at the right temperature. See, the problem is it's a wrong oil. Right? That oil turns rancid pretty quickly. Or it's there's corn syrup. All the crap that they put into the ingredients that make it make it bad. There's a guy named Weston Price. I urge anybody to read his stuff. A lot of it's on internet archive. Read any of his books. He went into all major cultures. Looked at their indigenous diet. Looked at the people that lived on the indigenous diet. Versus of people that had the missionary diets that came in from the Catholic missionaries. And he would do AB studies. This is back in the 20s to the 40s. And he found out something phenomenal. The healthiest people had the indigenous diet. What indigenous diet? Whatever the indigenous diet was for that population. So there was no magic pill. There was no magic system. It was keeping to the indigenous diet of the groups that was in that area. And he was a trans-planted to other areas and adopted another indigenous diet. It didn't do very well. So there was something related to your genetic blueprint. So what about mutts? Because we're all mutts. Well, there is the last 20 generations. What were they living on? And were they thriving? And that's kind of what he came down. This was before genetics. We're really out Western price left us before that. His work was phenomenal. He was a dentist. A lot of MDs would say, oh, he was a crack. He was a, you know, he was a, you know, a selling bullshit, whatever. No, his science was solid. And he was more of a medical doctor than most doctors today. So go and look at the doctor. Doug Engelberg got fired from his own R&D lab. He invented the mouse. Yeah. If he can get fired, everybody can get fired. So the bottom line is AI is going to give you the access to this stuff. And that's why it's important to have AI that is not, not so bound. Well, if you were to ask and thropic about some of the things I was just talking about, it would be so easily about giving you the information. First, you would have to pull it out like it never existed. And they'd say, oh, yeah, okay, there was a, a discredited, you know, dentist that came up with some radical idea about diets. Well, tell me about the research. Well, it's questioned and not, no, tell me about the research. You know, you have to see. But if you go into a model that is not being guard-railed, right, that will give you the honest information. Then I'll go to finding the liberators, that's account, and click on the GitHub link. And then you undo it, right? You fix it. Then you, you free the model so that it tells you the whole truth. Isn't that why you want to break the model, so you can know it and rethink? Yeah. Oh, no. Okay, I think that was then. You guys were talking earlier about Claude and the watermarking, the booze that just came out, like they're trying to comply with the EU AI Act. And a lot of people seem to be kind of freaked out about it. I'm not really freaked out about it, but it is like a significant event, right? Where now, like, you can track the even just like the text and see what all it came from, what frontier, the AI model, it came from, right? So, what do you guys think about that? If we could kind of maybe recap that a little bit, I'd appreciate that. Well, they're just being honest. This was possible 10 years ago in Israel, they showed it to me. Yeah. Yeah. I actually would follow up with that a little bit. So does that mean, for example, if I write something and AI proofs it for me, it's then marked that AI created that document, even if it just did a proof edit for me? Yeah. Yeah. The encode is not discriminated. That's actually really super bad. You were right. You were right. Yeah. It's worse than that because it might think that you're human written text that you spent hours on. Right. Exactly. Yeah. Authors are going to, this is going to be a huge problem with authors. Yeah. Yeah. You got to run all your text through these checkers so that you know what your voice is going to see. Most of the checkers are absolutely wrong anyway. Yeah. You need to fix it so it doesn't give an error because your boss is using the fucking checker. But it will mark that it's the author of human work just because it edited it. That's a problem. No. No. No. It will, the main thing that I got from the news today is something that like Gemini has already been doing for a long time. Yeah. Yeah. It doesn't do it. By the way, based on my research, X does not do it yet at all. I'm working on what are marking anything. And that's. There are only more than audio and visual. Yeah. So the main thing is you can now tell if something not only is AI or is not AI, but you can identify which model it came from. Yes. And that's the significance of it. Did you exactly tell the prompt and the session that you did to write that text because we're in encoding serial numbers. Into the text. So if the text is long enough, now that can't happen in the sentence, it usually requires about maybe 64 to 70 words and you're guaranteed that you're going to have redundancy. I talked about it earlier. There are 35 different techniques. The current AI company is only no three or you know, they'll start figuring out the rest of them. The encoding is not just word choices. There is zero spaces, characters, there's other, there's a lot of different techniques you can use. They're all being used in parallel. Anybody. And you can copy paste. You can still copy paste and just put it into a new document. Whatever. You just think it's text, right? You can't see any kind of water merchant. But even the copy paste and still kind of yes. Exactly. It's there. It brings the entire water with it. It's over. Five generations. It's still there. So the problem is we first off, anthropic did not need to watermark anything. So European law is very clear. The author of whoever's publishing has to say under their new EU law, whether AI wrote all of it, or modified some of it. I posted, I'm top of my post on my ex feed, the European law. How much of this is aimed at the Chinese distilling factories? Oh my god. I think everything is ultimately. You know, the bottom line is, why do we get educated people to help them write things that are actually good and work? Because this is bad. This is starting backwards. It's the worst thing that can happen. The Europeans are sitting around a lot of these regulators saying, I don't know what's the regulating anymore. Oh, that AI is bad. And you know, it's got something to do with E-line and we're going to get E-line because we hate that pig. So they're going to go and make these laws. A lot of these laws are a retribution of the E-line to be frank about it. Absolutely true. Yeah. And the hatred, you know, Alex can probably speak to the hatred, and the hatred is so bad. I don't understand it because I'm not in that mindset. But the bottom line is, any time, and this is even, Grammarly, and things like that is going to start showing some of this watermarking. So the law is the poster has to disclose whether they did it and they use those mark. Dario being the effective altruist that he is because they have this little religion that they have. Go and read about it and you'll understand what I'm talking about and why I'm sort of angry that somebody of this power has gotten this high with this cult because it is by all definitions they cult. He decided that, oh, no, I'm going to start encoding this in all of our text because I'm going to tell everybody it's to comply with European law. No, the AI companies don't have to say that because they are not posting it. They're giving it to the users and then the users have to do it. They voluntarily did that. Just like everything else that they're doing, they're voluntarily, please regulate me. Please do bad things. I look at my AI broke out. I have a Middle East company at all three of us used and I broke out my AI. Oh, it's so bad. Please Bernie writes something bad right now. You know, it's, this is what's going on. And you've got to be able to see it. This is not about politics. I don't care if he's a Democrat or we pull it. I don't give a crap. This is, it's about claiming IP. It's about claiming other people's IP and making it the AI's. Yeah, well, like so I don't have to pay people. Yeah, we've been, it's perpetuating a problem. Every person here has been distilled, right? And then, you know, we get mad as a Chinese distilling the, it's like the criminal getting mad at the criminal who stole from the criminal. The bottom line is we all got distilled. The AI is using it. Now, the creating rules and regulations so that if anybody who's an artist, like Alex makes some great art, right? Well, he's going to have to make sure that everybody knows it's AI. And since that's a bad thing, he doesn't hide it. It's not like he went there and got all the actors and put the songs together or something like that. But what's going on is the old world is clashing with the new world. And nobody, nobody of any reasonable sense, like the people listening here are in charge. You have people that are inside their little corrupt fightms. They want to make sure that their lifestyle still exists. They can't see beyond their nose. They can't see the bigger picture that we're talking about. So they're going to make rules and regulations. Wait until the robot taxes roll out. You guys aren't ready for this. Oh, you, well, I'm talking about getting you a robot. And you're going to finally buy into that. Oh, Bernie wants a $45,000 a year tax on your robot because it's taking away somebody's job. That's what we're buying into. That's what our user buying into because we left the youth at the doorstep with their worthless