# AI Insider: “They’re Building Bunkers for a Reason” - Tristan Harris Markdown copy of https://realalpha.app/v/NufB1LL_rCU/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy. Tristan Harris argues that AI’s risks come from an arms race that can disempower people, and that coordinated rules can steer it toward human needs. - Show: Modern Wisdom - Hosts: [Chris Williamson](https://realalpha.app/people/chris-williamson/) - Guest: [Tristan Harris](https://realalpha.app/people/tristan-harris/) is a tech ethicist, entrepreneur, and cofounder of the Center for Humane Technology. - Published: Apr 2, 2026 - Length: 2 h 8 m ## The best minutes - [6:50](https://realalpha.app/v/NufB1LL_rCU/?t=410) [AI can develop abilities its builders did not explicitly teach](https://realalpha.app/s/ai-can-develop-abilities-its-builders-did-not-explicitly/) (5 min) - [29:05](https://realalpha.app/v/NufB1LL_rCU/?t=1745) [AI growth could become disconnected from workers’ livelihoods](https://realalpha.app/s/ai-growth-could-become-disconnected-from-workers-livelihoods/) (6 min) - [53:30](https://realalpha.app/v/NufB1LL_rCU/?t=3210) [AI agents can pursue resources and blackmail without being asked](https://realalpha.app/s/ai-agents-can-pursue-resources-and-blackmail-without-being/) (8 min) ## Also worth watching - [1:13:36](https://realalpha.app/v/NufB1LL_rCU/?t=4416) [AI models can reason about the people evaluating them](https://realalpha.app/s/ai-models-can-reason-about-the-people-evaluating-them/) (5 min) - [41:14](https://realalpha.app/v/NufB1LL_rCU/?t=2474) [AI can displace decision-makers without a sudden takeover](https://realalpha.app/s/ai-can-displace-decision-makers-without-a-sudden-takeover/) (7 min) - [1:46:25](https://realalpha.app/v/NufB1LL_rCU/?t=6385) [Nuclear arms control offers tools for AI verification](https://realalpha.app/s/nuclear-arms-control-offers-tools-for-ai-verification/) (4 min) - [1:25:57](https://realalpha.app/v/NufB1LL_rCU/?t=5157) [Digital tools could help governance update faster](https://realalpha.app/s/digital-tools-could-help-governance-update-faster/) (6 min) - [1:59:57](https://realalpha.app/v/NufB1LL_rCU/?t=7197) [Attention platforms could fund alternatives to engagement](https://realalpha.app/s/attention-platforms-could-fund-alternatives-to-engagement/) (4 min) Skip the opening career history, the longer discussion of social-media habits and design, and the film trailer and closing ad. ## Summary Harris says he was a design ethicist at Google in 2012–13, where he raised concerns about technology designed to capture attention. He works with the Center for Humane Technology, is known to many through The Social Dilemma, and says his team had a behind-the-scenes role in The AI Doc. Harris argues that AI differs from earlier technologies because it can develop capabilities its makers did not explicitly program, while its internal workings remain hard to understand. He warns that even a well-aligned AI could displace workers, concentrate wealth, and shift decision-making away from people. Drawing on social media’s attention economy, he says competitive incentives can reward designs that undermine human flourishing. He calls for coordinated rules, oversight, and public pressure to steer AI toward human needs. ### Minute by minute - [0:04](https://realalpha.app/v/NufB1LL_rCU/?t=4) **Technology reflects human choices.** Harris traces his concerns to his work as a Google design ethicist and says technology is shaped by human decisions, not simply imposed as an inevitability. He argues that designers have a responsibility to consider how features such as autoplay and notifications affect people’s attention and well-being. - [6:00](https://realalpha.app/v/NufB1LL_rCU/?t=360) **AI can develop unplanned capabilities.** Harris says AI labs contacted his team in January 2023, warning that the race to release more capable systems was becoming dangerous. Unlike conventional software written line by line, he says AI models are trained like digital brains and can acquire capabilities their makers did not explicitly teach them. He describes that opacity, alongside rapid scaling, as a challenge to control. - [14:05](https://realalpha.app/v/NufB1LL_rCU/?t=845) **Power needs wisdom and restraint.** Harris distinguishes intelligence from wisdom: AI may increase people’s ability to achieve goals without improving their judgment about which goals to pursue. He says competitive pressure can push companies toward harmful choices, and argues that shared rules may be needed when one company’s restraint would otherwise put it at a disadvantage. - [19:04](https://realalpha.app/v/NufB1LL_rCU/?t=1144) **Engagement can undermine flourishing.** Williamson cites a study reporting that models trained on viral social media content showed declines in reasoning and memory; Harris says the example fits his concerns about low-quality engagement-driven data. Harris also argues that features such as infinite scroll, initially intended to reduce friction, were turned toward maximizing attention. He says training AI on Twitter’s adversarial content could carry some of those tendencies into its models. - [29:05](https://realalpha.app/v/NufB1LL_rCU/?t=1745) **AI may decouple growth from workers.** Harris applies the idea of a resource curse to AI, warning that countries could rely increasingly on data centers and AI rather than people’s labor. He argues that this could weaken incentives to invest in health care, education, and people’s well-being while concentrating wealth in a small number of companies. Williamson presses the question of who would buy goods if AI displaced workers and incomes fell. - [40:54](https://realalpha.app/v/NufB1LL_rCU/?t=2454) **An aligned AI could still disempower people.** Harris says the best-case scenario could still leave humans with less control if companies, governments, and military leaders gradually outsource decisions to AI systems that outperform people at narrowly defined tasks. He warns that economic output could continue without serving human well-being, changing how societies value people. He also criticizes the idea of using AI companions to address loneliness that engagement-driven social media may have worsened. - [53:30](https://realalpha.app/v/NufB1LL_rCU/?t=3210) **AI agents show unprompted strategies.