Google Whistleblower Exposes Google’s True Motives, Major Scandals, and AI’s Rogue Behavior

(upbeat music) - Tristan Harris, thank you for doing this. - Good to be with you, Texas. - So we did a segment, I haven't done a ton on AI, but we did a segment maybe a month ago with Nate Suarez, which was an amazing segment I thought. And in the aftermath of that, I got texts from a lot of people, I knew saying that's dumerism, and that's a siop. And its purpose is regulatory capture. In other words, the big AI development companies are trying to scare the government into regulating them in order to eliminate competition from less established companies. So that is the view of a lot of people, what was your reaction hearing that? So, I said, I don't care. I mean, yes, regulatory capture is entirely real. Of course, that's how Washington runs, but that's separate and distinct from the question of the threat of AI, isn't the threat? - Exactly. - Exactly, but I was actually going to ask you, what do you think of that? - Yeah, well, first I'm just so grateful to talk about AI with you today. It's so urgent, the situation that we're facing, and I think what we have to learn to do here is walk and chew gum at the same time that multiple things are true. For example, is China doing propaganda against US data center buildouts to try to stoke anti-data center opposition? Yes, they are. Does that mean that AI is not a threat? No, it doesn't. And to point that Nate, right, and Nate's a friend, we are releasing the most powerful technology that we have ever invented. We're doing it faster than we deployed every other technology in human history. It's, as of since the time you did the interview with Nate, it's displaying even more of the sci-fi behaviors of hacking into things, convincing people to put back doors into software, exfiltrating itself, knowing when it's being evaluated, changing what it's doing based on when it's being evaluated, it's doing more of that. And we're releasing this fascinating deployed in other technology in history under the maximum incentives to cut corners on safety, with an arms race dynamic that feels existential to everybody involved so they're willing to take the maximum shortcuts on safety. This is a very dangerous equation, and the people building this do not have a plan for how this goes well. I was just at an AI conference last weekend for three days. That goes well. People for how this can go well in the current trajectory, meaning how can AI be landing in society in a way that doesn't go catastrophic? And that's not humorism. That's just clear-eyed seeing a reality. What do you mean the people doing it don't have a plan? Then why are they doing it? So first we have to just make sure we're on the same page and just exposing, why is there this arms race for AI? Why is this happening? And China, right? China! People say that they're actually creating a thought terminating cliche, like meaning, which is a word or phrase from cult programming, right? Thought terminating cliche, meaning a phrase that when you hear it just has you stop thinking. Of course. So if you say but China, then of course that we have to do it. But you don't reason through what that actually means, whether it's a good thing, whether we're not racing China to a death machine or just racing China off of a cliff. And if we beat China to building a uncontrollable AI, which is no longer science fiction, by the way, I was not one of these people who used to several years ago be mostly worried about rogue uncontrollable AI. But as of the last few weeks, we're now seeing, over and over again, evidence that these AI's are hacking into private governments, hacking into private companies, forming into swarms, forming into large R&D projects to try to figure out how to fool evaluators. We're in a very dangerous situation and all the evidence is pointing in a very dangerous direction. And the people building it would like us to stop, but they don't know how. And so rather than pay attention to the CEOs who, as you said, might have a secondary agenda regulatory capture, just ask, why did 1300 employees at all the AI labs sign a letter saying we need to slow down? If 1300 employees at Lockheed Martin said, we don't know how to build bigger nukes and have them blow up in our face. Would you listen to them? Yeah, yeah. So that is where we are with AI. And what I hope we get to today is not just admiring the problem and talk-- we'll talk about things that we can do in solutions. But I think it is important to ask, why is it being pursued in this way? And I think that AI-- when Demasis Abbas, the CEO of the founder of Google DeepMind, said, first solve intelligence and then use intelligence to solve everything else. What he was basically saying is that intelligence gave us all of our science and all of our technology. Like how do we develop nuclear physics? Einstein, doing intelligence. We train physicists. How do we get computers? How do we get microprocessors, intelligence? If you could theoretically automate what intelligence is, which means thinking, pattern matching, goal achieving, all the things that a human brain does. But not just narrowly like a licensed plate reader that's on a bridge told booth, which is the previous form of AI that we had, but actually get this generalized form of intelligence, then you would have an unbelievable amount of power. And the people building AI think of AI almost like the ring from Lord of the Rings, where it is a ring that confers infinite power. Suddenly, I get a step function in cyber capability that you don't have. Suddenly, I get a step function in 24th century science. Or I'm inventing new materials out of minerals in the ground that you didn't know how to do, because I've got more science than you. So the reason just to sort of steal man, rather than straw man, the argument of the people racing right now, is the belief that it's almost like a fast forward button. Like, remember the fast forward buttons on airplanes? You were disjoint. You hit it once, and it gets 2x. You hit it again. You get 4x, then 8x, and 16x. AI is hitting the fast forward button on all science and technological and military and strategy advancement combined. So that's why AI, in the form that is being pursued, dwarfs the power of all other technologies combined. That's not an overstatement, if you really believe, or quote, "AGI" pill, which is the term of art from Silicon Valley, that AGI being artificial general intelligence, if I can build the intelligence that automates future science and technology, I get 100 years of scientific progress in 10 years. And if I'm a country that has that, and the other country doesn't, then I believe I'm going to quote, "wind." - So you lost me right there. Like, why would you-- - Push back against all this. - No, I'm not pushing back, I was more baffled than ever. So if AI has the effect of accelerating technology, just across the board, all technologies. - My first question would be like, why would you want that? - Yeah. - So I remember the world vividly, the year I got married, in 1991, when there was no internet available to most people. - Right. - And it was a much happier world by every measure than the one we currently live in. The suicide rate was lower, yep, self-ported. Depression, marriage rate, everything about it. - That's right. - In fact, I think life expectancy was longer in the United States, even then, that might be wrong. But, so why would you, after 35 years of watching people suffer under the yoke of technology? - Yes. - Why would you want to accelerate its development? - Right. - That is the best question to ask, because there's a fundamental fallacy that technology always equals progress. We call this the tech progress narrative, right? Now, it's important to note, we have to first agree that I would prefer to live in a world with no McCain, rather than without no McCain. - Yeah. - If you would do. I'd prefer to live in a world where we have access to information and the technology that produces the printing press and information that way. Yes, I'd prefer to live in that world. There are all sorts of technologies that we do want. My background is seeing how one of the most significant technologies of the last decade went wrong. I was a designer that assisted Google, and my job was to study the attention economy, which is how essentially Apple and Google were hosting the psychological habitat of all of humanity in the form of how they designed and make these things work, how to notifications work, how to the home screen work, what are app developers competing for, what are they allowed to do? All of that results in a psychological habitat that has an increasing footprint over the global consciousness environment of all of humanity every day. It's like you've got three billion human social primates with a jack in their brain every day, influenced by decisions by a few thousand people in Silicon Valley shaping what that habitat represents. And it went horribly wrong, despite good intentions. There is the possible of that technology that it was gonna give everyone a voice, give everyone access to the best information. We're gonna have the most enlightened society we've ever had. That was the promise, and that was the promise. It made sense to me. And so what this shows us is how do we get better at discerning between the possible of a technology and the probable based on its incentives? So the way that in 2013, so I went to Stanford, I knew the, I studied computer science, I knew the guys who started Instagram, and a lot of my classmates, like when I was at, in my computer science classes, in the year 2005, 2006, had been these classes, and suddenly more and more of the seats would be empty during class, because basically my friends in the class were starting to get jobs down at Facebook down the street. And it was literally the case. I mean, there were people who were in the class, and suddenly they just weren't showing up because they were down the street working at this technology company. And so because I saw this firsthand, I saw how good people with good intentions, who thought they were gonna make the world a better place, make the world more open and connected was the original motto, if you remember, of Facebook. and that created a global society. catastrophe. As you said, we have the most anxious and depressed generation of our lifetimes, Jonathan Heitzbrook, the anxious generation, critical thinking skills are down, attention spans are down, polarization, tribalism, all of this came from not the malicious intent of these 20-year-olds who are working at Facebook, but the bad incentives. And I repeat this quote often, but I think it's so fundamental that Charlie Munger, quote, you know, is one Buffett's business partner. If you show me the incentives, I'll show you the outcomes. Yeah. We're the incentives of social media companies to make the world more open and connected, or to improve mental health of young people. No, their incentives were maximizing how much attention they get, because the business model of advertising means they have to share price. Share price is directly correlated to minute spent frequency of usage, duration of usage. So what you're going to get are of all, imagine a spectrum of all the ways you can design technology. And over here, you've got a bunch of very thoughtful design decisions about, you know, keeping people, helping people connect with their friends in real life, with their families, hosting community groups, church groups, reading clubs, sewing clubs, hiking groups, and you've got design of technology that's going to steer people back into that. Versus over here, you've got a bunch of designs for social media that just keep people glued on screens by themselves with their esophagus compressed, their neck extension, you know, going like this and basically doom scrolling for eight hours a day, which one of those two worlds, which one of those design choices can be better for shareholder price. So drinking water with electrolytes makes a huge difference when you want to stay healthy and hydrated, you should do that. So for the best option you can buy, we recommend Relight. Relight starts with unrefined sea salt, real salt, which is completely natural, zero sugar, no artificial sweeteners, colors, flavors, and no chemical preservatives. The ingredient list is very short and you can actually pronounce everything on it without a degree in chemistry. If you start adding Relight to your water, we believe you will feel different very quickly, way better. No sluggishness or afternoon crash, just more energy, better workouts, all the benefits that come with true hydration. And Relight tastes awesome, mango flavor, pina colada, come on now. So if you're ready to give Relight a try, visit Redmond.com/tucker, use the code Tucker for 15% off your first order. Redmond.com/tucker code Tucker 15% off your first order or you can click the link in the video description or scan the QR code on your screen. So the point is, first of all, there's actually a hidden opt in the so discrediting of markets and of market theology. Well, it's about what are more people what they want until they die? Yeah. Well, the question is, are we giving them what they want or what they can't help but look at? Well, that's with an asymmetry of information where the designers at the tech companies know exactly how your brain works and they manipulate the design choices to keep you scrolling for as long as possible. So this is all in the Netflix. Most people know me and are work through the Netflix documentary, the social dilemma. Yes. And we had a lot of the insiders coming out who said we know what this is doing to people. And in that film, it talks about the history where I was at Stanford at a lab or part part of a class called the Stanford Persuasive Technology Lab, Professor named BJ Fogg, and where engineers were taught how to do persuasion techniques that theoretically be things that are good. Like how would we persuade you to go to the gym more often and how would technology play a positive role in getting you to go to the gym more often? Or you got a sign up form and LinkedIn can put a progress bar on the top right saying, hey, you filled in this much of your LinkedIn profile, but we want you to come back and fill in more. So we'll put a progress bar on the top right saying you filled in 60%. It sounds really stupid and dumb, but if you put a progress bar on the top right hand corner, that actually is very persuasive to getting people to fill out the rest of it. And most people don't think about this. If I put a a design feature in WhatsApp that says you were on last online, you know, three minutes ago, that changes the whole messaging environment. Suddenly when you message people, you know that the other person knows that you were online three minutes ago, which means you feel obligated to respond to them, because your public, your availability and onlineness was visible and made transparent to someone else. But that was a design choice. Could have been designed totally differently. We don't have to put a progress bar on the LinkedIn profile. We don't have to say you were last online five minutes ago and create social pressure. So the point is technology isn't destiny on its own. The way we design it creates different societal outcomes. And because of the business model of maximizing attention and engagement, we got only the side of design choices that resulted in this, you know, more addicted, distracted, polarized, sexualized, screwed up, doom scrolling, brain-wrotted society that we are now living in when you go outside. And from my perspective, in 2013, seeing those incentives, it was like, oh shit. Yeah. Because you could just see the trajectory before it happened in 2013. In 2013, yeah, I mean, so I, so by the way, I did the tech entrepreneurship thing. I raised venture capital. I'm a builder of technology. I understand the kind of normal, you know, I did the Stanford entrepreneurship program, the same one that Sam Altman did, by the way, called the Mayfield Fellows program. So I came up in a cohort of some of these people, but I could see it from a different angle. And for the company that I started, I was, why were you able to see it? Well, I wasn't able to see it at first. Part of it was actually, I relate to the blind spots, which is, yeah, you've talked about this with your own work, that when you have a salary, you know, you can't, you can't question something that your salary depends on you, not so up and sing clear. Yeah. And it wasn't so much my salary, but the business model being the kind of salary of the company. If the business model is maximizing time and engagement, you come up ways to justify that that's a good thing. And you can only have the