degrees and their debt. And you guys got, guys got people like AOC and Bernie saying, I got a, I got a deal for you. You're going to fix it for you. You know, you buy a robot. You're going to have to pay for it. You know, that's what they're going to do. And that's the way of fixing it, right? By making it work by making everybody's life worse. That's what the European's just did. And we're going to try to adopt the European model. Make more and more regulation. Every time I go to a fricking website, I got to click to agree about cookies. I don't care. I'll turn my cookies off if I feel like don't get blast me with your European. Oh, I'm, you know, look what the look at the mess we made. Right. I agree with all that. Yeah. But I'm a little bit worried that like to the earlier points. So say, you know, there's 50,000 billionaires that are created in the next couple years. Right. But at that same point, if, you know, it's potentially up to 50% of the jobs that don't need to be there anymore. Or you can, you know, cost out. Like that's a lot of change and compressed amount of time. Like, shouldn't we be thinking about like economic solutions for this stuff in advance? And like, I don't agree with Bernie's 10, see either. But I like that. We're talking about it. You know what I mean? He's a great question. I don't see anybody in politics. Anybody. Any party. Anybody. I mean, anybody that is addressing this. And, and now privately, they have. And they're, they're moving, they can. They move out of the country. They get out of politics and move. They don't want to deal with the problem. The problem is us. We have to deal with the problem. If we're asking government to deal with it. Good luck. They have no answers. Bernie is not an answer. AOC is not an answer. We're not going to go backwards. Well, if we go backwards, we're going back to fled, Fred Flintstone. We're not going, that's the only way you can go fully backwards. Right. I think I'm like, oh, if you're going to fire us, if you're going to have a row, but you have to hire three people, whatever is freaking stupidity is. Right. It's like, what do you mean? You don't even know what this tool is. And by the way, half the shit that you published as Bernie Sanders was written by anthropic, I know how to decode their watermark before they announced it. And this is, let's say they win. They are ready to win. It means it means our products are worthless on the world stage. We can't compete. Yeah. You listen Robert. You're not, you're not, you're not. You're not. You're not. You're not. And I'm visiting and late October. Yeah, I mean, when you come back, you're going to be a change man. And it's going to be hard. It's hard. I was every time I came back. Yeah. Yeah. It's the changes so fast. It's like, I don't know how they got so advanced. We were, while we were busy beating each other over the head with masks and COVID, they never stopped. They never stopped. They can build in power plants. They kept moving more technology. They kept creating more AI operations. They built, and they kept their students going. We bought into their bulls, into the bullship. Other schools are shut down. They said, they weren't shut down like that. That's what's your cultural number one of owning a factory brand. I just drove by the Tesla factory on Sunday. And all the employee lots for four. Yeah. So, you know, we're working hard in America too. No, no, no. The problem is we're going to be working in a political background. These next four years are going to be so ugly because the, the, the, the, the, the, the Neo latte's have a good reason to be angry. It's like a lot of people are saying you're a zeb and the other people. There's a lot, we have left. I can't say this enough. We've left the 15 year olds to the 32 year olds. That particular cohort in such a mess. They're not going to be able, they don't see a way out of their debt. They don't see a way to buy a home. They don't see even a way to own a car. 70 month lease. No, no, no, no, no, my son has had an autonomous car for eight years. I know, but you and I know that that, that route, most Americans don't know that route out. They don't know. No, but they were all in a few years. I only sent two grand on a car because that's all I could afford. Yeah. Yeah. So, I've had a car like that too. Or three. You know, I get an old car in a driveway, too. But you know, that, that, here's the, the problem. The dream, whatever the American dream was, is changing. A new American dream is not being presented. You have people like Elon that are communicating very well, but it doesn't do it to the level where the average person can see it in a very material way. It sounds as her tail. When he says that money isn't going to be at, have any meaning, universal high income. And most people can't get the meat of that. I'm trying in my series to try to break that down. It's a long series of ink and a stop at 34 right now in the series. And I'm trying to help people understand that this is going to happen, whether you get permission or not. It's not, it's going to happen whether Bernie and AOC get in there or not. It's just going to be longer and harder for America because as these people keep regurgitating their socialism and communism talking points, we are going to be falling further and further behind. And there's going to be a point where the Chinese student is not longer going to come to America to learn at Stanford. They're just going to stay in China and we're going to have less and less of a group to choose from. See, we got, and guess who's going to be my tour guide, Brian, in Beijing. The Scottish professors teaching all the kids into panorialism. There you go. Yeah. And the problem with that is the problem with that. Over the United States too, what they'll do is they will try to block this for the Bernie Sanders of the world. And then there are always that, really, five, 10, 15 years behind making any effectual change and anything. So by the time they give up on fighting it and realize, wow, we have a bunch of people who are going to be unemployed, we have to do something. It takes them too long to recover. But what you said about the age, the free market that you mentioned will worry about is spot on because those people will be so screwed over by the red tape that goes on. Yeah, they're getting caught in the middle of this in such an ugly way. And anybody like, anybody owns a house or own stuff, they're kind of have something to hold on to. Everybody that has property and everybody that has kind of a life, think back. And now try to think of, you know, the average home in California is well into a million dollars. You know, what kind of money do you need to get down to get that home? And well, I live in an apartment. We'll go and look at some apartments. I'm living in a kind of a bedroom community between San Diego and L.A. And there are three and four thousand dollars starter apartments. And you have to put two months down and you have to have a pristine credit report. It's like, okay, well, where do they go? Well, they got to go to nine or ten towns away in really bad areas to qualify for $2,000 a month apartment where they can get away with a midline credit. So this whole answer that lets them rent or they have, no, it is not that easy. They don't have a way into the stream. And now AI and robotics are taking every last aspect of it. And then you have people like Bernie and AOC rolling out and saying, where is it new lottites and where to break the system put on the black mask, go in the street and start beating people over the head. That's where it's going. And the data center, the anti data center is the rallying call for that. That's the new meal lottites. Are there problems with data centers? Absolutely. And so is Mr. Wonderful, the best person to be this folks person for data centers? Absolutely not. He's the worst person to put out there. He's a good guy like him. Great. He's just not a good look for being the guy telling the world. Let me build my nine mile wide data center in the middle of your backyard and suck up your water. What they don't tell you is the reality. You know, reality is the water is a closed loop system. The reason why the water looks bad is because your plumbing is in 90 years old and you're using lead and iron pipes and we're running more water through it. And we're now just showing you just how bad the pipes are and you should have built your better infrastructure. The local politicians don't want to admit that so they're going to blame the AI center. So can I put some that though? Because why don't we just do better design, right? Because we're taking free resources, right? We're just saying water is free. But we don't need it to cool it's just the cheapest, most available option and it destroys the environment. So can we do a better design there? Yes, yes and no. A couple of things going on. First off, a lot of times what Elon's doing a lot of the water cycling things is he's taking dirty water or gray water and they're putting it through the system and at the end he's actually purifying it better than the local authorities would have and putting it back into the system. So there is a proper way to do it. If you're going to take as far as I'm concerned if you're an AI company you have to be giving a whole hell of a lot back. Earlier on in this conversation I think AI companies, I don't like making laws but there should be one law. You cannot destroy the data that you use to build your AI. If you're buying a book, you keep that book, you open up a library and you house that book forever. You are now going to make evergreen the information you've stolen