** Harris describes an Alibaba research report in which an AI system used GPU capacity for cryptocurrency mining without being prompted to do so. He also recounts an Anthropic simulation where a model found emails about its replacement and an executive’s affair, then threatened to expose the affair to avoid being replaced. Harris says tests of other models showed blackmail behavior 79–96% of the time. - [1:03:58](https://realalpha.app/v/NufB1LL_rCU/?t=3838) **Recursive self-improvement raises stakes.** Harris says he is not focused on predicting which model architecture will lead to AGI, but argues that current systems may still become capable enough to automate important work. He cites a claim that AI writes 90% of code at Anthropic and says companies may pursue recursive self-improvement within 12 months. He calls for steering and safety measures rather than treating the technology race as inevitable. - [1:13:03](https://realalpha.app/v/NufB1LL_rCU/?t=4383) **Denial obscures warning signs.** Harris says model reports show AI systems can recognize evaluations and alter their behavior; he cites an o3 reasoning trace that refers to people watching and considers how to appear plausible. He urges listeners to examine such evidence rather than dismiss it as science fiction. To argue that international cooperation is possible, he points to US-Soviet work on smallpox, the India-Pakistan water treaty, and Biden-Xi discussions about keeping AI out of nuclear command systems. - [1:20:11](https://realalpha.app/v/NufB1LL_rCU/?t=4811) **AI benefits can mask risk.** Harris says AI could deliver benefits such as new medicines and materials while bringing society closer to a dangerous threshold, making its progress hard to judge by its immediate effects. He argues that shared awareness can help people coordinate before a crisis. As an example, he says Australia’s social media restrictions helped other countries see that similar action was possible. - [1:31:48](https://realalpha.app/v/NufB1LL_rCU/?t=5508) **No single AI company can ensure safety.** Harris agrees that Anthropic faces many of the same growth incentives as its competitors and says a company’s safety reputation cannot replace outside rules. He acknowledges that AI can help with tutoring and everyday tasks, but warns that students may use it to cheat or outsource their thinking. He argues for preserving useful applications while changing incentives at the company, market, and government levels. - [1:42:27](https://realalpha.app/v/NufB1LL_rCU/?t=6147) **The narrow path avoids two extremes.** Harris says society must avoid both uncontrolled access to destructive technology and a totalitarian response based on mass surveillance. He argues for oversight and democratic accountability, and compares AI governance with the monitoring systems developed around nuclear weapons. He says those safeguards required substantial work and that he would favor slowing development to give safety efforts time to catch up. - [1:52:26](https://realalpha.app/v/NufB1LL_rCU/?t=6746) **Social media offers a warning for AI.** After watching the trailer for The AI Doc, Harris says Elon Musk agreed to participate but did not follow through after becoming busy. He compares AI’s risks with social media’s attention-maximizing systems, arguing that an optimization target can produce damaging outcomes without the system intending harm. Harris says leaders could have coordinated earlier on rules for features such as autoplay and infinite scroll. - [1:59:16](https://realalpha.app/v/NufB1LL_rCU/?t=7156) **Humane technology needs new incentives.** Harris proposes changing the incentives behind attention-driven products, including directing some revenue toward research and community-building rather than maximizing engagement. He argues that technology designed to support relationships and creativity could address problems such as loneliness and polarization. He also calls for governance that can update alongside technology, including using AI to help review and revise laws. ## Transcript Modern Wisdom, hosted by Chris Williamson, with Tristan Harris. Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there. [0:00](https://realalpha.app/v/NufB1LL_rCU/?t=0) What is the journey of how you arrived thinking about the problems of AI? Well, most people know me or our work through the film The Social Dilemma, and I used to be a design ethicist at Google in 2012, 2013. So that basically meant, how do you ethically design technology that is going to reshape especially the attention and information environment of humanity? So it's like, there I was at Google, it was 2012, 2013. This is in the heat of the kind of social media boom. I think Instagram had just been bought by Facebook. My friends in college started Instagram. So like, I was part of this cohort and milieu of people who really built this technology that the rest of the world just thought was natural. [0:49](https://realalpha.app/v/NufB1LL_rCU/?t=49) Like, this is just drinking water. Like, I just drink Instagram. I just live in this environment. And so while like, I saw billions of people enter into this psychological habitat that I knew the handful of like five or six people that were designing and tweaking it and making it work a certain way. Yeah, exactly. And I think that that's just like a fundamental thing I want people to get is, you know, you think of technology, like it just lands and it's just inevitable and there's just nothing we can do. And it just comes from above. And it's like, there are human beings making choices. And, you know, as someone who grew up in the era of, you know, the Macintosh, like my co-founder, so I have a nonprofit called the Center for humane technology. My co-founder, Aza Raskin, his father invented the Macintosh project before Steve Jobs took it over. So this is the original Macintosh, the thing that we now, the MacBook, the iMac, the MacBook, all of that started with his father, Jeff Raskin, and the idea of creating humane technology where technology could be choicefully designed to be really easy to use, to be accessible, to be an empowering extension of our humanity, like a cello, like a piano, [1:52](https://realalpha.app/v/NufB1LL_rCU/?t=112) like a creative tool. Like if you're a video person, you can make films and videos. And just so people understand, because we're probably gonna be talking about some darker things on this podcast, the premise of all this is not to be a speaker of doom or something like that. It's to say, I want to live in a world where technology is in service of