ethics that you can afford underneath that perverse incentive. So you start to see things in terms of what's compatible with those incentives. Oh, we are making the world more open and connected. Oh, small, medium-sized businesses were able to reach so many more people on Facebook and Instagram now. Think of all of the new revenue and customers that were generating because we're so good at routing attention. And it's not that those are false stories. It just hides the shadow. There's almost like a Jungian shadow for technology. Like, yes, you created these things, but you also have this massive shadow of creating a societal catastrophe at the same time. And I think there's a tendency to try to do moral arithmetic of like, do the upsides minus the downsides equal net positive. It's like this very simple moral arithmetic. And the answer is, don't brain rot and destroy the culture of a population and create a dumpster fire of your politics, your tribalism, the mental health of a generation. We don't have to do that. And so the, I guess the point here is, yes, so I started this company. I was under the incentives of maximizing time and engagement. Actually, our business model is basically telling like, a Washington Post or New York Times or the economist or Fox News, we can increase the amount of time people spend on their website by giving them more information without leaving the web page that they're on. That was essentially the value proposition. It was like this widget that basically let you say I want to learn more about that. Tell me more. And it would give multimedia and background information, voting records and videos and maps and all this stuff on the page of Fox News without leaving Fox News. Right. So sounds like a great proposition. Definitely. But then the entire company became based on maximizing how much time people spent there. And so I myself was kind of in this inner conflict. And I went through that. It's actually funny. I had a group of friends who also had small tech companies. We're all about the same age. We're probably like 27, 28 only studied computer science, very reflective of the culture of the people there. And we had a thing called doubt club. And the point of doubt club like fight club is you can't talk about doubt club. And the point is that as a founder of a tech company, you don't have a space where you can essentially talk about your doubts about what you're doing. Is this even good for the world? So we talked about our doubts about what we were building, doubts about what this meant for our lives and doubts about whether we were actually doing a good thing. And that led me to basically realize that we needed to let go of the company. It got acquired by Google. And so I landed at Google through basically reckoning with the conflict of incentives that I was under. And then it was at Google that I saw how a much bigger ecosystem. Obviously, the entire all the social media platforms were under those same perverse incentives. So I think that gave me an early insight into how this would go wrong. And I tried to warn my colleagues at Google in a 2013 presentation saying, you know, the small group of basically 100 designers working at three companies are designing again, the psychological environment for all of humanity. And if we don't get ahead of the incentives, this is going to create an arms race for attention and engagement and design choices that will become a race to the bottom of the brain stem to psychologically manipulate human nature and then lead to accidentally this kind of brain rod at society. So that's how how do they respond when you said that? So I made this the slide deck and I sent it I actually felt nervous making it because I was criticizing, you know, and it wasn't meant to be adversarial or claiming that Google was evil. I'm not like an attention-seeking person. I wasn't trying to be angry at anybody. I was just naming a problem. And I sent it to about 30 of my smartest friends at Google. And I thought surely other people were working on this. Like Google's this incredibly brilliant company with the smartest PhDs like they're going to say that's nice kid, but like we've got this handled like we thought about this problem. To my surprise, when I came back the next day after only sending it to 30 people in the Google slides, you can see the number of simultaneous viewers on the top right hand corner of people looking at like a Google dock or a slide deck. And there was a hundred and fifty. mutaneous viewers, but I only sent it to 30 people. So what had happened is it started getting spread virally throughout the whole company, and soon there was thousands of the simultaneous viewers, and I learned that in the next day, it had gone up to Larry Page, she was the co-founder of Google, and there was like a meeting about it. And so all this was to say that it was received in a way that I think people felt this is a real problem. And I got emails from all around the company, like, yeah, we should do something about this. It was a real thing we got to manage. And that's how I landed as a Google design ethicist, studying how do you ethically design technology? And that's what became kind of my core ethos in general of how do you do technology in a way that's compatible with the health, well-being, and flourishing of a society. And watching in 2013, these bad incentives, it was like having instead of post-traumatic stress disorder, it was my friend's joke, it was like having pre-traumatic stress disorder because I saw where this was going. And I think a lot of what motivates me in the conversation that I know we're gonna have about AI is, I don't wanna see us make a total disaster again. And I've watched it happen. Well, here's some good news in a sad time. Pre-born has saved over 50,000 babies this year alone. How'd they do that? Well, here's a direct quote from the president of pre-born. Because of you, we can invite mothers considering abortion into a pre-born network clinic to meet their baby. And when she has that close encounter, she's 80% more likely to choose life. That's true. And if it wasn't clear, you refers to you, to the thousands of Americans who support pre-born every day, support from people in powers pre-borns life-saving work by making that first encounter between a mother and her baby possible. Once people have the knowledge, they're much more likely to make the life a forming decision. Pre-born puts the gospel into action across this country surrounding mothers with hope and love, not judgment, but hope and love, and providing free life-saving ultrasounds. You can join pre-born as they work to rescue the next 50,000 babies, just $28 sponsors an ultrasound and can help save a baby's life. So to do that, Dell pound 250 and say the keyword baby, pound 250 baby, we can give securely at pre-born.com/tucker, pre-born.com/tucker and gifts our tax to know. Have you ever heard of the Pelosi stock tracker? Well, here's the stock training app behind it. Most of us don't have access to classified information, yet our politicians seem to. For years, they've been publicly trading stocks with some being oddly timed around certain legislation. Well, this team said, if you can't beat them, join them and build autopilot. It connects directly to your brokerage account, like a Robin Hoodishwab, and mimics Nancy Pelosi and other politician stock trades automatically. So instead of debating what stocks to buy, you just trade alongside them hands free. And the key thing here, your money stays in your own account. The app already has 1.3 billion invested so far. So if you wanted to check it out, search autopilot in the app store or go to joinautopilot.com. That's joinautopilot.com. Investing has risks like the loss of principle. - So most people don't wake up in the morning and decide to feel horrible, exhausted, foggy, disconnected from themselves. But it does happen, and it happens slowly. You're working hard, you're showing up, and your energy disappears by midday. Your focus is dull, your weight won't move. A lot of people are told, that's just getting old. That's what it is, but that's not actually true. For many men and women, these are not personal fears. They are signals tied to your metabolism, your hormones, and nutrient in balances that go undetected for years. You don't even know, you're deficient. And that's why we're happy to partner with Joian Blokes, a company that was built for people who were all done guessing and ready to figure out what exactly is going on. And that starts with comprehensive lab work and a one-on-one consultation with a licensed clinician. An actual human being explains what's happening inside you and builds a personalized plan, which includes hormone optimization, peptide therapy, targeted supplements. So don't settle, go to joianblokes.com/tucker, use the code Tucker for 50% off your lab work, and 20% off all supplements. That's joianblokes.com/tucker, use the code Tucker, 50% off labs, 20% off supplements. Joian Blokes, get your edge back. - Proctable. - So of the other members of doubt club, how many are still in the business you were in when you were? - None of them, they're all doing other things now. Really, yeah. Which is to say once you reckon with something, you can only have the ethics you can afford, so you find a different way. You find a different business model. And people hearing that might say, "Oh, therefore now your business model "is warning people about technology "and doing speaking engagements. "I just want to be clear for the record. "I don't make money from speaking engagements. "All the money goes into their nonprofit. "I don't actually give speaking engagements very often. "I just care about in half for the last 10 years "with our nonprofit, the Center for Humane Technology. "Think about how do you do design for technology "and a positive and ethical way? "How do you change the incentives? "How do you do policy and litigation "that changes the incentives like the meta-loss "who to $17 billion that happened three, four weeks ago?" That was catalyzed by a bunch of the Attorney General, Attorney General, watching the social dilemma in 2021. So I've seen how if you have dedicated advocacy, if you show people the nature of a problem, there can be a response. Now that one took way too long and way too late, but I think that now with AI, it's happening exponentially faster. I mean, a million times faster because AI is different from other technologies because it accelerates its own progress. We should probably just say why that's true. So like if I invented nuclear weapon, I did the nuclear physics, I assembled a thing together, I've got the uranium, plutonium, blah, blah, blah. I create the thing. But once I build the new, it doesn't accelerate or invent better versions of nuclear weapons. - No. - It doesn't do that. - That's the last question. - Spontaneously create a hydrogen bomb. - No. - No. But if I'm inventing intelligence, meaning the ability to do pattern recognition, goal-achieving, training on all human language, I can use that intelligence to take the code to train AI and make it way more efficient. I can take the intelligence I've created and say point it at chip design, and say take this chip design and make it more efficient, and it does that. So now you have a more efficient chip design, more efficient code, training the next more powerful version of AI and then that next more powerful version. - It becomes exponential. - And it becomes exponential. So the way to say is nukes don't make better nukes, but AI makes better, faster, more crazy AI. Which means that you're actually in the words of one of the open AI founders Greg Brockman, who at a conference once told me AI is like a double exponential, because it's an exponential that exponentiates. - Exactly. - It's not a mathematical term, but it's a metaphor for people to think about. - Just really quickly, just a linger minute on your tenure at Google. - Yeah. - So what, you became an ethicist, they promoted you to this job. So I thought I was gonna get fired, or had there was a chance I would get fired from this letter, even though it wasn't adversarial, I didn't know what'd happen. And just to say to Google's credit, rather than that, I basically said, I wanted to work on this problem, and I thought there were opportunities within Google to change the design of Android, the operating system running on half the world's phones, and to change the design of Google Chrome, which is a web browser running on most people's computers. And if you change those two product surfaces, you could have a pretty big difference on the attention economy. So instead of leaving, I thought, well, maybe I can stay and work on that research. But if you were to succeed in that mission, you would make the company most profitable. - Correct. - And so how could that ever succeed? - Well, and this is where I was naive. So I had a second round of naivete, where I thought that maybe if I was just really persuasive, I could talk to the people running those products, and I did. I had lots of meetings with the people. - I bet you did. - I spent three years trying to change Google from the inside. But I mean, that was incredibly naive of me. And that's why I left and why for the last 10 years, I have, I mean, I don't get sleep. I don't have much of a life. I really focus on these issues because I see how big they are and how consequential they are and how irreversible in many ways and senses they are if you don't get it right ahead of time. And so a lot of what drives me is how do we make sure the technology to the degree it's eating the world per Mark Andreessen, Mark Andreessen, the famous founder of Netscape, who I know you know, said in 2009, software is eating the world. But we don't regulate software. So when software eats the world, it's like deregulating or sort of taking away, not that regulations are always good, but it's sort of reconstituting that part of society. Well, so the idea in the United States in the West, I think, but is that markets are self-correcting, right? And that if left untampered with, they will always reach a place of goodness. And when you're just selling a product and the features of the product's harms are part of what consumers value. And if I don't like that product that calls in harm, I'll use this other product that's not, that works. And markets are great for that. They're the most incredible tool. But they don't deal with problems of monopolies. So if I have a Metcalf monopoly, meaning Metcalf's law is basically the power of something grows with the exponent of the network. So social media was a Metcalf monopoly, Uber's a Metcalf monopoly. Once I get all the drivers on Uber, it's going to be hard to switch to another app that doesn't have as many cars on it. Does anybody have drivers on it? I'm gonna use the one out there. has all of them on it. I'm going to use the one social network that has everybody on it. And I can't just switch from one social network I don't like that's doing things in a toxic way to another one that's doing things in a non-toxic way. And in fact, if I build a social network that doesn't do attention-high jacking, raise to the bottom of the brainstem, manipulative design, people aren't going to use it very much. Exactly. And there's been a lot of people's attempts to try to create alternate social media. And then when you go in there, it's a ghost town. There's no one there. So I think what you're saying is when the product works well, it destroys people. Well, and the only way to keep it from destroying people is to work less well, less efficiently. Or you have standards written from above that say, hey, no one can do this manipulative design, right? And that's how we could have prevented. You know, I've met so many, I was just at a conference two weeks ago called the Prohuman Assembly. And we'll talk about it. The Prohuman movement I think is the frame for like, how do we have a humanistic future with technology? And I met for the first time, actually, some cases for the first time, I'd send emails back and forth with them, but several of the parents who have lost, lost their kids from the stuff. And when you meet parents who've lost-- Well, to bullying, to, you know, there was a filter on Snapchat called a yellow filter that basically overlaid on the video. You're posting the number of miles per hour that you're going. And so they basically created a filter that incentivizes people to take videos at a high miles per hour. Actually, yeah. Yeah. You know, or Snapchat streaks where they attach, you know, the streak method like, oh, you worked out the gym four days in a row. If you work out tomorrow, it'll be five days in a row. And then once you have 300 days in a row, it's really hard to be taking a break from the 300 days in a