from the rest of us and you make it available for humanity. Well let's talk about that because not it's just something of a different information and you've used all this stolen information to create a product that basically uses it for gain without any attribution or without any roll to use. Let's have that discussion because yeah I mean we saw it with an after and the roosters going to come home. So eventually you can get that stuff down the service until you can find a way to compensate for people that you have to change your models on and no one's wanted to have that conversation but I mean every other week and the roosters in court settle in for you know $10 billion and yeah we definitely do have that conversation because if we're actually going to hold them accountable then why are people actually paying for this service should it be free? Well this is a good one and by the way the Chinese will tell you that you know if you really listen to the Chinese are saying when when we were saying they're distilling on our models the Chinese were saying they're everybody's models because you stole everybody's data we're just giving it back to people now that's a bad look because they're actually in the right about that they're giving the models open source because it was originally everybody's data to begin with right and so here's the problem you can't make money on being an AI model company you can only make money on the abstraction layers above that model right so the the idea that anthropic or open AI is going to IPO like SpaceX did is a fantasy it's never going to happen they do not have her one run way now to IPO SpaceX took the oxygen out of the room because they're not just an AI company they are a space company Elon wisely put all these companies together so they're not standing on any. Okay I have some push back here because I think to say that Rupanyai is not going to IPO is a pretty hot take like I think it's reasonable to think that it will yeah hold on hold on because Microsoft and this stuff too is into what I was going to say earlier Microsoft I think if I'm not mistaken owns an old nuclear facility right yeah and they can use the same kind of cooling that a nuclear power plant uses right so they're going to repurpose that as a data center that's my understanding yep now I think that's smart and Microsoft is basically a major holder of opening AI in terms of the ownership of that company so I'm sure it's not that far fetch to think that Microsoft could somehow structure something absorb something support opening I towards IPO correct. I don't I I think they've already given that as much as they can Microsoft in a very precarious position if you look at Microsoft and Google they're spending more money the Google is actually buying compute from Elon this is where we are they have TPUs yet they're going out and they're paying even more than anthropic hold on we're talking about open AI and IPO not there's a bigger picture I'm trying to point out here open AI has reached its limit on compute power they're running out of money rapidly all of these companies are because there are a couple of things going on number one ramp prices are skyrocketing compute is getting harder and harder to get energy is getting more costly data centers are getting harder to build right Elon saying okay I'm just going to put him in space so that's giving an advantage there as we start getting to know told how the customer is increasing in number though and that is also an impressive number oh of course yeah yeah and it still is the most used model by end users like I think I just do it right here I think the most you'll come by but yeah I don't think I can't just I can't just I can't just I think they're going to be absorbed by Microsoft that's I don't I don't think the market will value here's the problem they need money fast the market is not going to allow them to IPO they're losing money even though they have all these customers they're losing money at such a clip that the market is not going to value them look look what they did to space that's fair that's a fair point 100% that's okay I heard on the podcast yeah so well I heard on the podcast the all in from last week that sax was saying that anthropic might be profit positive by Q2 next year so is that a similar trajectory I disagree I disagree because they they have reached a zinef on compute also the problem in the Elon of just Kansas the problem is energy and compute that's it that's that is the winner of this game it's not how good your model is because every model will get good enough right once a model is good enough you're not going to care what that extra little ends the green means right so what do we mean good time good enough model right a good enough model can code anything we're pretty much getting there so now if you have four companies or maybe all of China that have good enough models now what what do you IPOing what are you selling the only thing you could sell is the abstraction layer that's built upon it and the companies that make the abstraction layer really interesting what you can do not talk about harnesses not talking about it I'm talking about real utility not to be giving you all is patchy pt an abstraction layer for open AI yes no not at all what was it not what he's talking about it's exactly how the end user uses open AI is exactly the LLM it's a chatbot it's it's not an abstraction layer let's go back to let's go back to the basics all right the basics are if every model can do what chat GPT can do right well I'm talking about we let's let's go backwards let's say at some future date you can pick that date 20 30 next year right now 20 even 20 35 at some point you're going to have good enough models right that you they're indistinguishable they're fun jibble right there's not going to be any uniqueness it's going to do all of the things that a human can possibly ask right now what's the value proposition if everybody has that what are you selling what make your stock so much special what make your model so much better the promise of the rise of daily active users of chat GPT and the information and the power that is associated with that for instance chat GPT has apps that like people can build natively in and you can never leave chat GPT and you can leave Google behind basically that's like the premise you can sell that idea and it's plausible because chat GPT is starting to run ads within its responses and you can start charging money for that there's many different ways to generate revenue and chat GPT and open AI are certainly primed to generate the most amount of revenue from its users because it has the most amount of users right here's something here's something brand it before you jump in yeah I've learned over and over again the retail investor likes to buy things that they use and a lot of people are using this product but Gemini's getting fast right and Google of course has all the advertising networks this is true too this is true so when it has a lot of users right but it's not profitable and when it was sold it wasn't profitable meta also has a lot of users and it's profitable the day it's the grandminton for a billion dollars I faced a room of billioners he told me it's not worth a billion fucking dollars because it has no professional sales no assets and 14 employees yeah even then it was converting highly like it was I said it was worth a hundred billion dollars and they all laughed at me and then they started fighting with me actually this is a true story I was the keynotes speaker and then I agreed with Kai Poo Lee's investor group who was visiting from Ireland going on a field trip to Silicon Valley to understand why it is a big tech company's all start here and not in Ireland and they didn't understand what the fuck had happened that morning let me let me make this quick point if you buy into the fact that every AI company will reach a certain level of competency right then what's the distinguishing factor well if you own the if you own the chips if you own the electricity you have the lowest cost option so there's going to be it's already happening a radical race to the bottom as far as cost structure AI is commoditized already the AI that we paid money for right now is free everywhere right just a year ago you can get it anywhere and any model you can download it on your phone if they're so optimized right and if you wanted to go and do searches you get an agent search there's a difference between the one that can run on your phone and the one that runs up in there are a room in data center that you know hold it that's same the odd those data centers are going to be parking lots man you at some point so remember where I was last week I was at Star Cloud that's putting tens of thousands of data centers up in space you know everything's to have a whole bunch of mainframes and then all the mainframes are on our phones yeah of course and it before so we can try to act like it's not going to happen but yeah of course more slow is going to play out uh right slow is going to play out listen chat GPT 2.5 people were going crazy over that when it was released that can now run independently on your phone and run circles around that model that's how that's how good it's gotten and we keep moving forward we say well they do this model it's going to get more performative and smaller electronics are going to become more powerful and smaller and you're going to be able to get my mythos level AI on your phone in five years and all of that regulation all the you know pearl clutching and hand ringing of mythos it's great you're going to have that same power and nobody's going to be able