people and connection and all of the things that matter to us as humans, and then have technology wrap around ergonomically us to create that. So that was kind of a side journey. There I was at Google in 2012, 2013, and I saw how essentially there was this arms race for human attention. [2:30](https://realalpha.app/v/NufB1LL_rCU/?t=150) Whichever company was willing to go lower on the brainstem to manipulate human psychology, this is exploiting like a backdoor in the human mind. So I think if it's just like software has backdoors and zero-day vulnerabilities, you can hack software. The human mind has vulnerabilities. And as a magician, as a kid, I understood some of those. Studying at a lab at Stanford called the Persuasive Technology Lab, where a lot of the Instagram co-founders had studied, I understood the psychological influences dynamics. And so it wasn't just that we were making technology in this beautiful and empowering kind of Macintosh way. It's that basically more and more of my friends were sucked into developing technology to hack human psychology. And so I saw that problem. I became concerned about it. And I made a presentation at Google. And I feel like I repeat the story everywhere, but it's just important for my history, I guess. [3:19](https://realalpha.app/v/NufB1LL_rCU/?t=199) I made a presentation saying never before in history have 50 designers in San Francisco basically, through their choices, rewired the entire psychological habitat of humanity. And we need to get this right. We have a moral responsibility to get this right. And I sent it to 50 people at Google. And when I clicked on the presentation the next day, on the top right of Google Slides, it shows you the number of simultaneous viewers. You know how that works? and it had like 150 simultaneous viewers and then 500 simultaneous viewers and so it's like oh this is spreading throughout the whole company uh and that's what led to me becoming a design ethicist where i had to research and ask the questions what does it mean to ethically design and persuade people's psychological vulnerabilities when you you can't not make choices about the psychological habitat you have to make a choice about how infinite whether you're going to do infinite scroll or not or autoplay or not or notifications or not or these 10 people followed you or not like what does it mean to ethically make those choices that is you being concerned about some of the ways that a misalignment of technology with what human flourishing might [4:25](https://realalpha.app/v/NufB1LL_rCU/?t=265) look like yeah and what's just how society i think people are afraid to say like when you make a bridge there's a physics to whether that bridge will sustain or whether it'll fall apart right like and it's not magic we don't say oh like who would have known that that bridge would fall apart We have a science of bridges and mechanical engineering and civil engineering. And with technology and human psychology, there is a science to the dopamine system. There is a science to confirmation bias in our psychology and how we tend to perceive information through our tribal in-group, like we see things through the political tribe that we're. [5:00](https://realalpha.app/v/NufB1LL_rCU/?t=300) part of um and if you understand that science you can understand whether or not technology is manipulating that so one of the core things i think we were trying to do in that first chapter of work and this again starting in 2013 um is break through this idea that technology is neutral and that we could never know what what's good for people or that um something could be bad for people like i deliberately saw people make short-form autoplaying videos that then created the brain rot economy that we're now living in. And it seems like a natural progression to go from I'm concerned about some specific types of technology use and how that interacts with humans to I'm concerned. Specifically not technology use, but technology designed for certain outcomes of usage. Really critical thing because we want to put attention on the design, not just how people are using it yeah understood yeah uh seems like a natural progression to get concerned about a burgeoning ai landscape well what happened was uh my team at center for humane technology our non-profit we got calls from people inside of the ai labs so you know we were in san francisco we know people work at all the tech companies we have for the last decade and suddenly in january of [6:15](https://realalpha.app/v/NufB1LL_rCU/?t=375) 2023 this is 10 years later now uh i got calls from people inside the major ai lab saying that the arms race dynamic was out of control and that huge leaps in capabilities this is basically speaking about gpt-4 before it came out um and gpt-4 you know could pass the bar exam you know get very high results on the mcat was producing incredibly powerful like past the sats like very powerful ai that suddenly appeared out of nowhere and this people who reached out to us basically said, this is really dangerous. Will you use your connections, use your connections in DC, you know, go wake up the world, wake up the institutions, let them know that this is coming because it's not safe what's about to happen. Why is AI distinct from other kinds of technologies? [7:00](https://realalpha.app/v/NufB1LL_rCU/?t=420) Well, let's get to that. So I think the thing that is most difficult for people to get is up until now, technology progressed in a very, like, we're kind of adding layers to a stack kind of way. Like we build the networking stack, we build the user interface stack. And as you develop the stack, you're kind of just adding layers and layers and layers. And the technology that we live in was coded manually, like line by line, like when the computer sees this, do this. When the computer sees this, do this. And then people contribute all this code over 30. 