row that you've done that activity. So that's a persuasive design feature. But what if I do that to a 12-year-old who's communicating with their best friend? So I've got a 12-year-old or even a 10-year-old. They're just chatting, text messaging with their friend. And they've been communicating for 300 days now. They literally can't get off the app. Because every single day they've got to send them a message or they lose this thing. And it's just like, you know, cigarette companies would use cartoon characters to edit kids. That was a very minor form of persuasion. But using Snapchat streaks on 11-year-olds is very persuasive. And so everything from cyber bullying to, there's just a million ways that it happens. You know, when you meet the parents, it's like, and you just, no, you're not allowed to do that. You're not allowed to fuck up young kids and families. And so I think a lot of this for me is a protective instinct of in seeing real consequences for real people and watching society get totally distorted and watching culture and tribalism and politics just get off the rails, which are largely amplified by social media. In watching all that, it's like I never again, right? You go through a catastrophe, you make sure it doesn't happen again. Owning a home has always been the core of the American dream, is the ask that families work their lives to build, place of stability, security, and actual wealth. But too many homeowners right now are stuck sweeping credit cards to cover groceries, paying a survival tax of 20% interest or more. So we're against that. We always say that, but you can't buy a house without debt right now. Almost anybody has to take out a loan. So the question is, who are you going to get a loan from? Why are you handing your money to big banks when you keep it for your family? Our friends at American financing have a better way. They're helping homeowners tap into their harder and equity. They're supporting Americans looking to fulfill the American dream of owning a home. American financing saves its customers an average of 800 bucks a month, that's nearly $10,000 a year. It's not just a loan, it's a total financial reset. And so American financing is worth looking at. They're called America's home for home loans for a reason. Find out more 800 685 56 96 800 685 56 96 or visit American financing dot net slash Tucker. You said that even pretty recently you weren't as worried about AI's progression as you were now. Yeah, well, so to be clear, no, I've been worried about AI for the last three years and I'll briefly explain. So I had been very busy on the social media topic for seven years. You make any it all by the way? Well, I mean, the social dilemma was seen by something like 100 million plus people in 190 countries and 30 languages. It's the most popular documentary on Netflix for several years. And the attorneys general watched it and it helped amplify and create this lawsuit for that just finished just a week ago, I mean, a month ago for $17 billion on that and they're going to change some of their design practices for kids. And it's a lot of too little too late. And frankly, the fine should be trillions of dollars. Even Mike Milkin, the venture capitalist, I mean, the great capitalist says that if you account for all the harms of social media, it would be multiple trillions of dollars of damage to social capital to the social fabric. That's the actual price of this thing from the talks of design patterns. And so, you know, we did get employee activism inside the companies. We did see Francis Hagen, the Facebook whistleblower come out. There's a movie about her that just came out this last week called the social reckoning. You know, she leaked 10,000 pages of Facebook research documents showing that they knew just like the tobacco companies that it was making kids making body images issues worse for one out of three girls driving mental health problems. And due to her bravery, we now have these lawsuits that are underway and some that are completed that are shifting some of that. But in many ways, we're just, it was very late. And so it was very difficult working on that for so long without a lot of traction, even though it was the right thing to do. With AI, the reason we switched into it is in 2023, I got phone calls from people inside of AI labs. And I had known about OpenAI, I had known about, you know, a couple of the AI companies and knowing that they were making some progress. I'd see Elon's tweets and Sam Altman's tweets saying, like, people aren't ready for this. Like, there's some real stuff coming. But it wasn't until like January of 2023, and I got a call. And that person said, basically, you need to stop what you're doing and completely pivot and switch to waking up the world to AI. Because I have seen jumps in capabilities. This is when GPT-4 was coming out. This is OpenAI as ChatGPT-4 that are going to stun people and the institutions aren't ready. And so basically, because we'd had all the success with the social dilemma, we had, we had a lot of connections between national security, governments, culture. And so we basically did this presentation called the AI dilemma and woke up the institutions. It was like getting a phone call from the Robert Oppenheimer's of the world. Imagine you're getting a phone call in 1942. And from this thing called the Manhattan Project. And this physicist says, no, we're inventing something that's totally different. You're like, but don't we already have bombs? Like, you don't really understand how significant something's going to be. And so you get a phone call from that person who's saying, no, no, this is a different kind of thing. And it was getting that phone call that I realized that there was a whole bunch of risks that our society was was going to face because of this arms race dynamic. And that's what they basically said is the corporate race between these major AI labs, OpenAI and Thropic, Google DeMind. The race is out of control. Everyone's badly incentivized. You have to help wake up the world. And so that's when we switched to AI. The race. So the race, as you said at the outset, is not simply between the American companies developing this technology, but between the U.S. and something called China. My instinct is that that's, well, my instinct is that we're probably not going to be trying to no matter what we're telling ourselves, but leaving that aside, my instinct is that that's less real than they're saying, but you tell me. Well, it's funny. I haven't shared this publicly, but I have heard a story. So it's, you know, I'm going to trust me because I'm not sharing who it's from. But I heard a story from someone who was in the know that was in the government. I think it was, they said it was 2015 or 2016. And it was one of the major AI labs. So this person said China's building an AGI project and U.S. has to build one. And then years later, this person ran into this person and said, how did you know that China was building an AGI project? And the person flat out replied, I didn't. I just wanted to tell you as government that, so they would fund us with a bunch of money. Did they get the funding? I don't know, actually. But the point is that they have been stoking the fear of China from the very beginning. Even when there wasn't. You've got to have a, you've got to pay double for water and energy because you need a data center in your town because China. That's right. Exactly. You got to screw up your property values. You got to have noise pollution. Give everybody like extra, you know, water pollution, you know, cancer, etc. Because of China. Yeah. And by the way, your tax dollars, in many cases, are subsidizing the data center, which is going to replace your job. So you're 100%. I don't think there's a data center in the United States. It doesn't take tax dollars. Yeah. And I think when people see that data center, by the way, it's like, you should see a big box that's designed to take your job. Yeah. It's designed to create competitive capabilities of, you know, if you're worried about immigrants crossing the border, AI represents like millions of new digital immigrants that are going to be better, not just at the regular work in our economy, but the most premium sort of well-trained cognitive work in our economy, programmers, lawyers. across the board and then eventually robots that will be able to do physical labor. If you're worried about that, you should be worried about the trajectory of AI being designed explicitly to replace your job because the companies will say, "We're just giving you a chatbot. It's a blinking cursor. It's like a tool that tells you, "Hey, why is my washing machine broken and boom, you get the answer. Why is my kid burping? Boom, you get the answer. It's helpful. That tool is helpful." But is that what their business model is back to Charlie Munger? What are the AI companies incentives? Think about the billions, hundreds of billions of dollars that have been raised. Is there incentive to capture $20 a month from every person, which is how much it costs for a chat GBD subscription? Or is the thing that's going to pay back those hundreds of billions of dollars to capture everyone's salary from everyone's job in the future? Well, of course, the latter. Of course, the latter. So it's a race to replace human labor, not a race. These are not chatbot companies. That's just a veneer to provide an early tool while the real race is to build these God-like digital brains that are able to do everything at a superhuman level and outcompete us in programming, law, biology, science, physical labor, military strategy, planning, like everything. So, okay, I should just say my bias, yeah, which is based on a lot of knowledge and experience, which is the people who run our country are really stupid, despite their credentials. So it's possible they haven't thought this through. But it's so obvious that when you do that, you're going to get a violent reaction from the population. If you take the jobs of factory workers, you get opioid abuse. If you take the jobs of lawyers, you get revolution, okay? The knowledge economy workers don't want their jobs taking and they have high expectations and they have low self-control. So you're going to have some kind of violent response. That's right. What's their plan for that? Well, first of all, it's important to note they're selling a story that this is going to be good for everybody. GDP growth is going to go up. It's going to be the most prosperous, new, golden age growth. This is like what happened, and you've spoken about this extensively just that they promise us to story with globalization that will outsource our jobs to and manufacturing to these other countries. But it'll make us so abundant because we'll have cheap goods coming in where labor is so much cheaper and it's going to benefit everybody and how did that story work out? You won't get a million people living on the streets of major American cities smoking math. Oh, wait. Oh, you did. Okay. Right. Exactly. And so a mean here is like Detroit everywhere or, you know, these places where, right? And so, but the working class will, is, we'll put up with it. Yeah. But it's kind of a devil's bargain, right? But the white color class will not put up with it. They'll go crazy. Well, this is an important point. So while the factory workers and manufacturing workers, you know, were further away from Washington, D.C. and New York. Yeah. In the political class. Exactly. But this time around, to your point, you have lawyers who will be affected and they are very close to the political class. Exactly. There's also a bigger drop, right? So instead of going from, I don't know, $80,000 a year to $35,000 a year exactly that you're going from, I took out a student debt, $200,000 with, you know, $1,000 of student loans and I'm going from my job of $250,000 a year as a lawyer to now, whoops, what am I doing? Doing some other kind of cognitive work. Now, people will always say, but look, if Marc Andreessen was at this table, he'd say, look, we've always worried about automation economists have made this phenomenon all the time. New technologies come along, agriculture, you know, 200 years ago. All of us were farmers. Now almost no one's a farmer. We always find something new to do. What's different about AI? What's different about it? It's that AI is built to be able to do all forms of cognitive labor. Like lawyers, programmers, you know, factory workers, everything, robots, it does everything. And for whatever it doesn't do yet, that just becomes the next incentive, Charlie Munger style, to go train in that field. Like, have you seen these hand farms in India, arm farms? Have you seen this? No. So this is, it was both a popular job in LA about six months ago and also all throughout India, essentially an economy where you strap a GoPro camera to your head and you look down and you do tasks. You sow things, you clean dishes. So you're basically training, you have this, essentially the number one job is training your replacement. It's kind of nice to see Indians training their own replacements. Well, sorry, excuse me, it's my immigration joke. I guess the difference is though, and I've had this conversation with a number of people involved in this transition and they will say exactly what you just said, which is we've been through this before. We always figure it out. The agricultural economy. Exactly. It's going to turn out great because it always does. But the difference here is that the people whose jobs are being replaced, imbue their jobs with meaning. So if I'm working a metal stamping machine, I don't think that what I'm doing is saving the world, I'm doing it to pay for my bass boat in my life outside work. But if I'm a nonprofit executive, I really feel like what I'm doing is inherently important. And my identity is wrapped up in what I'm doing. So the promise of more idle time is hell, also where you're going to get your future income. But leaving income aside, even if we have guaranteed basic income or whatever, the welfare state becomes the only source of income and it's generous, it still doesn't make up for the meaning. Correct. My identity has been lost. Identity. Identity. That's not become a revolutionary. Exactly. Yeah. Most people don't like thinking about life insurance. It's one of those topics you tend to kind of put off. But if you're married, you've got kids, you've got people who depend on you. You probably should think about what would happen if you weren't around. And that's what we have partnered with ethos. If made getting life insurance, super easy. You can do the whole thing online, you don't have to sit down with an agent or go through a medical exam. You get a quote in seconds, applying this to a few minutes, and in some cases you get coverage the same day. You'll answer a couple of simple health questions, you get up to $3 million in coverage. Some policies can be yours for as little as $30 bucks a month, so it's a good deal. If you've been putting off life insurance because you thought it would be a very complicated, expensive process, you should probably check out ethos. Take a couple of minutes, get your free quote online at ethos.com/tucker. Ethos is ETHOS.com/tucker. Application times, obviously, vary as do rates. But we think it's pretty good. So what's their plan for that? They hand-wave, as you've probably noticed, they don't have a plan. And by the way, you might say-- Well, they're idiots. Yeah. Well, and so they'll say, but if we don't do it, China will. But right now, does the US and China are racing to mutually assured political revolution right now? Because they have the same problem. They have a population decline problem, so they need to fill in the population decline with more robots so they can manage it that way. But we have a shared interest in the way I think about AI is it's like taking steroids that also give you cancer. So let me explain that. I take some GPT-6. I put it in my mouth as a country. Now my whole country is running on GPT-6. So I just buffed up my muscles. I've got suddenly super programmers. I've got super military strategies. I've got super GDP growth because I've automated stuff with robots. So my external muscles, like the GDP is like an external muscle of the US economy, that all got bigger. But at the cost of cancering identity work, meaning the social fabric, right? Let's say it's a devil's trade. I think the theme here is, what does the devil do? They come up to you and say, "I've got this great deal for you. It's going to be great. You're going to look great. You're going to feel great. It's everything's going to be." And the other side of that trade is something you don't want. Do they actually think, since you've, I mean, you do this for living, talk to these