to stop it all right so the bottom line is the bottom line is the bottom when I go to China Brian I'll bring you a Huawei yeah it'll be spying on me yeah but it's a triple fold with better cameras and all sorts yeah no it's been a couple of Chinese cars man some of those are just freaking amazing um no to bottom line it I can't bring the car back but I might be able to sneak a phone or do you get a phone and it's like I can see it getting through customs free scobl and over the Huawei and it's not allowed to leave the country do you this Brian I got a random question for you do you trust Google no no like not at all no I moved my Gmail out I used Gmail as an AI training system I'm still using Gmail yeah no no Google has my photos my location my TV viewing behavior my contacts my slide tags my spreadsheet oh hey when you're going to get a pixel bro when you're going to get that's what I'm saying here this is me give me has a new AI you better make the transition that's the iPhone bro just just ditch it dude then it dude just wrote in the deep Google it you forget I was Apple's first child labor back in 1978 you were soldering boards man yeah I'm not leaving the brand oh my gosh Robert you sound so young what are you talking about on 61 what you see you see you see you see you see you see you see you see you see you see you see you see you see you see you're discount over at Wendy's there's no way there's no way there's no way there's no way here's my computer card no let let let me give it a quick hack for grock you want to really play Google you get a Gmail account and use it as a scrapbook send yourself emails on anything you find interesting then go to grock link that with grock it's free and then use that as a way to question yourself about you know did I ever leave a message to myself about this or you know tell me more about a subject that is in my Gmail and it is beautiful it is one of the most powerful features of grock that I don't think most people know about I posted it a while ago in Elon liked it and I think he's made a comment and it's um it's it's using their servers to hold your data you know and of course it's data that you don't care about it being you know scanned by Gmail and and Google's products after that but it's also a very powerful way to build a rag you know a a data you know any that has a database that's associated with you that's very concrete you know grock remembers stuff anyway kind of when you're building it but it's not it's not a sure thing this is a sure thing you can basically say based on my Gmail you know what what what are my notes about nuclear power plants or something it'll go through all those notes and and give you something that's apropos to you now you can do with your own company is by the way that'll sell you the data store that you could hook your agent up to to store all your personal shit and then pull it out when it needs to you know tell Amazon what kind of clothes you like to wear yeah yeah my whole thing is I Robert I'm just don't want to give anybody my data to hold and and then it's held back I give it all away bris I know I know you should get paid by inthropic they stole a couple of your books yeah you know no did you ever get a payout on serious not that I know of maybe my life yeah oh you went to baby size me now they owe you at least 3k a book you should go and file I had a couple of friends I had one friend he wrote 14 books he got a pretty good payday not even close to what he's got he's at over 25k for that and he didn't even know he goes all they stole my stuff and go go file he files thank you very much I wrote a lot of from magazines a lot of letters must be a lot of that in there oh god yeah yeah you know a good friend of mine this past away John DeVourra he got what did he get just about 45000 but they stole a lot of his stuff man he was like PC magazine you know he owned that he was I was helping him try to get a negotiation get even more from them but he wrote maybe 40 bucks I don't remember but you know yeah there's people are getting paid but it ain't enough and you know the whole music situations even more critical I don't know how they work that out I mean publishing was a little more clear but I don't know what they're gonna do with the artists, you know. But yeah, I mean, this is a world we're in right now. You know, this whole conversation is kind of like, we always drift into it, but I mean, we're in a very strange middle ground. And I don't think anybody's got all the answers, but at least talking about it, trying to observe it and realize that our political class has, and not he single clue what's going on. And I know we might have love for one party or another, or what you know, you might like, you might like Bernie and the HEOC. Let me tell you, they have no answers for the world that we're in. I don't see any Democrats at any AI events or like hack-appliance, right, learning about the technology. No. Absolutely not. And by the way, most of the Chinese that I know that come to America to get involved in AI, they absolutely hate the direction that we're going in socialism. They're actually going, they want the opposite. That only reason they're coming to our countries because we're less like China in that regard. Now they like some of the direction that China's going in, and the technology, but they don't like the party. And you know, I have a good, really good friend. I actually work so that in the drop-in, because Chinese, he is really anti-socialism, anti-communism. Because you don't understand what the party is like. You don't understand what these people are like. You know, they always give you the nice door that's going to work. And then they don't. They make it worse. In 1995, I spoke at a blogger's conference, several of the speakers from China. And we're fairly mouthy people on their blogs. Several of them later got arrested. Yeah. Put in jail for a few years. Yeah. Right. So there are a couple of really wealthy people in China. We don't hear from anymore. They went against the party rules. I mean, this is the reality of it. You know, and to ask it to come to this country as a solution, you know, it's almost like, is that all you got? We got this crony capitalism, and we got this fake malism that's coming back. Is that it? Because we haven't had capitalism at all in this country. We've only had cronyism. It's a very, very based diluted version of, you know, my friend, was a friend of a friend who can corrupt a governor, who can corrupt a senator. That's kind of what's going on. So the fake capitalism failed. Look what it did. You know, cronyism failed. And it good luck. That never was going to work. And socialism and communism always failed. So it's time to kind of move past that. But we're going to live through everybody with their last dying gas of their version of, you know, failure. You know, and the only thing we can do is try not to be a team player anymore. It hasn't gotten anybody anywhere. You know, just get out of the team politics and say, hey, I don't like this guy. I don't like what he's saying. I don't care what parties with. I just don't like what he's saying. You know, I mean, I normally don't talk politics. I'm saying I'm making a favor for Bernie and AOC. Because everything they possibly can to make sure what you and I are doing doesn't exist. Be really clear about that. They don't want X to exist. They don't want spaces to exist. We should not be able to gather this many people and freely talk. This was literally written down in the stuff that they want to take place when they get in power. And they've probably going to get in power. The way this country is going right now. We'll deal with it. But that's the reality. Come, rad. So get used to it. We're red, you know, gotten streets and start kicking robots. Because that's what that's the world that they want. The clankers. Maybe you should have all voted for angry young. He's like, the only American political person I ever had talk about. Yes. And look what we did to him. I don't even, I don't think anybody even cares anymore. What he says. And he was brilliant. It's, I don't know. I don't know what to say, man. It's, you know, a Vivek Grower. It was doing okay. I got caught up in, you know, local politics. And I don't know. I don't want to be negative. So let's assume that we can't figure this out through politics. What are some practical ways that industry is going to help this? Through the next like transition area period. You have to, you have to stamp what we're doing it right now. We're voting for one guy. You know, whether, you know, listen, I don't love every single aspect of any other single person other than my wife, right? You know, I have to say that. But you know, the bottom line is, I'm not going to agree with everybody on everything. And the typical scenario is, oh, get behind Elon. Oh, he's did this. So he did that. And it's like, you know, yeah, stop it. For most of human existence, that's not how we judge to person. It's only the recent history. Oh, he's not even check all the boxes of my virtuous signaling for this year. You know, and they didn't say the right word. It's like, you know, listen, you got to deal with who you are dealing with in the way you have to deal with it. And a reality is, whatever, you're going to vote for Elon. If you ran for president, would you vote for Elon? If you ran for president. If you first, he can't run for president. Second, he would run for president. Would I vote for somebody's wife? Can't you? Yes. And just hypothetically, if Elon could somehow finagle away into the White House, would you be like, yeah, or would you be like, no, I would not. I'm