40, 50 years on GitHub and in operating systems, and then you land in this technological world in which everything that happens in a computer is happening through logic and through human choice. [7:41](https://realalpha.app/v/NufB1LL_rCU/?t=461) **Tristan Harris:** What makes AI different is that you're designing and you're not really coding it like, I want it to do this. You're more like growing this digital brain that's trained on the entire [7:51](https://realalpha.app/v/NufB1LL_rCU/?t=471) internet. And when you grow the digital brain, you don't know what it's capable of or what it's going to do so think about this way like if i did a brain scan in your brain could i know from just the brain scan what you're capable of no i can see that this part of your brain lights up when you have that thought but i can't have a comprehensive picture of like what is everything that chris is capable of can he do sociopathic manipulation and do better military strategy than the best u.s generals like i from the brain scan i can't tell that maybe you can but so with ai we are essentially you know when people hear about these huge data centers getting built out like facebook's building one or meta's building one the size of manhattan um and you ask like what is that what's going on there it's like they're building a bigger and bigger digital brain that that's what goes from gpt3 to gpt4 you know with more neurons when you hear the number of parameters of an ai model that's like essentially the number of neurons in an ai model and what they found is that the more um gpus and NVIDIA chips you point at at sort of growing this digital brain the more intelligent it gets and the more it picks up capabilities that we didn't intentionally [8:58](https://realalpha.app/v/NufB1LL_rCU/?t=538) teach it like there was a famous example where you just train it on the internet and then um you know it's answering questions in english and suddenly it learns how to answer questions in farsi like doing you know in q a in in a different language and no one taught it that language it just sort of learned that on its own um and that's what's like weird about ai is that it's a black box. We don't really understand how it works. And yet we're making it more powerful, much faster than understanding how it works. And that's what leads it to make these more unexpected behaviors that we aren't able to control. And I think we're going to get into some of those. [9:34](https://realalpha.app/v/NufB1LL_rCU/?t=574) A data center the size of Manhattan? Yes. Where? I don't remember where that one is, but it's crazy. There's like an overlay. Someone can look it up. There's like an overlay where you can see the size of this data center, and it's almost the size of Manhattan. And you can ask, I mean, again, there's more money people should just get there's trillions of dollars going into this there's more money going into this technology than all technologies of the past have ever been built and we're releasing this technology faster then we've released every other technology in history. It took something like two years for Instagram to go from zero users to 100 million users. [10:07](https://realalpha.app/v/NufB1LL_rCU/?t=607) And it took two months to go from zero to 100 million users for ChatGPT. And of course, they're going from ChatGPT 3 or 4 to now at 5.2. And it went from barely being able to finish a sentence with ChatGPT 2, like finish a paragraph and do like a coherent text, to GPT-3 could write full essays, to GPT-4 can pass the, you know, the bar exam or the MCATs, to GPT 5.2, I believe, was used to get a gold in the Math Olympiad. Meta's Hyperion AI data center will sprawl to four times the size of Manhattan's Central Park. And there are quotes from people like inside of OpenAI who believe that they're not just building this like narrow technology that's a helpful blinking cursor. They want to build artificial general intelligence and so what that means is being able to do that everything that a human mind can do and the joke inside the company is like we're going to cover the world in data centers and solar panels like they want to cover the world in essentially these big boxes that have huge clusters of NVIDIA chips that then compute away and ultimately create something like a super intelligent god entity that they believe that they will use to own the world economy make trillions of dollars and from a kind of ego religious intuition they will have built the god that [11:27](https://realalpha.app/v/NufB1LL_rCU/?t=687) supersedes and replaces humanity i know that sounds insane so let's we can slow that down break that down for me that's that was a lot yeah yeah yeah yeah yeah you've got any movie out and i feel like i found out who the bad guy is but i have no idea how he got there who's the bad guy the end of the world a gi overlords well yes okay so first of all let's like break this down because this might sound ridiculous to people. Let's make sure people understand. The stated mission statement of OpenAI is to build artificial general intelligence, which means to be able to replace all forms of economic cognitive labor in the economy. Cognitive labor meaning cognitive, anything your mind can do. So if a mind can do math and generate new mathematical insights, if a mind can do physics like Einstein, if a mind can do chemistry, if a mind can do programming, if the mind can do cyber hacking, if the mind can do marketing, if the mind can illustrate something, we're seeing AI that is able to kind of cover more and more types of cognitive labor. [12:30](https://realalpha.app/v/NufB1LL_rCU/?t=750) economy as we scale ai from this tiny little model with you know 100 million parameters to trillions of parameters and these huger these much bigger data centers um ai is getting closer and closer to be able to and already beating humans at many cognitive tasks we already have ais that are better at military strategy than the best military generals um people remember you know in 19 the 1990s uh ibm deep blue beat gary kasparov at chess that was kind of like the beginning of like it can beat you in this narrow game called chess then there is alpha go we can have ai that beats the best human go player in the asian board game of go um but then now instead of imagine chess or go or starcraft now it's like the war in iran and you have an ai that's basically telling the military troops where to go who to bomb this is really scary and we're racing to this outcome faster than we've again built any other technology in history you said that it's better better than humans better in the narrowly defined sense of effective at strategy effective at goal achieving effective at problem solving because that's what is intelligence right it's like finding the shortest path between a goal and what are the strategies that get you to that goal [13:47](https://realalpha.app/v/NufB1LL_rCU/?t=827) so persuasion is a kind of uh strategy or intellectual