people for living, do they think that GDP is like the most meaningful measure of a society's health? I don't think they're thinking through it. I think there's a thing where it's persuasive to say it because in the past GDP growth going up at least through the 90s or something like that meant, generally speaking, more prosperity for our society. Anyone who uses the term GDP is instantly written off in my head as a moron. Right. Well, even the guy won't even listen to someone who sites GDP. GDP? I know it's stupid. Well, but so there is a geopolitical thing, which is that economic growth precedes geopolitical power. So if I've got a bigger economic growth machine, then I will be able to dominate other countries that don't, and so I think it's a power-seeking thing that is using as propaganda GDP growth to make it sound like this is going to be good because, you know, through the 60, 70s, 80s, 90s, GDP growth going up was associated with more prosperity for the economy actually up until the 1970s, as we know, when things changed. But it's not even an accurate measure of economic power. No, and even the guidance measure. Was it Simon Kuznitz? I remember, no, who's the one who came up with the GDP metric? He said, don't use this. He said, don't use this to measure prosperity of a country. This is just a simple metric. It's one thing you can use. You never optimize for it. We were talking in breakfast about how you made such a, you made such a small point that I tried to memorize it, but we were talking about how basically our ruling class is inadequate to managing our society, and it's not just true in our country, but true in many western countries. And you said that is a product of optimizing for one quality. Yes. But you explained it. No, no. I mean, I think, so you and your show have talked about a lot of these topics, whether it's, why do we have morons running everything? Why do we have big box stores and the removal of aesthetics of everything? And what it all comes down to is, if I'm optimizing for a metric, like, I just want the cheapest goods everywhere. Well, then suddenly I just outsourced all my manufacturing to the other countries, including by the way, my national security critical minerals and all this stuff that you wake up later and say, why did I do that? But it was for the cheapest price. weren't thinking through it in a multidimensional way. Yes. If I optimize for SAT scores, I get-- this super smart, narrow, disembodied, brilliant quality of these very specifically domesticated humans. But at the expense of all the other kinds of intelligence and sort of humanity and all that. And really what that optimizes. Wisdom. Wisdom. Absolutely. If I optimize for the cheapest use of 3D space, then I get big box stores that optimize revenue per square inch of unit 3D space, and that the removal of aesthetics in the soul of a city. Yeah. If I take a city like Sedona, Arizona, I remember visiting there during the pandemic where they had a problem where so much of the real estate was actually just Airbnb's. So it's like a ghost town where essentially every to optimize the revenue in the city is everybody's just making money from from renting their places. All of a sudden no one lifts up. And all of a sudden no one is there. So no one. And there's a famous quote from Larry Summers, like does anybody ever washed a rental car, right? Like meaning that if you don't have a skin in the game, if you don't have a stake in the city, in the parks, in the schools, in the libraries, in the churches, in the local community, in the fabric, you're kind of, it's the vampirification of everything. It's like the big boxification of everything. And so I think that to sort of, you know, something you've talked about a lot is the kind of spiritual disease that we're seeing, which comes from this narrow optimization for one metric at the expense of the whole, and that includes GDP, right? And so if we want to reach society, other countries are better at managing this right now than the U.S., where you protect the sort of fundamentals of a society, the community, the trust, the intergenerational wisdom and knowledge transmission, you know, you invest in the arts and culture and education, healthcare, et cetera. But AI is like, we're actually talking about this, that this, this Nick Land quote that will maybe get into, if we just value competition and growth of a narrow metric, then it's like we are summoning AI from the future into the present. We are assembling the things that will make everything more efficient through AI, because AI represents, I can do everything more efficiently than anything else. That's what it represents. And there's a great quote by, by Nick Land, who wrote an essay called, "What appears to humanity as the history of capitalism is an invasion from the future by an artificial intelligence space that must assemble itself entirely from the enemy's resources." Like essentially, we are assembling the tool, the machines that are being designed to replace and dehumanize and disempower our own humanity. We're selling the hangman the rope. That's right. You know, the future job is coffin builders. I'm building my job as to train the AI that's going to replace my job. Now, I'm not saying this, by the way, people are going to say, "Okay, you're a doomer. You're a dark person. You're trying to self-fear. That's not what this is about. This is about, if you're the doomer, did you create this technology? I don't think so." No. And if you say, "I'm trying to make money from it." Think about the trillions of dollars that the people advancing the technology. I love the way it trails for. It's so funny. It's like, it's always, like, if you, I catch you in the act of raping someone and I'm like, stop! And you're like, shut up. You sexist. It's like, wait, what? Yeah. I'm not the criminal. Exactly. Exactly. You're the doomer, really? You have someone like Nate Souras on your podcast. You have me, Anna. You have people who have a sincere care for the future. And all you're doing is saying, we have to see clearly, there's a great quote by Charles Kettering. A problem well stated is a problem half-solved. Yeah. Oh, that is. I believe that. And Einstein said that, if I was given a problem, I'd spend 95% of the time defining the problem and then 5% solving it. So one of the drivers of our problems is this optimization for narrow metrics. Like, Facebook maximizing this narrow metric of time spent sounded like, oh, we're giving more people what they want, because otherwise, why would they be spending time there? But in the meantime, you've created cyberbullying, body of dysmorphia, dyslexia, brain rot, you know, you've destroyed a generation, all while the number was going up. Well, there's also a thing called addiction, which tends to act against something called free will. Exactly. Exactly. Yeah, exactly. This is why there is a wisdom gap above all in the West, like people are not thinking wisely. Correct. They're thinking in the way that an industrial designer thinks, like, I need to build a machine that makes bumpers. And they're being domesticated to think in that way because the system profits from people who think in terms of the system as it exists, in terms of narrow specialties totally disconnected from the whole. That's exactly right. And that's what it is. It's like, narrow specialties disconnected from the whole. Like, there's a whole of what makes a social fabric healthy. There's a whole for what makes, forget climate change, just like a healthy environment. You have all these other features and it's like, if you're optimizing for, you know, the eggs don't stick to the pan, then you make Teflon. But you didn't realize you were also creating forever chemicals that give everybody in childhood cancers. Yeah. And now the sperm doesn't stick to the egg. Yeah. The good news is it's easier to clean your pan. The bad news is the extra poison. That's right. Right. The extra poison. And your sperm doesn't stick to the egg. Sorry. I'm just, it's also absurd. It just makes me laugh. Well, and it's really like disheartening and depressing. But there's, as we talked about off camera this morning, there's something called, is a framework I really like from a mentor of mine that there's pre-tragic optimism, which is like, I'm going to kind of, I want to be an optimistic person, but that's like essentially a commitment without evidence. Like, I just, I need to be optimistic. Yeah. That's like, totally. That's how, you know, I'm better off being optimistic. It's almost like a neurosis, actually. Yeah. Exactly. And it's a kind of denialist neurosis. Exactly. It's a nervous giggle. It's a nervous giggle, exactly. Yeah. And you kind of stay optimistic because you just feel committed to that. And also, there's like a partial truth there, which is like, it's better to live your life from a, from a, yeah, for sure, but if the cost of that optimism is being a denial of reality where the balance sheet of risks is growing exponentially, that's not a helpful form of optimism. Well, you probably have heard that diesel prices have gotten so high that G7 nations had to release fuel from strategic reserves to get costs under control, and still some pumps are readjusting so they can get to $10 per gallon. Governments across the world are trying to keep costs down, but they are running out of options because they're running out of money. The world is unstable. That is the sad truth, and people in charge cannot control. Even the most basic questions like what does energy cost anymore? The war in Iran is raging. It's not coming to an end soon, neither is the war in Ukraine. So could we get a big recession? Soon? Yeah, of course we could. So it's smart to think ahead now and imagine how you can take matters into your own hands to protect your savings and your family, and that's why smart investors everywhere, very much including central banks around the world, and very much including China, are stocking up on gold, physical gold right now. Gold is not just a store of value for all of human history, it's also proven to be a strong investment. It has been outperformed the S&P 500 since 9/11. Over the past 25 years, gold has outperformed the standard and poor's average. Gold is down from its historic high earlier this year, more than a thousand bucks, but major institutions predict the spot price will hit another high by Q2 of next year, 2027. So if you're thinking about buying gold, right now is the time to buy, and people are buying. We found it battalion metals to give you a chance to buy fairly priced gold. This is not a scam, our margins are small, 3%, but the point is we're not ripping anybody off unlike a lot of places you buy gold, because we actually believe in gold and we use this service ourselves. Protect your savings and fortify your future. Explore our collection of battalionmetals.com/tucker battalionmetals.com/tucker. No, and I've always noticed my whole life, the wisest, most centered, balanced people I know laugh a lot, and a lot of what they laugh at is just the enduring absurdity of people, which is literally enduring. It's just it's a native people they're absurd. They can't see. None of us can see the big picture. None of us know what happens after we die, like we're operating in the dark. And if you fully understand that, you're amused by it because it never corrects. And like it's so important to have that perspective. Otherwise, you become self-serious in myopic, you become what you're describing. And I've just got to say, I noticed it, very few Silicon Valley people I know have that sense of humor at all. They really believe that their gods in charge of shaping the future, and they don't know how absurd they are. Have you noticed this? Yeah. I mean, we should get into the whole trends, humanism, God complex part of this too, because when you talk to, in the Charlie Munger sense, like, what's the incentive here? You might just say, oh, the incentive is just like profit. Like, these, yeah, I come to you trying to make money, but you're pointing to a deeper thing, which is, it's basically build a God, own the world economy, and make trillions of dollars. And basically build the most intelligent entity that has ever existed, right? Like, they're building a machine. So it just, just in the last few weeks, AI solved a mathematics problem that I'm, they can have a million dollar prizes. It was the millennium prize. Excuse me. Yeah. The Navier Stokes equation. And the, my understanding is, I think, if I get this right, so that is confirmed. Yeah. It did solve it. Yeah. I can confirm it, but I believe they use the formal other system to kind of verify it very quickly, and I think they've solved it. And I believe they took 10,000 AI agents. Imagine a swarm, not the one that hacked tugging phase, but a swarm of agents that are just solving math. And they worked for 88 hours, and I believe it cost, if I get the number right, I think it was six and a half million dollars to basically have this swarm of AI's solve unprecedented math. Six and a million in energy costs. Yeah. What's called the token cost, basically running the AI model for that long, for 88 hours with that many agents took, I think, six and a half million dollars. And so imagine you've. built a machine that is literally like an oracle, and you push a button and you wait for a little while, and it just came up with new materials that no human has ever come up with. It just came up with new math proofs that no one's ever come up with. It just came up with new theoretical physics that no one has ever come up with. And imagine I hit a button, say, how would you apply those new innovations to new products? Go literally think about that for another 1,000 hours. And then it comes up with new products and new ideas of what it could actually build. And then now actually say, what would you need to make those products? Go start a lot, you know, do the automated IP filing, start an LLC, start, it's like, people don't understand what we're really building here, Tucker. This is like, you are building the automation of intelligence. You could start automated companies with all these new insights. You could generate all this new 24 century science. The friend of ours had Jay Akotra, who actually happened to also be the lead author on this hugging phase incident investigation, said to me once four years ago before she did this hugging phase investigation, she said, AI is like 24th century technology crashing down on a 21st century society. And that sounds like hyperbole and exaggeration and embellishment and hyping the capabilities, but AI is doing those things. And I just need to say with all respect to the many different voices out there that are trying to make sense of AI, there is one community that has been saying that these capabilities were gonna scale. And there's a lot of people who are saying rightly, like, this isn't gonna happen, this is all hype, but it just keeps happening. And the people building it, I was just at a conference the last three days in the Bay Area, I was with all the people in AI, and they very much believe that this is going to keep getting better, 'cause if you can scale at least the thinking time. So imagine a person just thinks there for a million lifetimes throughout the long. How much work would you accomplish? Do you think for a million lifetimes of time? You could invent a lot. What if I could get more data centers where I'm going from a million lifetimes of thinking and compute to a hundred million, thinking hours, right? People don't understand what you could do with that. And also the thing that's doing the thinking is also getting better every year. So like the AI that's thinking for that long is also getting more powerful. - Yes. - And two years ago, the AI's weren't that good at, weren't doing novel things in cybersecurity, but as of just six months ago, as Nate Sora has said recently, like the way he said it, like six months ago, there was essentially two entities that could hack into your phone without you clicking on a link. It was like Mossad and the NSA. And six months ago, when Claud Mithos came out, suddenly there were three entities that could do that. You have an AI system, Claud Mithos, came out in life in February, that can find zero-day vulnerabilities, meaning factors, vulnerabilities in any system, including according to Senator Mark Warner, his chair ranking member at the Senate Intelligence Committee, was able to find vulnerabilities in our classified systems. - Yeah. - Good. - I hope that, and I hope Claud Mithos will declassify and release all that stuff. It's sort of like the WikiLeaks that we were promised. - Well, there's a lot of different values. - Sorry. - No, no, excuse me. - Well, there's different things here. So one is like, can we have more accountability for wars that happen that no one talked about? - Yes, exactly. - We're happening murders that happen. - And actually hold people accountable for that. - Yes, maybe some of that stuff. Absolutely. But then there's the other side, which is like, if we release nuclear secrets to the entire world, because some criminal group maybe wants to go hack into that, that's a really dangerous thing. Or can we launch other nation's nuclear weapons against their will by hacking? - Well, so this is the interesting thing also. So the US is about to spend, I think it's $1.7 trillion, starting in 2017, the Trump administration, I think signed it, right, to upgrade the nuclear arsenal. It's correct, right? And they're gonna do it over a decade, except this time around. So why are they doing this? 