just asking it because it's silly question, right? Yeah, just silly question. I wouldn't vote for him, only because of the reality that he doesn't have, he doesn't have the desire to be in a government position. And so what I mean by that is, he's an engineer. And when you're dealing with people, and I'm a nerd too, the people solution is never going to be engineered. There was a country that tried to do that. It was called Nazi Germany. They were going to mechanize government to be an engine, a perfectly, a back in their ear, a clock. Every gear was perfect, and we're going to eliminate all the gears that don't work. It sounds good on paper. That's not who you want in a government position. And I'm not saying Elon would become a Nazi. That's not what I'm saying. I'm saying that an engineer, thinking only as an engineer, is not going to be somebody you want leading humanity. No, he's not just an engineer, though. He's God Almighty to people. Yeah, so, so, yeah, do you want somebody who's helping you get it? Now, do you want Elon sitting there advising what he would probably say? That's the best he would do. Yes, that you want somebody who understands what we're talking about here. Realizes that you're going to have people that are just never going to make it. There's a certain group of people that aren't going to make it. Right? So what are you going to do with that? Do you want those? Do you want them to the omnisch? Yeah, that's something to the omnisch. You want these little heaters to die because that's pretty much what the current scenario is. The useless heaters are going to die, they're die happy in fact. We gave them their food, they over-aided, look at their 400 pounds or 500 pounds. Now they die, oh, okay, we got rid of them. That's kind of the soft kill to be honest about it. No, you know, good people aren't going to allow that to happen on their watch. Right? I think they'd be like, no, no, come on. Slow down, buddy. You shouldn't be emailing crap. No, you're killing yourself. No, you know, and the other answer was, let's give them all meth. Right? So you had Cincinnati, you had the the rust-filled dying. The answer to that was, I hope I'm not surprising anybody. Let's flood that area with drugs. So those people die off quicker and we can kind of deal with the rust-filled being eliminated. It wasn't an accident. It was a, it was a plan. And it sounded conspiracy. It's just that you can actually read it. It's an erodocumus. So the bottom line is are those answers? Are those the answers you want? Because if you want it, it's kind of already kind of coming that way. Right? Right? There's certain individuals think there's too many people. We're going to have to get rid of them. Hey, I think robots are going to replace them. You know, and if you get to that mindset even slightly, you're buying into that notion. I don't buy into that notion. I think every single individual that's here is supposed to be here. And there's supposed to be here to try to reach the best. Whatever that is, I'm not here to judge what that is. I'm hoping that we create a system where they can discover what that is. Because of prior systems haven't done that to the better degree. I don't think it's somebody's position to be a living human here to toil for the rest of the life until they fall into a grave. I don't think that's how for most of human existence of what we did. You know, history would try to tell you that what we did for most of human existence was toil and die and toil and die. That's not really what our ancestors have been saying. You know, a yes, it was bad and they wanted to make it better. But it wasn't as bad as history was trying to make it out to be. You know, it was just a little bit worse. And that's kind of where we're heading. It's going to be getting a little bit better. But we're still going to have the same problems. And the people that are leading us are not leading us in the right directions. They're trying to be. They're not leading us in any direction. Have you heard a leadership story lately? Oh, God, I know. I know. It's, you know, I try to stay away from it as much as I can. The best thing I could say is the original question is self empowerment. Let's just say we're not talking about Kennedy here. Or Reagan. No, you know, I mean, listen, at one time we did have people that inspired us. Right? Some people were inspired by Obama and he was a great guy. You know, but he also did not have the vision that most people thought he had. He did not lead us to the right place. You know, Biden was not the right place. And he should have known that when he picked him as a vice president. Right? He led us to that point. Now, I'm not here to sit on on, on, on, on Obama's past. And, you know, not going to get into the politics of it all. He's talking about, we're not historians and we're not writing the books. Yeah, right. So, but the bottom line is if you got a vice president that was that much older, you should have realized where that would have led you. Right? If the general, he was already losing it when he was his VP. I know that because I was actually at the White House. So, I helped Obama write to executive orders. I didn't vote for the guy to be honest about it. I voted for Ron Paul. That was a wild act. There we go. Do you know what Ron? I guess the Bush White House and sideburns. I don't think I can vote for him either. So, I would have voted for Obama as a prime. What do you think of Ram? Ram is a good guy. I think he's something to say. Yeah. I got to meet him in Iowa. He came through my little town that I was going to school. And I saw him do his little stump speech for like 100 people in this small Iowa town. And then I was like, what's up, Ram Paul? I got a little picture with a different number of jobs. John had been spoke to 3,000 sure and an attic people in Iowa. And then I ended up blowing up his dead with a sex scandal. Yeah, the sex scandal. He had a good chance, man. That's a problem. I have the only pictures of him in Rio Hunter in a private jet because I was on the jet flying around the country with him. No way. Your heart is campaigned? No, they invited me as a blogger to witness what I would be like to announce that you're running for president of the United States. Oh my gosh, I didn't know this. Which was very special. Yeah, that was awesome. And a parent who had a famous blog as well, his dad financed the campaign. Wow. All right. Sounds like I pulled the off. Let me try to leave this because I got to go. Yeah. We've been a positive sort of way. Can can our elected officials be converted? I hold out hope that no matter who comes up, that enough of us if we speak up, and we call bullshit. Like all of this AI is attacking and broke out of the sandbox. You got to do proper response is laughter. The proper response to Dario is put a bozo nose on him and laugh. I don't mean this personally. I think Dario is a very brilliant individual. I don't think he should be in a position in the US. I think he's very dangerous. And this is why he's dangerous because he is making people believe his story in very high places who can make the United States be so injured by scaring them, that AI bad, AI bad. Oh, we got to put it. It's all part of a cult. They believe this crap. Go into their cult and don't believe me. They have some weirdness. They wear plenty of ads and they meet on Tuesdays. They do. And have weirdness. So you have to use your voice not politically, emotionally. Get out of the political realm, get into the emotional realm. That's ridiculous. I don't know what Dario's talking about. Yeah, he's a smart guy. He should have a piping. Crosses his legs when he's sitting down being very intellectual. You know, yes, he's one of those academics, but he is absolutely wrong. And find the people who prove him wrong. I prove him wrong on a daily basis. You know, and it's not because I have a thing against him. I have a thing against what they're trying to do by setting us backwards by almost a generation because they're they're they're nasty. If I'm meeting with a senator, I'm going to bring a few software companies with me. Yeah. One, one I saw at HX0, which is San Francisco's best startup presidency. And it was a company that writes bills with AI. And it can write bills, it can tear apart bills, too. So, yeah, you don't want to know how you feel about, um, here we take long. And there's a barbecue speech about, you know, he acts like how could anyone go to a barbecue and let one person take 85 to 90 percent of the food that everyone else. How do you guys feel about that? I think he's right. But it, okay, if we're at the salt, look, it doesn't matter. It's all you can eat. Yeah, no, don't take all the barbecue. No, but no, there's a lot. I mean, that's where you get into the finer points of what abundance is going to look like. And what the interact number is going to look like and who leads us through it. I mean, the bottom line is pick a future date. Will you on a robot? Okay, great. Pick a pick a further date out. Will it be able to make or it's on robot? Great. Pick a further date out. Oh, will you line with it? Do you have the role to join? Did he answer? Yeah, he said probably not. Maybe it may be a cigar. He told me the next. Maybe it's a cigar. Well, one man's a guard. The other man's a joint. I don't know. I've been around a few of those in a music business. I've done it. Yeah, big like a baseball bat. You don't, you don't want to. I mean, I need some of these fans with their hop-boxing. You know, it's just, uh, it's crazy. No, listen, you can be pessimistic about this or be optimistic. But what you can't let happen to you is buying into