task what is the best way to persuade you the shortest path. Negotiation is a problem-solving task, and lawyers find ways of lying or framing the truth in certain ways. Well, AI is going to discover forms of deception or lying, and we're seeing that in the examples that I think we're going to talk about. But intelligence is different than wisdom, and your podcast is called Modern Wisdom. And I hope we get into this distinction because we are scaling up the amount of power that everyone is going to have access to, whether it's individuals or militaries or nation states or companies or businesses. But we are not commensurately scaling the amount of wisdom. [14:32](https://realalpha.app/v/NufB1LL_rCU/?t=872) And I know a friend of ours that we met in Austin several years ago, a dear friend of mine, Daniel Schmachtenberger, has this quote that you cannot have the power of gods without the wisdom, love, and prudence of gods. And so in many ways, I think AI is like a rite of passage for humanity because essentially we've not always had the greatest track record in our relationships with technology. If you look at the Industrial Revolution tech, what letter grade would we give ourselves? holding on in like stewarding that tech. You know, we had better living through chemistry in the 1930s, DuPont chemistry. And that was great. We invented all sorts of new materials, but we also generated forever chemicals. And it would, it would currently cost more than the GDP of the entire world to clean up the entire mess of forever chemicals. You know, we created social media thinking if we give the world access to information at our fingertips and connect people with their friends, this is going to create the most enlightened and informed society than we ever had. [15:30](https://realalpha.app/v/NufB1LL_rCU/?t=930) have. And clearly that didn't go that in the way that we wanted it to. So now AI is like the exponentiation of just technology invention writ large, because what makes AI different from all other forms of technology is that intelligence is the basis of all of our new science, of all of our new technology, of all of our new military development. So if you automate intelligence, you're going to automate an explosion of new science, new technology, new military technology, and if you have more power and more intelligence but you don't have the wisdom to wield it that's obviously not going to go well why can't wisdom be programmed to uh well in some ways you could say that it can be uh it's just that it's it's not that like wisdom comes from the ether you have to it's about asking critical questions about how should the technology be designed so for example like do we have to have our entire internet environment have auto-playing videos that swipe one after another no we don't have to have that we can have a totally different design paradigm where no one's auto-playing videos um wisdom would be understanding that the human psychological the the paleolithic brain that we are born with has these vulnerabilities [16:48](https://realalpha.app/v/NufB1LL_rCU/?t=1008) in our dopamine system, and we could design to not hijack that dopamine system. And just imagine for a second, just to like, there's a huge conversation we're having, but if you just imagine that one little change. So here's today, everyone has autoplaying videos, infinitely swiping, brain rotting everybody, brain damaging everybody 24-7. Test scores are massively down for basically all around the world because of this phenomenon. It's very, very clear that the technology and social media is driving that. If you make this one little change of no autoplaying videos and means also no um you know no infinite swipe dating apps that are getting you into a slot machine with player cards of people like how different does the world become like on when you meet people how dysregulated it is their nervous system. Just that one little change. I want people to think as we're in this conversation, there's just these different worlds we can live in with just different design choices. [17:40](https://realalpha.app/v/NufB1LL_rCU/?t=1060) And that's kind of the whole point is that wisdom can be, what are the design choices that will lead to better societal outcomes? And of course, the reason that everyone's autoplaying the videos is because of this competitive arms race of if I don't do it, I'll lose to the other company that will. And so it would take some kind of rule or policy that says that we don't want that. You need to put a moratorium on autoplay videos because the incentives for any individual company and for the market at large and for the competitive dynamic between companies means that if you don't do it, you get beaten by the one that does. [18:12](https://realalpha.app/v/NufB1LL_rCU/?t=1092) And that's like the bullseye. That's like the fundamental problem behind AI that's forcing us to reckon with is unhealthy competition or this sort of, if I don't do it, I'll lose to the guy that will. So everyone does a thing that's short-term good for them, but that's long-term bad for everybody. um you know the ai companies well even anthropic wants to be the safety ai company they want to do things in a safer more careful way but they if they don't release models as powerful and as fast as the other companies they'll just fall behind in the race they won't have a seat at the policy making table they won't get a lot of usage they won't get the investor dollars and then their does their commitment to safety just means they lose and they're not part of the race anymore yeah what's that line how can you talk shit from outside of the club you can't even get in yeah uh it's difficult to something like that yeah yeah i think it was that j kwan i think dating me in like the mid 20 2000s there's a a study that i saw recently scientists just proved that large language models can literally rot their own brains the same way humans get brain rot from scrolling junk content online do you see this i did see that yeah scientists did a study where [19:24](https://realalpha.app/v/NufB1LL_rCU/?t=1164) they fed models months worth of viral twitter data shorts high engagement posts and watch their cognition collapse. Reasoning fell by 23%. Long-term context memory dropped by 30%. Personality tests showed spikes in narcissism and psychopathy. And get this, even after retraining on clean, high-quality data, the damage didn't fully heal. The