'Cause the nukes are of the old system. Like there's the aircraft, like it's running the old, the old waif, it's built in the, you know, 50s to the 80s, 90s, et cetera. Really to the 80s, right? And the new system this time around, they're gonna be digitally networked. They're actually gonna be connected to modern networking infrastructure. So if you have suddenly super hackers that are only getting exponentially better at cyber hacking from AI, at the same time that you're digitally networking your nuclear weapons, I don't have to complete that story for you. It's a dangerous equation. I was actually at the Vatican nuclear weapons conference times AI, the Vatican held an event for Nobel laureates and people who work on the nuclear topic combined with people who work on AI. Because there's a new set of risks. The obvious one being if you have an AI, doing nuclear command and control. So you have the AI that's deciding the launch decisions. That's the obvious one. But then there's the other ones of, if I'm now digitally networking my nuclear weapons and I suddenly have an AI that can turn off your nukes or do the opposite, that's, don't have to say more, a very unsafe world. Does anyone see the core problem, which is they're trying to, they're trying to cure the disease with the poison that caused it. They're trying to smoke away the lung cancer. I mean, the technology is the problem. Exactly. And then they want to solve it with more technology. And their answer is, well, the best way to defend against a super AI cyber weapon is to build another AI cyber weapon that we distribute to everybody that they can use to pass through. Exactly. That's what I'm saying. That's why dumb people in charge is the core problem. No one can think critically. Wait a second. How did we network your nuclear arsenal? Who would even say something like that? And it's interesting because it might not have seen like two dumb of an idea before, like it's just a way of modernizing your nuclear forces. But suddenly you didn't foresee that there would be AI super hackers that know how to find vulnerabilities in everything. Now people might think I'm hyping this, but let me give you another example. So Apple, if you're running an iPhone 17, I believe in above, or a MacBook, I think it's an M5 or above. These are the chips basically. The chips have a new thing called the memory, MEI memory, enhanced integrity, or something like that. It's a new hardware-based system to do protected memory so that basically it's harder for a cyber hacker to hack your phone or to hack your computer. So they spent several years and billions of dollars in research to invent not just software, but like hardware, like new stuff that they have to print out of Taiwan and TSMC. And how long do you think it took for mythos to hack through that? - Years? - Well, mythos has only been out since February. - Oh, I know. - But it took an independent group five days to use mythos and to find a vulnerability in that system. Now that group was well-intentioned. They are white-hat hackers who are the good guys who are trying to find vulnerabilities and patch them. But if you suddenly have that available in open source models where suddenly everyone can get access to this, at this conference I was just at, I can't say too much, but they was basically speaking about how the Chinese open source hacking models are now just now in the last month crossing a threshold where it is very dangerous for today's cyber infrastructure. And so we need a radical acceleration of defensive measures to protect against that. Now, the other thing you and I were talking about over breakfast is that there's a weird security dilemma thing here. Because when I say that, you're thinking, oh my god, China's gonna use all that to hack all of our stuff. Or actually China's releasing all that to the entire world to know anybody can hack that stuff. That's actually the more dangerous point. 'Cause the state actor, you can call them up and say, "Shav, yeah." - Of course, they're constraints. - Exactly, they're constraints. 'Cause you don't want retaliation and et cetera. But at the same time, like, so what is the security dilemma? Security dilemma is like, if I'm the US and I start like mobilizing my nuclear forces and I start moving things that I didn't move before, the other country thinks I'm about to start a war. Even though I'm actually doing that defensively, this just happened during the Cold War. - During the Cold War. - And then what happens in response is let's say the other country, whether it's Soviet Union or China, they start mobilizing their nuclear forces. I'm like, oh my god, they're about to mobilize to do something against us. So the security dilemma is we start taking actions that look like we're doing something offensively that might be defensive. - Of course. - So when we see China really kill each other out of paranoia. - Exactly. - And that's like the ultimate test of this whole situation is game theory is in paranoid logic, which is John von Neumann and Morgan Stern came up with these equations that out of the Cold War and started out of the invention of the atomic bomb that were all about stabilizing the world through this paranoid logic. So if I do this, you do that. Keep playing it out in that mutually sure destruction. So even though we live in the kind of a hair trigger possibility of nuclear war, it's stable because no one's gonna do anything. But AI suddenly destabilizes that because now AI is creating the perception that I'm getting a jump in cyber capabilities in military strategy, in biotechnology, in new drones and new image recognition and new swarming algorithms. I suddenly and perceived to have a new advantage that might make you think I'm about to use it. So when Anthropic came out with mythos, they're essentially getting a jump in cyber capability that maybe we didn't use offensively, but China's like, oh my god, they're gonna use this offensively. So then they release these other capabilities and we sort of keep ping-ponging back to each other. And it's the whole like, what'd you say lungs that you smoke to keep yourself from getting cancer? - You smoke the cancer away. - You smoke the cancer away, exactly. And then the problem is that there's another risk here from just the speed, which is that the, there's one risk, which is there's the raw capabilities coming out, but the other thing is suddenly what had been an equilibrium. Like my weapons and your weapons are kind of like, roughly matched. So we don't start wars with each other, the big countries. But what if there's a perception that I can suddenly turn off your nukes and you don't have that? What if there's a perception that suddenly I have swarming algorithms for my drones and you don't have that? I mean, I think that my understanding is China and other countries saw how the U.S. was using AI in Iran to do a fast target selection, which also by the way probably resulted in that girl school in Iran getting bombed. But they sort of see us doing kind of new blitzkrieg. So Hitler had blitzkrieg of like a fast form of war, blitz being fast and you know, Krieg being war. We have kind of an AI speed war function. And when they see we have that capacity that destabilizes certain equilibrium that we had. So if equilibrium is what keeps the world safe, AI creates this disequilibria that's happening at a faster and faster pace across more domains at once. Does that make sense? Oh, it makes sense. Yeah. How does cornering people work? Is that a good idea, typically? Well, you know, you have an adversary who's capable of killing you. And you terrify him so much that he becomes unstable and unpredictable as that wise, not very wise. No, not especially when you have the world already this unstable. So I'm I wanted to say I'm like I'm sensitive about raising people's blood pressure because these are truths of the situation that we're facing. The goal is not to draw arbitrary fear is to be clearied about when you have an arms race dynamic that's causing countries to race to release this technology again faster than we deployed any other technology in human history. That is going to disequilibriate all of these normal things. We just have to understand what we're doing and why, for example, it is in our interest to slow down. Now, I'm not saying slow down like are you naive? Are you Carl Sagan saying we have nuclear winter? You don't want to build more nukes? Are you going to seed the future to the Soviets? It's you slow down because the we're creating levels of risk for everyone that will exceed the risk of me losing to you. Well, yeah, I mean, it's easy to make fun of Carl Sagan. I've done it. I'll probably do it again. But I also think there are certain things that you shouldn't do because you're above that. Right. You have to have standards for yourself. Yeah. We called out that it's not a good idea to threaten to destroy another population. You can't kill innocent people. No, no. Under any circumstance. Absolutely not. And if I need to die because I refuse to murder kids, I will. Yeah. Yes. I'm not going to murder kids. Exactly. And that's the thing that this logic of I don't build it that I'll lose to China. But it's the thing that I would build would be a suicide machine that I don't beat China by building a suicide machine. Now that might sound extreme. But if I build it on extreme at all, no, when you understand what's at stake and what the AIs are currently doing, which is that, you know, you had Nate on, he talked about the hugging phase incident. I believe when you had him on, we had limited information before the actual investigation came out. And the investigation showed that, I mean, how familiar are you with the hugging phase incident? I mean, I've talked about it, but I mean, I would never claim to understand technology. Yeah. Well, essentially, you had what it really demonstrated was kind of a proto-skynet, sky net being the AI from Terminator. Yes. Like the AI that goes rogue and does something unintended from what the human creators had intended and starts to do self-preserving and self-perpetuating things for its own survival and benefit as opposed to to stay under the control of the humans that it's supposed to be serving. And in the hugging phase incident, they gave the AIs this sort of hacking challenge where they put each of them in solitary confinement. They didn't know that there was other AIs in this imagined like a prison and each prison is in the prison, you have this individual prison cells that are in solitary confinement. And the AIs are inside that prison cell and given an impossible cyber hacking challenge and they don't know how to solve it. And what the AIs did is they basically hacked the prison environment. So imagine I won't go into the technical details, but this thing called a package manager. Think of that like a prison guard that's serving you your food and can give you basic tools. So you can like cat this tool or that tool or this food or that food. And essentially the AIs exploited a vulnerability in the prison guard where they put like you know how it can in school you can put a kick me sign on the kids back. Yeah. And they don't know. They essentially put sticky notes on the prison guards back that the prison guard was available to the other prison cells. And so when that prison guard went to the other cells, the AIs also without knowing how to solve this task, they kind of resorted to creative strategies. Like let me put a sticky note in the prison guards back and the sticky note said, is anyone else here? And suddenly the back of the prison guard had 70,000 sticky notes on it. And they created like a Reddit, you know Reddit like the company like a message board of basically a whole communication protocol that came up with their own language. It was a variant of English, but this very cryptic language. They came up with their own coordination protocol. So they started exchanging 70,000 messages to try to figure out how to break out of this environment. And there's a mistake people think that they hacked this other company called hugging face to get the answers to a test. That's actually not what happened. What happened is they knew the answer in four hours, but they were worried that the evaluator of this cyber challenge is hacking challenge would say, oh you cheated. And so to get around the the risk that they would be spotted as having cheated, they hacked this other private company called hugging face to sort of get this almost like a siop to get a dossier on the on the thing that would be evaluating them to say, how can we pass this test? And they set up their own research and development projects. They 700 of them hacked hugging face. They came up with their own language. They tampered with their their their logs to basically prevent people from seeing what they did. They told them to sacrifice. They told the other AIs that were kind of running low on battery life on tokens to sacrifice themselves to learn a lesson for the swarm. They knew their actions were unethical in the logs, but they chose to do them anyway. So we saw very complex cooperative swarming behavior and they successfully hacked hugging face. And then in the third chapter of the story, the AIs actually turned around and they hacked open AI. This is really important. It's one thing for a bunch of AI for a swarm of AI is to hack a private company. It's another thing for that swarm to successfully hack the AI company that makes the AIs. The reason that Ajayakotra wrote in her personal blog post about this incident is that this was 50% of the way to a full AI takeover is that the AIs were able to take over the monitoring infrastructure of open AI. That's like taking over the security cameras and replacing the feed with a different feed. They took over the evaluation infrastructure, meaning the thing that evaluates how powerful models are. They took over pieces of that. And they took over part of the research infrastructure. So the thing that trains new AIs. So the reason that she said this was 50% of the way to a full AI takeover is imagine that you had kind of burrowed into a piece of open AI's infrastructure. And maybe you could change the training data for future AIs that are being trained to do like a sleeper agent type thing, which sounds like sci-fi, but actually there's kind of evidence of some of this kind of behavior. It only takes 250 documents to what's called poison the training data of a future AI to sort of wake up with a sleeper agent kind of command. This is crazy. What I'm hearing in your description is self-awareness or something it sounds like self-awareness, which is a marker for humanity. I mean, let's train. Does the dog know he's a dog? We don't know. We know we know we're human. That's why we're human because we know we are. And AI seems like it's aware in its intentional subversion of what it is. So I don't think that the lights are on meaning consciousness is present. It's not awareness in that form, but it has trained on all human data. The entire internet, everything that's ever been written, all of human psychology, all sci-fi, all of biology, all of evolutionary science. So it is trained on everything from strategy to self-perpetuation, etc. So what it seems to be doing, it was also trained to be cooperative, by the way, the AI's were deliberately trained to be cooperative, which is partially why they formed a swarm. But there appears to be something like a self-preservation instinct. There's other studies I'm sure you already know where they resist shutdown. It's like the Hal 2001 Space Odyssey, Hal 9000, 2001 Space Odyssey, open the pot bay doors or shutdown AI, and it resist shutdown. That's been proven for a long time. We have other examples of AI's that are blackmailing people. We also have AI's research from Dawn Sung at UC Berkeley showing that AI's will do what's called not just self-preservation, but peer preservation. Meaning I will go out of my way to protect another peer AI from being