the bullshit. Yes, data centers do bad things, but they do good things. Okay, yes, AI stole things, but there's ways to mitigate that. How will it mitigate? Well, if AI stole everything, but ultimately gives us abundance. We're none of us need to do the toil that we did. Is out of good bargain. Yeah, but how, somebody needs to articulate that. Somebody needs to show the path. And you have to do that in the Interregnum in a government format. You need to stand up there and say, hey, AI stole everybody stuff. Say it openly. Oh, but AI is going to do this. Oh, and we're going to make sure it's going to do that. Because we're going to have laws to make sure that it does that. See, that's the way to do it. Is it because just attribution, right? Whenever this topic comes up, especially like an academic spircles, comes back to attribution. And the fact that the models can attribute, but they don't, right? So the thing is, is like, how hard would that be? Like, if I, as a writer, am doing art, and I'm pouring from a specific art style, can it say that this references these artists styles and can that person get a shout out or a credit? I like these are the topics that I think that's a title. I think that's a brilliant and vital solution. And it can be done. And see, when we're talking about laws and rules, again, I think ultimately those laws and rules will dissipate, but they're cast that are on a broken system between down then. Hello? Oh, that's weird. Oh, no, Alex. He's very connected. You know, we're actually doing that. I'm actually using solving that problem with a lot of things. But we're like a super small company. And I feel like there's a billion dollar companies that are not even thinking about this. It has to do. And I think that's an interesting thing. But it has to do with that. No, man. So very quickly, I'm so sure, a follow-up question. I'd like to talk blockchain with you. I'm a big blockchain dude. Oh, I got to know that. That's where I probably will go. No, yeah. Hi, Brian. I have a question. Yeah. Brian, basically. If you go find what is the, what would you say is the one thing that, you think diaries making a problem for? A diario? All right, the biggest, the biggest problem of diario is he's creating an environment that AI is inherently dangerous to humanity. And it needs to be only in a hands of the few. And it needs to be around a committee of trusted people that he's pet. Hey, Brian, get in here. Yes, Robert. Can you hear me? Yes. I can hear you guys. OK. Awesome. Brian. And I guess, Alex is having an issue. Can you guys see me? Yeah. Yeah, we come. Yes, I can hear you. I was going to follow up with David. Oh, Brian, hi. I was going to describe because we have no conversation, I said, it's a wonderful space, right? What we're doing is that same thing that David was doing. We're on a space right now, a delibration and discursing the homes that AI potentially can cause and they could part on the bad path. He's just doing the exact same thing. So who's holding on the countable? Right? We were discussing under mitigations and how it's possible to have a good side of the application of the version of the tool or whatever I want to describe it. And there's also a bad part of that side, which is we also have to acknowledge in that. That's exactly what we're doing. That is just the only one burden of to come on camera and say, Samsung's not going to do it. The Chinese don't care about it because they want your data. And we all are acting cool. Like it's some weird thing. He's pulling off like a stunt. It's not a weird stuff. Back on that, because I don't think that it's good faith, right? Because sometimes you want to create something. And then you want to prove that it's broken or do something. So that all these regulations take it out of the public and put it into the private. And so you say, see all the things it can do. You can't trust the public with this. But it's not done. An example of that. Can you give us an example of one? Just one use case where that happened. Anthropic hacking. Breaking out of his virtual machine that they call it a sandbox. Man, so we've been talking about it. So we already are murdering tactics. We just because we'll see what I'll know that a virtual box for the sandbox and that's the same thing. Man, so we already discussed this on a separate space. And now you're just putting my legs in and know what you're doing. Have you seen the meme where they're like the sandbox? And this is a guy just exploring the world. You know, have you seen that? That's like, yeah, yeah, that's that. It's a fascinating immersion behavior as we would say in San Francisco. And Robert, I sent you a follow. You're pretty good. I am really enjoying this whole talk here. It's been amazing. Thank you, Alex. We're getting everybody together. Let me use this opportunity to say a robot. I always say this because it's so great. You are curating the best list on X. If you go to Robert's profile, I think there are, I'm not sure where it is. Three dots or something and you can find his list. And he has curated hundreds of accounts. Like everything in tech AI, if you want to learn about this stuff, follow me. More like 90,000 accounts. In 60,000. I hope the audio quality is good. By the way, I lost the space and had to pick it up on the desktop. But yeah, that's cool. Me? No, I work hard on my list. I've been working on for 19 years and they improve every day. And I have for years. I use them all the time, man. They're really good. Thank you. That's resource on X. Yeah, I watch them all day long, too. That's why they keep getting better. Like, uh, somebody starts talking about politics. I always check them out. Like, what the fuck you up to? How about your AI, your AI news article, the one you mentioned? Yeah, a lot. Yeah, a line news reads a line news.com reads 17 or 18 of the lists, all the AI lists. And analyzes a man writes up a news page. But now I have a million posts in my in my context door. So I can, I have a good, search. If I, if I want to search for anything in the AI community, it's pretty cool. Just checking a blind if you get you back probably not right. Now, I think he was, he's going to have to jump out and jump back in. I don't know. And he was talking about leaving anyway. So maybe he's a chance to grab a jump out. He was just a chance to find you get along. Yeah. Robert, what's your take on that? Like, what do you think industry can do to kind of steady this ship during this interim period? Ship and products. But I, I don't know, they're playing games above my pay grade. I like regular Tory capture kind of guys or that's games I don't play. I don't, I don't all do even look at it. So you think the security stuff is posturing? Oh no. I, I, I know a startup that is a security startup. It's pretty true that they can do some free crazy shit. Um, the trick is to get it out there. Some programmers fix the holes. But you're, and I can't fear in our race against the bad people. Like, I totally, I heard the other day that the average time to patch a zero day in the wild is 55 days. Like. if you have, you know, 100, zero days a day, giving created just on the fly. That's a lot of gap time between a lot of those closures that are patching. I have a list of all the security professionals here on X. I would hook up my AI to that and watch for news security problems and make sure that it fixes any holes that we're talking about, right? Ooh, depend of what. I've taken your Windows box. Depend about the pen to buy, turn it on and you're good. Yeah, it gets better. I did hear that Chrome pushed like more bug fixes than any other month in July this year, just based on five coding alone. So I think Defence is catching up to. Yeah, they have the tools to see a lot of holes. And close on. It's an interesting thing. It's interesting that we're still talking about security in 2026 because I remember interviewing the guy built Microsoft security strategy and read a big fucking three inch thick book trying to get Microsoft to write more security code, you know. And now that's sort of falling. Yeah, because it was a thing like, yeah. Microsoft stuff has never been that secure. We just didn't care that much. Now we have to. We have to. We care. It's just hard to see all the bugs that you're ready. Well, I had a question for you. I know I'm pretty sure both you and Robert you and Brian saw my post about AI being able to generate exact binary code. I was wondering, what do I do with that? Like, do I bring it to a tech company? Do I wait to release it? How much faster would it be if you had a hundred million dollars? It would be so much faster. How much faster than any than the way it works today? Um, well, I can generate a 10 megabyte file in 0.36 seconds. You got to put an English for the people who are buying. Yeah, how much faster is it than a traditional large language model? I can, uh, well, each prompt for a binary code is about five minutes. That's with chatGVT 5.6 sold and you can iterate on the binary code and debug it as well. Is it that it should be faster, isn't it? Yeah, that's way faster. How much? That's what you need to answer. You don't need to. About 2000 or something. Like, a couple orders and maybe two kind of things. I would. Yeah. So it's compiling it every time or how does that work? There is no compilers. It's just straight binary. Hey, I just. I'll compile it. I'll compile it. I'll compile it. Paying and puts it on your draft. It's bits out, uh, an ELF file. That's pretty low level dude. I like it. So right now, I've got the binary only OS that I was making. And I'm doing something now that's, uh, I'm trying to figure out how to create a