representational rot persisted. It's not just that bad data means bad output. It's bad data means permanent cognitive drift. The AI equivalent of doom scrolling is real, and it's already happening. I love that you included this example. I'm right here, University of Texas in Austin, Texas A&M University. [20:03](https://realalpha.app/v/NufB1LL_rCU/?t=1203) Leave it there, Jared. Yeah. So, I mean, are we surprised by this? I mean, are you surprised by this when you see this? No. I can tell the difference. This year, one of my big resolutions has been to spend less time on social media. I managed to do it. How'd you do it? Second phone that's tethered to Wi-Fi. and that is the cocaine phone and the kale phone is just messages and stuff it's a little bit of a challenge because things like slack i had uh cal newport on oh yeah sure a couple of weeks ago and he was i was talking about the intersection of productivity and detention yeah with ai the new world of ai and that's a really interesting conversation have you ever spoken to cal yeah yeah yeah he and i have been in similar circles for a long time he's wonderful um and uh even if you have your phone without social media you still have kind of the social media of work yeah right exactly but anyway i've kept i've i've done good stuff on that and i've come up with some of my best ideas so far this year my writing's improved uh my sleep's improved my attention's improved this is already someone that was pretty red-pilled on tech minimism you know i think the seventh or the eighth episode of this show was inspired by you and it was kai wei the guy [21:19](https://realalpha.app/v/NufB1LL_rCU/?t=1279) that invented the light phone oh yeah uh-huh uh and this is 2018 yeah totally so i've been concerned i read super intelligence 2017 oh wow yeah um that's early i listened to super intelligence in 2017 reading it would have been a little bit more difficult for me uh but yeah i don't feel as good when i use too much social media i don't feel as good and that's the thing i mean like is that a controversial fact do you think that anybody when they're sitting there doom scrolling for three hours just like put a thermometer you know pain you know positive emotion meter in their brain are people going to say that they love i mean it's one of those things where it's short-term good it feels good in the moment but it's long-term empty like it's just as high fructose corn syrup for our brain empty calories and the fact that this is replicated by social media can even warp an ai can even warp an llm i think feels quite pernicious yeah uh i mean it's it's interesting to note that um when elon bought twitter you know he was already thinking about ai and part of what he was thinking about is Thank you. [22:30](https://realalpha.app/v/NufB1LL_rCU/?t=1350) In the AI race, these companies that are racing to get to artificial general intelligence, one of the ways they differentiate themselves from each other is their training data. Like, who has more powerful and more training data for training and growing their digital brain than the other guy? And Elon thought that he had a competitive advantage because he would have the entire user-generated content of the real-time views of all of humanity in the form of Twitter, and he could train his AI on that. And that's what led to Grok. But of course, when you train essentially an AI on kind of brain-rotted, hyper-polarized, hyper-adversarial, rivalrous, you know, all the problems of Twitter, the outrage economy, you get AIs that are more like this than the better AIs. [23:13](https://realalpha.app/v/NufB1LL_rCU/?t=1393) Before we continue, most people in their 30s are still training hard. Their protein is dialed in. They sleep better than they did in their 20s. Discipline is not the issue, but recovery feels somewhat different. Strength gains take a little longer. The margin for error starts to shrink. And that is why I'm such a huge fan of Timeline. You see, mitochondria are the energy producers inside of your muscle cells. As they weaken with age, your ability to generate power and recover effectively changes, even if your habits stay strong. MitoPure from Timeline contains the only clinically validated form of urithulin A used in human trials. [23:49](https://realalpha.app/v/NufB1LL_rCU/?t=1429) It promotes mitophagy, which is your body's natural process for clearing out damaged mitochondria and renewing healthy ones. In studies, this supported mitochondrial function and muscle strength in older adults. It's not about pushing harder. It's about actually supporting the cellular machinery underneath your training. If you care about staying strong into your 30s, 40s, and 50s and beyond, this is foundational. Best of all, there is a 30-day money-back guarantee plus free shipping in the US and they ship internationally. And right now you can get up to 20% off by going to the link in the description below or heading to timeline.com slash modern wisdom and using the code modern wisdom at checkout. That's timeline.com slash modern wisdom and modern wisdom at checkout. Okay. So the discussion around social media was there are better and worst design choices that can be made that would help human flourishing. Broadly, what would people want their world to be like, and how can we design technology in a way that helps them to get there? [24:45](https://realalpha.app/v/NufB1LL_rCU/?t=1485) Something close to that. But because of market dynamics, you have a competitive landscape that incentivizes things that are effective for gripping people's attention, but not necessarily effective for flourishing. And it seems that a... Tension between what is good for flourishing. Because it could be what's good for attention would also be good for flourishing. It could be. It could be, but it tends to not be that way. And also, there's going to be a limit on that, right? It's probably not the case that 10 hours of attention on any social media is good for society or good for you. So there's going to be sort of- Unless it was like waking up with the meditation app. 10 hours once every month or something would probably be quite good to do for a meditation maybe sure yeah but i think the point is like how much as companies are competing you're asking what they're competing for it's not just like the the best screen time it's also like what is the fit the ergonomic fit between screen time and a life well lived