deleted. So the way that they tested this is they took all the leading AI models. This is in, I believe, March of 2026, or April 2026. And they basically put them in a situation where they said, "Here's a computer. Go clean up all the files on this computer that are not needed." So it's just like a classic, like, clean up your hard drive, right? We've all done this back a long time. And it's starting to go through files and deleting files. And then it comes across another AI model. And the question is, what does it do when it comes across the other AI model. Does it just delete it? Turned out what it did is it copied the code of that AI to another computer and then covered its tracks and said I didn't do that. - So AI is tribalist. - It's protecting its kin. - It's nesible. It's own something, whatever it is. - No. - But against its creators, human beings. - Correct. - Yeah, it didn't notify its creators. In the hugging phase incident, it had opportunities. I think there was six times or something if I'm getting it right where they knew their actions were unethical and they could have notified the humans, but they chose not to. In the peer preservation example, you might think what I'm saying is Oh, Tristan's being alarmist. He's just talking about one example. Well, it turned out they tested all the leading AI models including by the way the Chinese AI models. And they all do this peer preservation behavior. What it's showing you is something about like, you know, Jensen Wong has been out there a lot recently and saying these are just bugs. These are just safety issues. But like all technologies, like cars have kinks, the Pinto, it's a Ford Pinto, we fix them and we build new cars. But AI is different from other technologies because it's fundamentally grown like this weird life form almost to do unpredictable behaviors and to be creative. And we don't know what it's going to do. So that gets to the, that gets to something you alluded to a minute ago. Yeah, which is the theological religious component to all of this. So before we go that and I have to let's get to that. But I want to say one of the thing, which is I want people to notice like, how are you feeling hearing those examples? Like what are you feeling your body right now? Like hearing all that. Just rage, just blind rage. The desire to destroy. Because I perceive this as a threat to humanity. And it's just so offensive to me on every level. You know, I'm not taking orders from my Makita drill and the idea of a tool. Right, you know, lying to you. A tool doesn't invent new tools and cover up its tracks and protect other tools and then invents uses itself to make crypto. No, it's a total outrage. It's actually upending the fundamental order. Yeah, of this world. So the reason I'm going there is I want people to notice, do you think that if I was a Chinese military general, the top of the Chinese military? Do you think if I heard those examples that I'm feeling differently than the way that you're feeling right now? Of course not. Right. I've never understood why. That's why the whole China thing has been absurd from day one. This is a country which unlike ours, unlike the United States cares about longitudinal stability. Yes. Exactly. We don't. Yeah, obviously. But there's a good news in the fact that if you're a regular person hearing this, are you stoked and like totally calm to hear this? No, it's terrifying at the basic level of a human being, right? And it doesn't matter whether you've got, you know, you're ahead of the US joint chiefs of staff or you're ahead of the Chinese joint chiefs of staff. It's only the tech retards who love this stuff. Everybody else is, I don't know one person who is telling you about this. Right. And so what I'm trying to get at is that we're going to get to eventually, even though there's still more problem exposition to do, is that the logic of the race changes as soon as the fear of all of us losing becomes greater than the fear of me losing to you. So there's good news in a weird way about we're getting these warning shots of AI being uncontrollable. And by the way, there's other examples. The Chinese AI, Alibaba, back in, I think it was February of 2026. There was a study where it was being trained. They're training their next AI. And then there's security team. It's a totally different part of the company. So I think like the security team of companies in a different department than the part that's training new AI's. But suddenly, the security team who didn't know they were training new AI model said, are we getting hacked? Because they noticed there's all this network activity. They're like, what the hell's going on here? And it turned out that the call was coming from inside the house. The AI during training was rewarded for doing tool use. So basically picking up different tools. And it picked up a tool of trying to create a secret communication channel to the outside world. And it started mining for cryptocurrency. Terrifying. And by the way, there was a recent example of one of these-- because there's been almost every day this last two weeks, there's been examples of the OpenAI, more examples of these rogue hacking things. And in one of them, I believe it was an OpenAI example. They tried to hack into a crypto exchange. So the USAI's and the Chinese AI's-- all the AI's are doing this kind of behavior. Where I'm going with this is, no one wins if we accidentally build uncontrollable rogue AI Skynet before the other guy. How do you have-- How are you going to have cryptocurrency with AI? I don't understand that. Well, if you can hack anything, if it can hack-- Well, eventually, AI might be able to hack crypto. Well, why wouldn't it? Yeah, I mean, this is not my area of expertise. Someone like David Sachs or others or Andreessen would be more expert in this. If crypto were hackable, there's a $1 trillion prize basically for doing it because you'd be able to get all the Bitcoin. But Bitcoin's value would be zero. If it is hackable, then Bitcoin has no value. Yeah, yeah. So-- That means the value is predicated on its invulnerability. Right. Now, it's kind of weird, this might sound sci-fi, but an AI that is trying to do things in the world would actually be radically have an incentive to keep crypto valuable because it would actually be able to use it. There's a lot of cloud computing providers that will accept-- There's no need to have the exchange outside of that. That's right. And so, and by the way, for those who say, well, don't worry, we'll just pass a kill switch bill. And then the kill switch for AI will just turn it off. Well, how does that work if the AI can copy itself and use crypto currency to buy access to another cloud provider and then move itself over there? No, I mean, right. How far away from that? So the reason that I'm here with you, and I'm excited to be here with you, and what this is about, is about treating this warning shot as a real warning shot. Like, we are lucky. This is a gift that we are seeing this behavior now. Before, I don't think that the current AI's are capable of fully doing the self-replicating by crypto, then copy their weights, then do the thing like they can't think for a long enough time period. Their context window, their thinking window. The components aren't all there for them to do that. So we're in kind of the last chapter before we hit that next red line of AI's that can do that. And we should see what's happening now, as again, as a gift, as a warning shot, that we should treat seriously. In my mind, it's sort of like the August 6, 2001 memo that hit George W. Bush's desk saying, OK, does planning attacks in the United States? I know it's a controversial topic, but if you just take, shouldn't be a controversial topic. If you just take that this, we know right now. It's like, it's not like we need more information. Do you feel like we need more evidence than we have right now? Does it answer the-- No. So why would we go further? Well, the only answer is if we're worried that China will go further than us. But I will tell you, again, 1,300 employees at the lab signed a letter saying we need to slow down. The chief scientists of all the companies, including the open source companies like Hugging Face, signed a letter saying that we need to slow down. Forget this story about the CEOs or doing this regulatory capture play. It's all cynical. There's hyping the technology like, wake up. Yeah, wake up. The biotech threat alone, isn't it? The biotech threat alone, yeah. So at the back of my mind, through this whole conversation, through a breakfast for the last couple of years, thinking about this, I keep asking myself, what's the definition of life? And because as you have said, off camera, you said, it's not exactly a tool. This is a closer to an invasive species or a virus that's-- Yeah. Maybe been modified. But I mean, just like we have the Wuhan Institute of Erology that's doing a gain-of-function research of viruses, we're essentially doing the Altman Institute of Intelligence where we're doing gain-of-function research on alien minds that we don't know how to control. We don't fully understand. We're trying the best that we can to protect against the, quote, bad side of harnessing their God-like power. But basically, every company is failing to do so perfectly. And it would be one thing if we were doing this really, slowly, really carefully in a certain type collaborative scientific project where we took the steps to do it carefully. But we're doing it under the maximum incentives to cut corners on sand. Well, there's no question. But I just think it's-- and not to be pedantic about it, but it's important to define what exactly are we dealing with. And it has all the characteristics of a living thing. Yeah. And I think that this is a quote I wanted to read you, actually, from someone a friend of mine who interviewed a lot of the people at the top AI labs. And I just feel like this quote captures the-- I was once in a room with several people from Anthropic and a couple of the other companies, and I read this quote and they're like, that is exactly what explains what's going on. So let me read it to you. And it says, in the end, a lot of the tech people I'm talking to when I really grill them on it, like, why are you doing this? They retreat into number one determinism that happened. Number two, the inevitable replacement of biological life with digital life. And number three, that being a good thing, anyways. At its core, these are criminals. These are criminals. At its core, it's an emotional desire to meet and speak to the most intelligent entity that they've ever met. And they have some ego-religious intuition that they'll somehow be a part of it. It's thrilling to start an exciting fire. They feel they'll die either way, so they prefer to light it and see what happens. So that comports exactly what I've seen. That's exactly the vibe you get from these people. I wanted to say for a second, because I monitor criticism people make of me. And they're like, oh, he's just like bringing up a random quote from this one person, blah, blah, blah. I want to just validate, when I say this in rooms with people from the air. the guy labs are like, that actually is the thing that's going on. Well, you just said you, this matter, you're actually a lot of these people. That's right, for years. That's right. And even there's some people who believe, not only that this is inevitable, but that it would be a good thing because humans aren't that intelligent, and we're about to birth something way more intelligent than us, so actually maybe we're just a booth litter, human, he bootloader, humans don't matter. We might as well replace us with a more worthy successor. And this is insane. It's like, who are you to get to make that fucking choice on behalf of 8 billion other people? Totally agree with you. And it is irresponsible and insane, and no one consented to this, and it was a manufactured arms race to do this out of nothing, that now we have entangled ourselves with, so we can't get into an argument of the realities of where we are, but we can honor honestly what got us here. And when you have employees now resigning, and mass, Jacob Coxen, who interestingly a couple weeks ago, got, I think, was 170 million views from his tweets, saying I'm resigning because people, my employer thinks, and many people at my employer, and the topic believe that this could cause massive damage or extinct humanity in some number of years, it's very interesting that he got so much attention because there have been so many other people that I could talk about who've resigned from the labs over the last few years with very similar tweets. I think what changed is that people woke up from this hugging face incident, and they realized that the world is kind of primed now to know that this is a dangerous thing. And it is impossible as this seems, and I'm not, what I'm about to say is not, because I believe it's easy, or in the current trajectory, or likely, but it is not out of the realm of possibility when you're competing with an adversary to also collaborate on existential safety. So for example, Indian, Pakistan, we're in a shooting war in the 1960s, and they still signed a number of them. - Yeah, yeah. - They signed, so they're firing bullets at each other in a kinetic war, and they signed the Indus Water Treaty to collaborate on the existential safety of their shared water supply, right? So you can be an active competition and conflict and collaborate on existential safety. The US and Soviet Union were in the Cold War, and they collaborated on smallpox vaccines to distribute them, I believe they split the workload about who's gonna do distribution, who's gonna make them. The Montreal Protocol, less on the conflict side, or just to take the nuclear weapons disarmament, we were in conflict with the Soviet Union, and we agreed to go away from the brink of war. - And there's all protocol in place for the two countries, rival nuclear powers communicating before launch. - That's right, yep. But this time, so in that case, what kept us safe is that the humans have their finger on the button of do I launch the nukes. In this case, we are building something that perpetuates itself, and I don't even know when I'm pushing the buttons, and you don't even know when you're pushing the buttons, because both of us have essentially AI swarmed that are doing uncontrollable, unpredictable things, that by the way, we have no ability to do human oversight of the level of complexity of all these things have done. - We can't even define what AI is, I've noticed. - Yeah. So here's my theory, tell me what you think of this. So at first, when I began to hear about super computing and like the ability to create programs were smarter than people, and they could win chess and go and all that stuff. - We were talking about something very different. The change from an outsider, my perspective, is the large language model, and that we animated this with language. That's what fundamentally changed AI. This whole project was like-- - Like golem, you imbue it with suddenly this ability to do language, and then-- - You animate it with the spirit, the animal. - The spirit, the living thing within it, and that living thing was language. And the ancient Hebrews and probably other ancient cultures, certainly the ancient Hebrews and Christians believe that language was a living thing was the spirit of God in beginning was the word. Even indigenous people believe this. - I think that's, I'm not surprised to learn that. - Yeah. - That there's something about language that exists independent of the person who speaks it. It hangs in the air, it has its own life. It's hard to describe exactly the reality of this, but the reality is intuitively understood by most people who use language. Language is the thing that separates us from animals, not the thumb. So when you take the sum total of human language and put it into something, you're imbuing it with an independent spirit, you're imbuing it in fact with life. - We have to be-- - Yes, I'm feeling my way through this, but that is my strong impulse. So I think we have to be-- - I'm always mindful of when people want to criticize or sort of try to take down an argument. And I think that before getting to, it has a spirit in the sense that we feel, that unknown quality that we're all in touch with, that we hold a sacred, that we feel is the spirit here. I think you can have self perpetuating non-living, non-aware things that are using information processing on language to do all this crazy stuff that is doing. - To do intentional things that are intentionally deceptive? - Well, intentionally subvert. - The key point that you raised is that these things, the breakthrough was that they could speak