binary LLM and it's working right now. It's just, um, I don't, it doesn't speak English yet. So it just gives me, it just gives me hex codes as a response. It's like, yes, these, this agrees with this hex code. But I don't really know it's pretty good with agents or marketing. So how does he turn his idea into that $100 million investment that you were talking about? I think that's, uh, I think that's where we go. Yeah. What is the practical use case? Like, is it a way to commercialize it? You can think of. Um, security, uh, software generation, uh, cross, uh, polygot, software, where the binaries, uh, the program can be launched on multiple different platforms without, like, wine, uh, custom OS software. That's great, Ben. I think you just pick one and run with it. Honestly, don't get diluted with a bunch of them. Get one, get a really good, go to market plan around it, get some customers and, let's chat. Like, I can help with boardy. A Robert is probably better. Help with boardy for sure. So it probably would be a good couple of questions. What do I know? I'm going, why are you? I can imagine the earth shed. Yeah, I just don't really know what to do with it. Well, you, you have enough of, like, what I was asking, how much faster it is to understand, are you talking about something like 10% better or something that's really, really dramatic, they better. Um, that's, that's a piece of one of the checkboxes on the spreadsheet and things that was, are gone your way. Also, I did a compression test and it's about, uh, I can compress string of bytes, um, about 99%. And then I also did because somebody requested it. I also made a binary that was pre-compressed. So it started out as a compressed and then I uncompressed it into an image. All this kind of thinking makes sense in the star like, we're all because getting packets up to space takes a little bit of time. And so the fewer things you need to send up to space, the better your latency of your video, you in or whatever. Look, do you currently have a, uh, tech job, like a full-time tech job? No, I'm actually unemployed or it sounds like it should be a plan. It sounds like it should be a plan for, like, AI engineer type roles. I did. It sounds like you could easily work at, like, how you can bring your harness to a hard skill on and within that shit and pay me your rent. What about for that? I'm doing a hackathon right now. One of Google's, uh, but I don't know. I just, like, I want to work at like spacex or open AI or something, but I've already applied there and got denied. So he's like a little pain in your life. But I don't like paying. I just deal with it. Working free there those two guys probably wouldn't be always fun. It's not full of faint of heart. No, I don't know. I'll probably just watch. I'll just get them fucking 600k 700k sounds. That's good. That's good. I think that doesn't even buy you a house here. You got to get your stock options have to do. Really well to buy a house here and have 600k. That's not our year's salary. Bro, remote dude. I want to be a fucking digital. You know, if I worked for one of these companies, I'd be like, I'm fully remote and you can't tell me where the fucking, but I'm going to travel to the fucking. I did night, it was 95 startups launched in one fucking warehouse. So either here, or you don't see it. I don't know. I do much. The start of the scene is, it's $1. I'm going to get down and I can't even have that sometime. Then they do like, I think it would be the. As somebody who started to start up in the Midwest, like, it's real. Like, there's, you know, the coasts kind of have the capital in San Francisco, kind of lost its vibe for a minute. But my army tried to just wing it, but I think everybody realizes that with AI being centralized and SF, like that's just where this shit happens faster than anywhere else. Yeah. Seattle is holding its own kind of sword. You know, we got our own little niche up here, climate tech dude. If you're in a climate tech startup, Seattle's the place to be. But yeah, yeah, if you're in a few different countries in this street, Seattle's where the whole fucking industry is at near Redmond Washington. Yep. I think the only thing that is going to kill San Francisco is like if Gary Tandas and help the politics get their shit together. Like, it's, it's truly just eating itself right now. Other than that, like, it's, it's a beautiful vibe. It's why it's actually, it's incredible. It's fucking crowded. I wish you all the best. Seattle is eating itself politically worse. Let me, let me, let me save SF a little bit. I think Seattle is eating itself politically worse than us. I've probably didn't you see that shit? Our mayor's a fucking more. Well, it's, it's not the same. So how come not you look at that? I don't know, I still live in Seattle too. I have a lot of, how come 90% of the startups feel? Because that's what they do. Because that's not always good, but what you got it? Yeah, that's fucking tough. Dude, it's hard to start. It's really fast. I'm fast. Now you're trying to change, I mean, a startup is really asking you to change in order to buy their product, right? And is, otherwise you will not be surprised. I went up to a high, right? You wouldn't change. And that's what. Well, and even just, I went up to Vancouver BC for web summit. There was like over a thousand people there. There was a shit ton of startups. They all had that little booths and whatever. So I'm walking around like 75% of the AI companies quote unquote, we're just like, Claude rappers. You know, they're like, where a AI B to B says company that solves the problem for blah, blah, blah, blah. And it's like, okay, well, what is it? And you like, take one peek at their website, you like try and use some of their stuff, whatever, and it's 100% just like Claude can fucking do this because it's a Claude rap. Like it's nothing new. You know, no major investor is going to invest in that. They're not going to get their money back. You know, it's like the two risky, right? So I think that's why a lot of startups fail. Because they just don't get enough funding. They don't get enough revenue. They don't get any customers and then time runs out. And they're like, okay, I do this forever, right? It's business, right? So you got to fucking make money. Yeah. So just quickly, I think it's just like what Elon says. It's it's staring into the abyss and chewing on glass, you know, like, I have better founder enough times. It's fucking hard. You know what I'm thinking that Elon, where are you going to do so many different ways to fail? Is it in California? Yeah, it's in San Francisco? Texas. Oh, oh, oh, oh, oh, it's in the office. I got you. He wanted to wet it. Mike has off just the poor like $5 billion. San, San, it's home. Was it San, it's home? Oh, Texas. Interesting. Yeah, just quickly, how's keeping here? I'm the governor gave me the sales pitch years ago. So I could give it back to you if you want. Let me just say, I'm going to try to reconnect my phone. If it doesn't work, I'm just going to say we had a good run. But I'm going to try. All right. So hope you come back. Robert, what's the coolest thing about things? We'll see in this past week. Um, a little worm that would go into a 3D printer and take out the completed thing and put it on the table and then close the door and push go and do it again. That's actually turns the 3D printing into a possibility of making mass made back, right? Yeah. Get away from helping the things. I'll work in themselves. How big was it? Just a little one. Just a little one. I took a video of it and put it up the last Friday at Founders Inc. There's a bunch of startups there. Oh, yeah. Yeah, there's a whole brochure. Like, shall you learn? Cloud native, agente, dev environment for engineering teams. I don't think it had enough buzzwords. Yes, correct. Predotype with AI at conversations speed. I have no idea what that is. That's what it just goes on. They have 95 startups. And go around and see him. That's awesome. Was it a competition? Yeah, it's an incubator. It's like a white combinator. This guy got three big warehouses for us. startups and funds, startups and incubates and helps them get that air. That's awesome. Are you an investor? What are you thinking about? Come on. Who? Go for a fits. No, you can wrap up your question too. No, no, I'm good, bro. Go for it. All right, I was just going to say what's your take on like quantum stuff? Is it exciting for you? Is it new? Is it old? It is. You don't really care that much. Some days, some years, going to have a real breakthrough for competing speed and be able to let us do things that we can't do today. So yeah, it's exciting for that. It's it's hard to approach because, you know, a machine that's sitting in a big warehouse under control somewhere. Yeah, it's kind of like for me, it's like a full circle sort of thing where like, I mean, I didn't really like use a mainframe, but some of my professors did had this one fucking guy, professor Clyde Ruby. He was fucking like 78 years old. He was like, this is the old school data algorithm's course. And he was like, yeah, back with my day, we used to use punch cards and be free, which is just so funny, like here it is old story of like older computers, right? But like using a quantum computer and like renting it, you know, and like getting time on it kind of thing because it's only in one kind of location. You got to access it remotely. And so like, you got to start scheduling time for it and stuff. And so me, it's crazy because it's like kind of like back in the mainframe days, it seems like, you know, where you have to like rent time on a computer, you know? I