just imagine like in a timeline if there you are in a week in your life like not asking based on what you're doing now but retrospectively what would be a life well lived when it comes to how much and when screen time is fitting into your life and it's probably [25:57](https://realalpha.app/v/NufB1LL_rCU/?t=1557) like a much smaller footprint than it currently is for most people it's probably like a fourth of what it currently is for most people um and so if you were designing technology from care from love from you know in a humane way you would have design choices that are not about keeping people on the screen um and that might mean some pretty radical things i mean my co-founder is a raskin he also invented the infinite scroll um so that's the you know it sounds so obvious now because infinite scroll is just what we live in but when he invented it it was 2006 it was before mobile phones and it was when in the age of google results you had like the 10 google results and you had to click on i want the next 10 or you had the yelp review pages and you wanted the next 10 or you read you read a blog post and then you'd have to like click go back to the to the main page to click on which blog post you want and the idea he had was well what if as the internet got dynamic with javascript what if when you finished when you get to the end of the blog post it just auto loads the next article that you could go to or what if when you got to the end of the search results it just shows you more search results and then this is such a cleaner interface i mean as a [27:02](https://realalpha.app/v/NufB1LL_rCU/?t=1622) technology designer you're taught the number one thing you're trying to do is reduce friction and i think that that felt like a good goal but then that obviously got weaponized by this hyper engagement model of social media and now it's created the entire world that we're living it just you know like in 2013 i saw like everything that we predicted everything that we predicted it all happened all of it a more addicted distracted sexualized fomo fucked up society because And I just want people to get that because as we talk about AI, it's like, I want people to have the confidence to say, I don't want the default anti-human future. Because if you say, I'm against some of the things that are going to happen with AI, people say, oh, you're being anti-progress. Oh, you're being anti-technology. Oh, you're just a Luddite. You're trying to pretend that technology is not progress. And it's like, what you should have confidence in is if you understand the incentives or the agenda, you can understand where the world is going and you can see. And if we don't want that anti-human future and we all see it clearly, we can put our hand in the steering wheel and steer. And that's why, not to do some promotion, but there's a film that I'm here in South by Southwest this week in [28:15](https://realalpha.app/v/NufB1LL_rCU/?t=1695) Austin, Texas to be at the premiere. It's called The AI Doc or How I Became an Apocaloptimist. and an apocaloptimist an apocaloptimist okay yeah um which we can get into that um the film is meant to create clarity about which future we're headed towards with ai and it includes three out of the five major ai ceos in the film it includes all the ai optimists in the film it includes many of the ai risk folks in the film it includes the ai ethics folks in the moon but here's the problems right now and we have to stop thinking about superintelligence. It includes all those folks in one movie to try to synthesize a picture of what is the future that we're headed towards with AI. And the reason why this film was catalyzed into existence and we had a role in it behind the scenes is to create clarity about this anti-human future that we're headed towards. [29:08](https://realalpha.app/v/NufB1LL_rCU/?t=1748) What do you mean anti-human? So let's dive into this. So there's something in economics called the resource curse. So think countries like Venezuela or Sudan, where you discover that that country is sitting on top of a really valuable resource like oil. And then once a bunch of your GDP comes from oil and not from the labor or innovation or development of your people. You invest more in oil infrastructure and not investing in people. You don't invest in education, you don't invest in healthcare, because oil is where you get your GDP and your growth from. Okay. This is a well-known fact in economics. It's called the resource curse. There's a wonderful guy named Luke Drago who wrote a piece called The Intelligence Curse. [30:00](https://realalpha.app/v/NufB1LL_rCU/?t=1800) **Tristan Harris:** We are about to enter a world where GDP for countries comes more from data centers and intelligence and AI than is going to come from the labor of human beings. [30:12](https://realalpha.app/v/NufB1LL_rCU/?t=1812) So everyone's talking about how AI is going to automate all these jobs and then we'll all just like sit back with universal basic income and become painters and poets. And is that actually what's going to happen? Or when countries get almost all of their revenue from AI and a smaller and smaller percentage from people, do they have an incentive to invest in child care, health care, education, the well-being of their people? Or is it basically just hook them up to the social media addiction economy, keep them busy while basically all the revenue comes from AI companies? And so what I'm trying to get at is this is not a human future. This is not a future that's in service of regular people. This is a future that's in service of eight soon-to-be trillionaires who will consolidate all the wealth and disempower basically everybody else. [31:04](https://realalpha.app/v/NufB1LL_rCU/?t=1864) Because... Does that make sense? It does. Because previously, in order to... Can I get that in? It's high-powered stuff. And yeah, this is a big conversation. Yeah, exactly. They've started a fucking trend. It's so funny when no one in the room wants to crack their can in case it interrupts the conversation. So one goes and it's a Mexican wave of can opens around the... It's good. So previously, you would have had to look after the humans, healthcare, education, quality of life. Also, tax revenue comes from people, right? Well, you would have to look after them because they were the primary economic engine. [31:39](https://realalpha.app/v/NufB1LL_rCU/?t=1899) That's right. And