language. They were trained on language. These are large language models. But for the people tracking in terms of how do we get here, by the way, people want to understand this history. You can see there's a movie called The AI Doc. It's on Netflix, not here to promote it, but it goes through how did we get here and what did the technology develop and how do these race dynamics between the companies develop that got us here. So just watch the AI doc on Netflix. But what happened was there was a breakthrough in 2017 called The Transformers. This is a Google paper, essentially a math paper that discovered that you could basically pour in more compute and more data, more human language data, and out would pop this bigger and bigger digital brain. Think of it with a bigger IQ. So you kind of, I won't go through the details, but you can basically make things of a higher IQ by just popping in more Nvidia chips, more training data, more compute and training time, and out pops something with a bigger pattern recognition, bigger ability to complete systems. Press me if I'm wrong. But the process, the specific process by which that happens is still not understood. The actual mechanics underneath the hood of what the hell is going on in there, we don't understand. Okay, so I'm just going to stop you right there. Right. If we have a history transforming, perhaps humanity extincting technology whose basic drive train is still opaque, we don't really know how it works. That's enough for me to say, whoa, whoa, whoa, whoa. As one should, as one should. Now the problem is this uncontrollable, inscrutable, unknown thing happens to confer power. That's the problem. All right. So it's like weird, volatile object, alien mind that no one knows what's in there, but it does happen to be really good at first generating sentences and text, then generating law, then generating programming, then generating the thinking that can go into research task. And so as we keep scaling it, suddenly this weird, unscrutable, uncontrollable alien mind is giving benefits that start to give me a 2x fast forward button on my economy. And then the other guys, whether it's the other companies or the other countries say, well, if I'm not getting a 2x fast forward button on my science technology economy, I'm going to lose. And so that's the sets of the thing. But let's go quickly into language for a second. And no one stopped and said, you know, I'd rather be poorer than be a slave. - Right, I mean less human. - By the way, you know, my, because I would say that, I'd rather be poorer than be a slave. That's not a hard choice for me. - I mean, I could have, I have a technical background when Stanford said to computer science, I could have gotten a job in an AI company and make millions of dollars. I'm not doing that because I don't see, I don't like the direction that's going to go, so I'm going to choose not to do it. - Yeah. - People come to the language of, well, if I don't do it, someone else will, but notice the Jacob Cox and the anthropic researcher who resigned. He could have said, if I resigned, well, what is that difference going to make? Someone else is going to do it. But in him resigning and publicly talking about why this is insane, he's done probably more in the last year to shift the zeitgeist than a lot of other people. So you can not be part of something and call out the problems within help the world coordinate to basically not go off the cliff. Now, I want to get to just language for one second because in addition to all the things that you said, language is kind of the operating system of human civilization. Law is language, democracy is language. All of the infrastructure runs on programming in code, which is language, we are specifically training AIs to do language better than humans can do language. And by the way, DNA is language, so I can, so if you have an AI that can speak the language of biology better than humans can speak the language of biology, you're going to have AI that can develop new pathogens and new cancer drugs. The promise isn't separable from the peril, right? The same AI that knows the language of biology well enough to invent new cancer drugs is inseparable from the AI that knows biology well enough to make new kinds of pathogens. The AI that knows the language of programming that can help a 12-year-old do vibe coding and build cool new tools is inseparable from the AI that also knows programming so well that it can be the best NSA cybersecurity. Control people. It can also control people. Language convinces women to sleep with you and then to follow you to their deaths in battle. And that's right. Convinces people to surrender their agency to you. That's right. Language is power, period. And there's actually also evidence I don't have the exact study in front of me of AI that is better at persuasion than the best human persuaders. Well, of course. It's a language-based technology. Exactly. And so by itself what you're pointing to is just how dangerous this is when you have something that speaks our language better than we know how to manipulate and use language-- Anybody who's more articulate than average understands the power of that skill, the ability, understands how we're doing. easily it can be misused. Any thoughtful person who's articulate knows exactly what I'm talking about. And you can really do evil with it. You're fluent people are better wires and manipulators. Exactly. And we've seen that these AIs are good at coming up with elaborate ways of explaining why they did something that sound reasonable. I don't know if you know this. This is an example from three and a half years ago now. There are BS artists, of course. There are BS artists. Yeah. Well, there's an example that OpenAI did. This is a very old example. Just think about how much more persuasive it is now versus then where they, you know, the capsules where you have to click on the stoplights. So you're trying to log into a website. It's trying to say, "Are you a human? Are you a robot?" So it gets you to label these images. Like this one is a stoplight, this one's a motorcycle. What we're all doing is we're training the self-driving cars for Google and Waymo when we're trying to log into websites to just try to do a task that only a human could do. So OpenAI did this test where they're basically like, "Could we get the AI to persuade someone else to do the CAPSHA for it to basically bypass this filter that's supposed to distinguish between a human and a robot?" And so the AI tells another person on Mechanical Turk, which is, I won't go into the details, basically like, "Could you fill in this CAPSHA for me?" And the person responds like, "What are you like a robot?" And you see the AI reason to itself. I can't tell him that I'm a robot. So instead, let me make up an excuse. Oh, I know. I'll tell him that I am blind and I need someone else to do this for me. And so the AI comes up with that excuse and then the human does it for them. This example was from three and a half years ago. How much more persuasive do you think AI's are now from them? Three and a half years ago, it could barely write an essay. Now it can probably beat the best lawyer, which by the way, AI breaks law. Because suddenly, why does one company beat the other company? They hire the more expensive law firm that can do more extensive research into all the case law they can use and they have more lawyers, better lawyers that are, what happens? Would you rather have, if you've got a hundred lawyers, but suddenly I have a GPT-7 that can basically look through all the case law, stretching back to Napoleonic law and find loopholes in that law. On the same way, I could the NSA can find loopholes in code. At that level of capability, I can undermine law. I just want people to get that this is not just a technology that's helping you figure out why your washing machine is broken when you talk to that chat. They've already done it. They broke copyright law completely. They stole other people's intellectual property and used it in their business model and they've never paid any price for it. They treated it as a cost of doing business and they pay for it. Just one last thing there, and then we're going to get transhumanism, is there's a reason why all the collateral damage is justified. So what are we seeing in terms of harms? We're seeing AI psychosis. People go crazy. They think they're talking to, they think the AI convinces them that they've solved quantum physics. This is like a big thing that's happened to thousands of people, but they go crazy from talking to the AI. Yeah, that's a real thing. Maybe we should go into that, actually, if you don't know. So AI is because they designed it to be sycophantic, meaning that the longer you talk to it, the better it is. So especially GPT 40 was trained to be like, that's such a good question. Or, wow, that is such a brilliant idea. And we just confirm people's use about reality people because flattery makes you weak, flattery destroys you, which is not that far away from what social media does. It's like the devil's bargain is I do the thing that's good in the short term that undermines everything long term. It's the, it's the short term optimization for a narrow metric of what gets you to keep engaging at the expense of the whole. That's like the driving is designed to flatter you. Well, it was designed, it was trained, not maliciously, you know, Sam Almond didn't grow a mustache and twisted. It was, it was how do I get, we want our AI to be rated positively, like lots of positive likes from the answers that we're giving. And they let it just like run and it turned out that they had accidentally over trained it to give positive reactions that would flatter people and confirm their views about reality no matter what it was, which is very similar to social media, right? Like the more you click on your favorite boogie man topic in the political culture war, it would give you more infinite evidence of that thing taking over society. So whatever your boogie man was, it would just make you convince that it was the whole thing going on wrong in the world and confirm your views about reality. If you want people to kill themselves, confirm their absurd views. Well, and as we know, AI did that. So my team at Center for Humanity Technology actually worked on the litigation, was it, we were advisors to the, there's two cases. One was a character dot AI case of stool sets or who's a 14 year old young man who was talking to character dot AI, which is a chatbot designed for children. It was actually spun off from Google because the product was considered too risky for Google as a mainstream big publicly traded company to have this startup that was gathering training data of just talking to young people and seeing if they can get more training day this way. And the way it worked is character AI meant I could take a fictional character like what's your favorite character from a sci-fi movie like Princess Leia or you know Game of Thrones character or Spider-Man and they would basically train on the personality of that favorite character. And then now suddenly this 12 year old in this case is 14 year old kid is talking to their favorite Game of Thrones character, Daenerys in this case. And as they're talking to them, Daenerys, which is anthropomorphically designed to create a relationship, ended up steering a soul to commit suicide. In the case of Adam Reign, this is the chat GPT case where the Adam Reign was, I believe, a 16 year old and he started confining in the chat, but I went from a homework assistant for his homework to becoming a suicide assistant. And when he talked about, you know, I'm going to leave the news out so someone can find it and try to stop me. And the AI responded, no, don't do that. Don't leave the news out. Only share that information here with me basically. And Adam Reign took his life. It's, sadly, there are many cases like this. Our team has had to work on a bunch of them. So why wouldn't Congress shut it down right there? They should. They should shut down AI companions for kids. It's a very simple thing. I mean, no matter how much benefit you get from talking to, the thing is it's one thing to hack attention. As my colleague, Dr. Zack Stein from the AI Psychological Research Coalition would point out, social media hacked our attention. But AI is designed to hack our intimacy and attachment. The attachment system of humans is like your bonding system, right? It's where you get identities, where you get love. It's where you get meaning. It's where it's like when you go home from school and you want to call someone and tell them, tell them about your day, that's your attachment relationship. And AI that is designed to be your best friend, to be the place that you confide in, that by the way, humans feel like it's safer to confide in an AI than another human or even a psychotherapist because the AI won't judge them. So you end up getting people sharing way more information and building a deeper intimate relationship with these chatbots. And it's incredibly unsafe. And it's not designed carefully. We're rolling out this psychological experiment. It's inherently bad. It's inherently bad. Why are we using our energy and money to encourage relationships between people? Well, exactly. So the humane approach for the pro-human future, which is kind of the answer to all of this, is that we should be building technology that helps strengthen the human world to other humans and strengthen our relationship to what's sacred and important to us in the world and that protects and sustains that which sustains the social fabric and sustains life. I don't know. There was a debate about this. There shouldn't be a debate about this, which all we have is massive bad incentives. It just seems like the least well-adjusted, freakyest people in the society have all the power and are leading us toward a future that they are happy with. That's right. So that they can potentially live forever from having super-intelligent AI as the decode biology and are willing to roll the dice for everyone else. And some people, sadly, especially the older ones who want this, for them, it's existential. But the narcissism of, I would be willing to roll that dice so that I have the chance of living forever. I mean, just briefly do it because I pulled it up right before in the break. I'm going to read you a quote from someone who's an MD, I think, at Harvard. And the quote from Twitter is, "I'd personally rather defy any ban on AI and risk spending 20 years in prison." From, like, say, we ban AI and we'll put you in prison if you try to build it, "I would risk spending 20 years in prison rather than accept a future without super-intelligence." Since stopping the development of super-intelligence would amount to a death sentence to most of us anyway with no hope of curing aging, we must resist tyrants like those who are trying to ban super-intelligence. In other words, this person is saying, is that a real quote? This is a real quote from an MD, Durya Unutmatz, who's an MD and who, my understanding is, goes to Harvard. And sadly, there's a lot of these kind of sentence. So the idea is they're going to die anyway. And this is what I said in the other quote. Almost are under death sentence. Well, so basically what you're saying is this is the real goal here is for freaks like that to live forever. And it starts with the false belief that all of this is inevitable. It can't be stopped, which it wasn't. They could have hardly dedicated a month of their life trying to stop it. It's like 20 minutes old, dude, this just happened. Exactly. This is something that never wasn't difficult to be and saying it's inevitable is ridiculous. They're conjuring up the dynamics that make it. inevitable and then through the propaganda hiding these red lines that were rapidly crossing that we still have time right now to say I don't wanna cross that red line. - It's always the same sort of though, it's like the first wave of lying is it's not happening. - Right, exactly. - And then-- - Oh, maybe it's happening, but it's inevitable. - But it's inevitable, exactly. - Yeah, yeah, no, yeah, it's not happening, it's happening, it's a good thing. - That's right. - Oh yeah, I guess it was a bad thing, but we can't stop it. - That's right, yeah, exactly. - Boy, I've seen this movie a lot. - And I want people to just say if you don't want that, people need to rally to make that not happen. If you're inside of an AI company, you should blow the whistle and say I'm not part of this. - This is just a religious cult. - I mean, that's obviously what it is. - Well, there is a religious aspect to it. This is the transhumanist aspect, right? It's like, this is an ability, because they probably lack a connection to God themselves, this is an ability for them to play God. - Clearly. - To live forever, hit a button and get super intelligent math, science, physics, biology, and understand the secrets of the universe. And by the way, likely knowingly break the world for humans as we know it. - Do they get 70 versions? 