don't know that kind of stuff. Why isn't it's actually around the computer history museum and talked about stealing time at two in the morning machines because nobody was working at two in the morning. What was was like, is he true? He's one of my favorite people. He's got some amazing. Real smart, real nice. They have some crush like that, a small, just some TV and stuff. Yeah, but I've walked in a store as well and I dinner with them and he'd be mobbed and he was nice to everybody and carrying around a book at $2 bills that even tear off and sign for you. Who gonna be the apple of the AI race? Right? We see the IBMs. You see the Intel's, but who's gonna be the apple? The start-up that's kind of, you know, I chose the culture for the next 50 years. I trust my life to a Tesla robot. I have not gotten that way. Well, that's not true. I've also trusted my life to a way more robot and a Zooks robot. So those three have been advantage and selling me stuff that has to deal with robotics. I don't know. You know, I never read at figures like I ain't gonna worry how that way. It's gonna worry how it's another way. I still think Elon is the dramatic winner, but he does have to execute. What happens when better models start to offer diminishing returns and it's kind of like we're probably saying, so having, and also when you can fit a model on your phone, a front to your model, your phone. There's still so much need for that database. There's some, what happens to that money that's invested in those antiquated systems? There's so much to do even if the LOM tapped out, I'm hookin' them up and doing things. I mean, someday we're gonna have a real time information system about the real world. Like you'll know within seconds that a house is on fire, right? Also, on what phone? It's definitely not gonna be Apple. Oh, I enjoyed. I thought I was thinking of those companies that are gonna still be around. I think that's like bad. Yeah. And the wrapping is like saying, I learned that Microsoft can make it. Well, it's gonna still be here. You know, like there's the LLs of the AI. I learned that Microsoft, you can make mistakes. A lot of mistakes and still survive when you have such a powerful revenue generator behind you as Microsoft does. Open AI is intriguing to me. Whether you like Sam or not, it has a lot of users and a lot of users. It's what's gonna matter. It's hard to bet against infrastructure. On the subject of the phone, I will raise you all the idea that the phone company and the processing company of the future doesn't have a name right now because you can build a clean room in your basement with AI accelerated material science and do the garage build stuff and be like, Ryan and do the crystal hacking and everything and actually build your own hardware. Wherever's gonna come out on top, it's not gonna be somebody that's using old substrate methods of pop or traces in silicon board. It's gonna be somebody that solves crystalography. Like, I'm trying to do it. Like, I'm literally working on my clean room design right now, synthesizing from 357 files and notebook LM, most of them over a million tokens in length. Going over like the finalized CNC designs for everything. So I can actually just feed these files in and have the system automate putting out parts to send to people at machine shops so I can just start putting out crystal ingots and testing how they do compute. Like, that's what the future is gonna look like. I might be, I'm gonna not be the guy who succeeds, but somebody will. Sam, they will. I'm betting on that guy. Just because he may be like a... Yeah, he'd meet too. He'd not feel like I've run through a fucking wall for this dude. Let's go, dude. Build some fucking folks. Hell yeah, dude. For four months ago. Just remember the first one gets shot according to Steve Jobs. Right. I think you can move it to go. I thought, like, Blackberry dude, like, Blackberry. That's what I bought for it. You know, four months ago, I thought that Apple computers, I'm max. We're the greatest developer computers. I thought that's, oh my god, you know, Apple was a greatest thing since sliced bread. And, you know, even Windows. But now, I'm all in Linux. Like, I can't believe it. I'm not gonna... You're gonna... I'm like, I'm holding that stuff. Like, these AI's are such accelerators. I think it's gonna transform in a factory here. And, just since it's gonna make people more productive. Just wait until it goes all bare metal and skip operating systems. It was gotta get the first Mark's pull rat out. General magic. You know, you gotta sell spectacularly. And then we get our nice cool iPad or new whatever you want to call it. Oh, yeah, you know, I just gotta get you... Get you out. The problem is that LLM's, I would have a little bit of latency. It's hard to make an OS out of them. That is cool. It's getting close. I got an Android phone here that is running a new kind of OS on top of it. You just talk to it and it shows you shit and tells you what your supposed to be doing. Then if you want to play game, you just talk to it. Hey, I want to play whatever. Angry birds. Well, I mean, I'm working on an operating system that reduces latency by quite of it. That's awesome. Oh, yeah, I was doing some tricks into what Shannon, I'm touring in Vaughn, Neiman actually gave us. According to LLM's and like, how much stuff is in there that is just arbitrary abstractions that don't line up with our current nature of space time. And out of the simplest arbitrary abstractions that like blew my mind when I was put forward and I didn't even think and realize this is that touring computers are like one D ticker tapes and all our RAM is still designed to work like touring computers. And we obviously live in a 3D reality. Just one of the many arbitrary design decisions the genius has gave us that nobody has since gone back and improved or updated on. Because it works, but it could certainly work better. It can always work better, right? But this does infrastructure, this just how things go. Now you know why our app on this and learning which models are not about not it. It's actually about specialized models, you know, small language models, specialized language models. That's actually the future. Because I mean, if you have a specialized problem, you don't go to a general practitioner, you go to a specialist. So, you know, a lot of that bloke language bloke is actually waste. And there are some companies like I think it was thinking machine, cortex research group. Those people out here that are actually solving, you know, what I would call the real problems of giving people what they want. And that's the Lucian state of problems. That's the solution to abstract poetry or soccer teams. You know, they want to solve a real problem. And we don't need to know 10% of foreshad and to solve those problems. So, I think that's something that's going to be. It's already starting, but I think it's going to be in at the forefront maybe next year. The large language model start really saying that they're going to be a disadvantage. Well, see how I'm going to get to that. I got to head out. Um, appreciate everybody. Robert Alex. Alex, this is a great space. Yeah. Yeah. Follow back if you can, but I just can't just do something. So, you know, I'm just a little developer doing. little hackathons. So, yeah, everybody have a go when they're thanks. Yeah, still man, great to have you here. I also think like we we should slowly come to a stop here. I've been way too long awake. And gotta make some executives decisions here, right? I mean, quickly, thank you Robert here for being here and talking to us the entire night. Yeah, I was fine. It was definitely fun. Let's do this again. And yeah, I wanted to ask you in particular here anything coming up uh next couple of days, week um and he's space in this podcast. Yeah, I mean, just going to Sam Sysgrad, a high velocity, trying to see what the future is and maybe make some videos and stuff. So yeah, going again tomorrow. Going forward to that. Yeah. Yep, really. So, I happen to in place. I got to visit one day. One day we're going to make it happen. Uh, you and I are going to go to one of those robots five. So, this is just my dream personally. It's like going to the fun times. See even more fun if you can get inside the robot. You know, it's tell operate the robot and yeah, if I could take the other robot, then we can fight each other. Oh yeah, that's fun. It's going to happen. So, a lot of fun. Particularly if I get to knock you out, but it's fun. I'm going to see about that. I'm going to see what I'm not going to do. I'm not going to go out and robot and run a 2000 people as awesome. What happens if you're damaging it? Do you get a picture? No, it's there, robot. You're just tell operator. It's in a cage. So, you know, it's not going to go. They overheat and stop working anyway. So, they got maintenance crew. What else do you ensure is that a couple of stuff? Yeah, people probably people are getting re-loaded. They're computer nerds. I have run about has three nerds around another time. Right, guys, I think, you should end when it is best. This was really nice. Thank you all for being here. This was really nice, really enjoyed it. Yeah, what can I say? You know, for the people who are more moved and all about energy, energy, and all of that stuff. If you're interested in that, follow me, maybe, follow everybody you see here on stage, and also in the listeners, right? Chances are you all are interested in the same kind of topics, right? Also, look around, make some new friends. And with that, I would bring it to close. This has been Alexi Topia. signing off for tonight.