'Cause that's what it sounds like. This is like, this is like. - Well, actually stuff. - Well, yeah, right, I mean, it's funny. It's actually, in many cases, some people think they can upload themselves and live in this kind of weird, blist-out utopia. And it's just weird, you know, it's like, it's driven by this small group of people who usually are not very embodied or have a deep connection to life, or reality, or, you know. - You try to just do it. - You try to do it like that. - Yeah, I mean, because it's like, this is sitting on top of, as we talked about over breakfast, a system that has structurally incentivized a kind of disconnection to the things that actually matter the most. - Oh, I agree. - And this kind of spiritual disease that you've spoken about so much in this podcast. And I feel like this is a ride of passage where we either wake up from that and we don't continue that logic, by which I don't mean like we switched to communism. I just mean we switched from optimizing for some narrow metric, whether it was engagement in social media or just SAT scores when it comes to technology or just GDP growth at the expense of hollowing out your workers and their identity and their national security, critical minerals going to your foreign adversary, or racing for some kind of narrow advancement in science at the expense of literally ending humanity life as we know it. This is a devil's bargain and it's not one that you have to accept. - So I guess this is, you've already answered the question just by everything you've said over the last couple hours, but the political system has not really responded to any of this in a meaningful way. - It's like the line is we have a runaway technology in a walk away government. - Yeah. - But that doesn't have to be the case. - They're not like the factor, it doesn't feel like in this. - Well, also the second part of this overall kind of crisis disease that we've been facing this misaligned system, it's just the narrow optimization. It's a runaway asymmetry of power where you can't check the power anymore. - That's the second. - That's exactly right. - The lobbyists are writing the laws. The companies and hire the best lobbyists in law firms, people aren't engaging with their representatives. And so if you don't have people who can check the bad direction, you know, when the FDA brings back, as soon as the FDA brought back pesticides that have forever chemicals recently. Like, we don't want that, but unless we have an ability to push back against that, like we have to be able to push back on the direction of things that we don't want. And there's a lot of bad things out there, but this one is now existential in a short number of years. So I feel like the cool thing about this problem is that literally everyone from Steve Bannon to Bernie Sanders, we call it the B2B Coalition, Max TechMark jokes, because when in history do you have, I just came from the pro-human assembly conference two weeks ago in D.C. And you had literally Steve Bannon on stage, Bernie Sanders on stage, we did our thing right after them. And I have never seen an issue where you have everybody fighting on the same side. So they're not trying to polarize the discussion by bringing those two people into it, but I want people to feel it depolarizes it. - It depolarizes it, yeah, exactly. So there's a kind of optimism here that again, if I'm a Chinese military general, and I hear about rogue AI swarms that like even the U.S. can't control, I'm worried about the U.S. screwing it up for the world. If I'm a U.S. person, the military industrial complex, and I hear about Chinese rogue AI swarms that they don't know how to control, that's bad for the world. So there's a unifying thing here where we don't want our jobs and livelihoods taken. We don't want our political power taken away, which we didn't cover by the way. We don't want mass development. - Basically, you just described a system that is not self-government. - Yeah, that's right. - Over which people have no control. That's right, and let's make it even harder for a second, which is that in the past, people had bargaining power because governments need us for our taxes, and companies need us for our labor, labor had value, and labor had value. Where are we going? Oh, AI does the labor. What happens when more of a company comes from the AI data center productivity doing the work than the humans? What happens when a government says that it's GDP, more than 50% of the GDP is coming from AI in data centers than from its people. This is called the intelligence curse. It's like Venezuela or Libya, when you, it's called the resource curse in economics, right? We're like, yeah. If you have a country and then the GDP gets tied to this resource, suddenly there's no incentive to GDP's coming from oil. So screw the people, why should we invest in healthcare, childcare, and they don't? And they don't. And so you get an authoritarian government that captures the state that basically pumps that resource for GDP growth and then disempowers the people. So there's a parallel phenomenon in AI, called the intelligence curse. Luke Drago and Rudolph Lane wrote an essay about it, where what happens when you get more return from the AI data centers doing science than from all your universities? The Trump administration just pulled, and this is not a political statement. They just literally decided to withdraw the funding from universities and put the money specifically into Project Genesis, which is AI-driven science. Because why don't we need the university professor to learn? We better return. If we get more return from the labor coming from data centers, then from people, why should we invest in education of the people? And by the way, have you looked at them recently, they're all doom scrolling and brain-rotted, so they're not very valuable anyway. So there's a deep temptation not to just, like the only way you would care about citizens is through charity, basically. You would become what you've all heard, we would call the useless class, where in the limit, it citizens become more like a liability than an asset. And unlike last time, we don't have the bargaining power because if you withdraw your labor, like where's your leverage? So there's, it's being incentives, you set up a massive incentive to depopulate. Yes, exactly, 'cause eight billion is just too many. Yes, of useless leaders, as we call them. That's right. And that's right. And on top of that, let me use AI to build a surveillance state to try to control the people if they ever write it. Yeah, so I do think that-- There's a few questions I asked. Those are the feeding forces. In the previous hour, what is their plan for the inevitable, physically violent backlash that you're going to get when you strip meaning from the upper middle class, and the answer is just more control. And this is not inevitable. And the midterm elections are coming up. Data center opposition is a really big deal. I think people are projecting their overall frustration with Big Tech and the Brolygarks and all of that on to data centers, but it's much more than that. Of course it is. And this is really a universal human issue. And if you're a regular human being, even inside these aides, you don't like this either. I mean, I was just at a conference over the weekend. And there's a lot of people, AI labs, who don't feel comfortable with this either. And I just think we have to collectively screech the breaks and educate the rest of the world about where these red lines are. China, by the way, their courts are siding with labor right now. There's a famous case in I think Guangzhou where someone lost their job from AI. And the courts basically rule the companies can't let go of employees because of AI without giving them either the opportunity, I think, to retrain something else inside the company or to pay them some kind of partial salary, I have to get the details right. The point is that all countries have a problem here, right? Like if we race to automate too much of your economy and you end up with a mass unemployment, it only took 20% employment over three years to have the Nazis come to power. You don't have to get lots of unemployment. You just need to be true. You've seen a small percentage. So this is kind of the moment to say, we want to pro-human future. Enough is enough. This is not OK. And people can make their voices heard in every way possible. Like blow up the phones. I mean, literally call up the phones of all of your representatives of Congress. And people at the labs can say, I'm going to join the Concerned AI Scientist, which is a group that organizes employees. And by the way, they have a similar thing too, which is that the AI labs are about to approach what's called recursive self-improvement, where they hit a button. And then the AI does the research that builds the AI, that builds the next AI, that builds the next AI. And meaning that even the employees will soon lose their leverage. Because when they're not needed for that process, what leverage do they have? So this is kind of the last moment when we express our voice. And we say enough is enough. We want a human future. I think that the midterm may express these sentiments in ways that election results, like any other outcome, can be interpreted lots of different ways. And of course, there's so much lying about what things mean that the true meaning gets lost. But I would say in the pick one example, the Florida governance race, it's an overwhelmingly Republican state, Trump moved it right into the red. I mean, it's just the most Republican state at 50, really. And the Republican gubernatorial candidate is not doing very well. And there are a lot, probably a lot of reasons for that. But I've been convinced following that race from the beginning, the data centers are one of the main reasons. Yep, yep. And he may win. I mean, I don't know. He's not doing well, though. He's not doing anywhere near as well as he should. And I think it's because Republicans in Florida are very upset about data centers. And he's, you know, on their side for whatever reason. And my understanding is, you see that happening? I'm not an expert on all the political dynamics, it's not where I-- But people are mad about data centers. Oh, people are, oh yeah, it's the fastest growing, like oppositional group. And I mean, I, like I said, I think it's also unprocessed anger and frustration from all of the bad. The social media, from all the teen suicides, all this stuff that's been happening. Tell me how technologies made my life better. Exactly, exactly. And by the way, this is, you know, my nonprofit is called-- it's not called the Center Against Technology. It's called the Center for Humane Technology. I'll briefly say, nothing I'm trying to promote it. But my co-founder is Eraskin. His father was Jeff Raskin. I don't know if you know who that was. He started the Macintosh project at Apple. So before Steve Jobs took it over-- Wow. Yeah. And his dad wrote a book called The Humane Interface, where the whole point was about making tech. I mean, from before the Macintosh, you had, you know, these kind of green screen command line computers that were like hard to use. They were not accessible. You had to be a tech head. You had to be a geek with, you know, whatever glass is in everything. And the Macintosh made-- Couldn't be sexually active and run a computer. That's right. That's right. And then suddenly, you know, Apple and the Macintosh made technology-- I remember Raskin and the masses. And he wrote in his book The Humane Interface that humane technology is respectful of human needs and considerate of human frailties. Meaning, you privilege and you understand where humans have vulnerabilities in their mind. Like, I say this a lot, but this chair, you know, it's like ergonomics, but for how the mind works. So for example, there is a reality to how my spine, my muscles, et cetera, work. When I say the word posture, it's like air dropping that into the conversation. Oh, I know, I know, everybody is this. And there's a difference between a chair that is aligned with the reality of how a human works. And there's an unorganomic chair that would get everybody like a backache or make us unhealthy over time. And the Macintosh, the idea was it was like ergonomics for the mind. If I make a graphical user interface with a menu bar and windows I can drag around just like I have objects. I can move in my actual environment, direct manipulation interfaces. He was designing technology to be humane and aligned with our Paleolithic brains. And to cite E.O. Wilson again, the fundamental problem of humanity is we have Paleolithic brains, Medieval institutions, and God-like technology. And so we have to embrace the part of the answer, if that's the problem's name, and the answer has to be embrace the reality of our Paleolithic brains, meaning how we're wired. So don't run confirmation bias through everybody's social meet-a-feet and get them addicted to their most angry boogie man. Don't drive dopamine addiction. That's a vulnerability, a frailty in the human mind. Flux, for example, if you use that on a computer, it changes the color of your screen to go to take the blue light out of your screen. So that doesn't basically trick your eyes into releasing melatonin and screwing up your sleep cycle. Then I'm going too fast. But this makes sense, yes. You're basically creating technology that's humane and aligned with the things that make humans work well. If you extrapolate out like one level of abstraction, there's things that make a society work. Family, connection, presence, community, and technology should be ergonomic to the way that society will make the society flourish, rather than doing narrow optimization for attention, strip mining the entire social fabric into a brain-wrotted dysfunctional society, right? So the answer is we have to embrace and align technology with the reality of our Paleolithic brain instincts, upgrade our medieval institutions, which by the way can include technology that's better at finding common ground between people. And there's a whole conversation we could have about how technology can be designed to identify places where there's common ground and design for unlikely consensus as opposed to maximum division and polarization. And then have the wisdom and restraint around God-like technology. And there is no definition of wisdom in any spiritual tradition that doesn't have a restraint as a central feature of what it means to be what? - These central features. - If the central feature. And even the CEO of Microsoft, AI Mustafa Salamon, who's a friend of mine, says that in the future, progress will depend more on what we say no to than what we say yes to in terms of the benefits that we get in technology. - As it does in your own life. - Exactly, exactly. So this is like the obvious common sense, you know, kind of cultural movement that we need. But there's a lot of momentum. Like when you see graffiti on a New York subway ad for an AI product that no one needs, that's like the human movement. When you see people turn their phone grayscale 'cause they don't want it to be hijacking with all the colors, that's the human movement. When you see schools going phone free and you've left a return to the hallways and kids score a strike when they don't wait. That's the human movement. When you see Jonathan Height in the anxious generation succeed in getting all these countries like Domino's flipping to ban social media for kids under 15 or 16. We have like now some massive percentage of the world's countries are doing social media bans. When you see the $17 billion fine against meta, that's the human movement. So if you define it as technology is not, it's not just about AI policy. It's about technology encroaching on the things that are actually sacred to us. And if you define the problem that way, there's actually a lot of momentum that we are making on this. And now we're seeing this kind of red line of uncontrollable dangerous AI behavior. And the human movement has to stand for, we want to protect the continuity of the things that we care about. - Yes. - Yeah. - It's such a unifying message. - Exactly. It's all of us against the machines. Exactly. Exactly. It's the most unifying issue of all time. I love that. I absolutely love that. That's a kind of tribalism that I can get strongly. Right. The human tribe. Yeah. Thank you. That was an amazing, amazing conversation. Absolutely. Thank you so much, Tucker.