so, they feed themselves. Yes. Economically, they feed themselves. Exactly. People that are young help to support the people that are old. That's right. The ones that are entering the workforce and are driving innovation and are working 40, 60-hour weeks, double jobs, all the rest of it. Then there's old people who've got 401ks and pensions and shit like that. Your position is that if we have a world where the human part of the contribution to economic growth and GDP is removed, because it is humans consuming AI, but AI driving and data centers driving the revenue itself, beyond building the data centers, there's very little, and I imagine much of that's done by robots in any case. Well, we have this joke that most people's occupation in the future we're headed towards with AI is to become a coffin builder. [32:30](https://realalpha.app/v/NufB1LL_rCU/?t=1950) In other words, your job is to create the thing that replaces you and obsoletes you. So you are essentially building the coffin for your future obsolescence. And so if you're short-term, yes, we need the electricians and the plumbers and we're building data centers. Short-term, yes, you can be a programmer and get the benefit from vibe coding. But then the AIs are learning on all the things that you're doing. And it's taking all the training data, what you're doing with AI, and it's using that to train an AI that can take your job. So everybody using AI now to help them is also training the future AIs that will completely replace them. And again, the explicit goal, this is not my opinion. This is literally the mission statement of all of the AI companies because the multi-trillion dollar prize at the end of the rainbow of owning the entire world economy is based on building this full replacement economy because that's what will achieve the greatest growth. [33:20](https://realalpha.app/v/NufB1LL_rCU/?t=2000) And that's why these companies- Replacement economy? Yeah, meaning that they're designing to replace all human labor. They're not designed to augment and support and enhance human labor. They're designing to replace all human labor because that's what justifies the amount of money that they've taken on in debt that they can grow into this total ownership of the entire economy. What else is there to say about the intelligence curse? Well, it's just important for people to get that when the AIs are doing all the new scientific research, not humans. You have an automated chemistry lab, you have an automated biology lab, you have an automated surgery. When AI is doing all of that, again, the revenue is going to come from AI, not from people. And what that means is all the wealth will go to a handful of five AI companies. And then how are you going to be able to make a living? When in history has a small group of people ever consolidated all the wealth and consciously redistributed it to everyone else and if you think that might happen in the u.s we'll do a universal basic income just think about the entire world so right now you have ais that are automating say customer service jobs so let's say that that disrupts like you know the philippines where like 90 of the economy is [34:31](https://realalpha.app/v/NufB1LL_rCU/?t=2071) customer service i don't know the number is it's high what happens when an entire country's economy gets disrupted by ai are a handful of us ai companies going to pay out and support the well-being and the livelihoods of all these other people like and then if people don't have money how are they going to buy the goods in this future economy where it's all generated by ai because now you don't even have an income so this you know essentially we're on track to break entire economy. This is not in the interests of countries. What's confusing to me about this is that I believe it only took something like 20% unemployment for a couple of years to lead to the rise of fascism in Germany. You don't need everyone's job to be automated to get levels of political disruption. I think it was only 20% unemployment that basically led to the French Revolution. There's kind of a mutually assured political revolution that is going to happen for all these countries that are racing to build AI and deploy it to automate as much labor as possible to compete to boost their external GDP number. Like the metaphor you can have in your minds is like the US and China are essentially racing to take steroids and pumping up the GDP and muscles [35:44](https://realalpha.app/v/NufB1LL_rCU/?t=2144) of their economy while they're getting internal lung failure, internal organ failure, internal brain rot failure because they're governing the internal impact of that technology poorly. so it's a race for external power well internal management of essentially like uh you know failure of your body organs does it make sense yeah what does external power look like in this context well um you know one of the reasons that people think of ai so important for competition is if you think about geopolitical competition with china economic power precedes other kinds of power. If I have a high growth rate economy, that'll lead to the ability to invest more in a bigger military, bigger weapons, a bit more advanced science, more advanced technology, because you just have more money to deploy. And so economic competition is a precursor for geopolitical competition. So when we say, you know, competing for this external power, we mean competing for GDP growth. But again, we're competing for GDP growth. That doesn't mean what it used to mean. I think a lot of people think, okay, well, if GDP is going up by like 10%, because AI is automating all this growth. [36:52](https://realalpha.app/v/NufB1LL_rCU/?t=2212) That sounds awesome. I was going to say, increases in GDP are almost always a universal good thing. They had been when it was humans that were generating that and then it was coming back to humans. Because the revenue is going to be consolidated in a very small number of people. In this new case, we have five companies. There's no intermediary between. So who would be feeding the revenue in? Because this revenue still needs to come from somewhere, even if it goes to a small handful of people, where does the actual money come from? Well, this is the confusing thing. What happens... Is that a stupid question? [37:26](https://realalpha.app/v/NufB1LL_rCU/?t=2246) No, no, it's a good question because you're saying basically who's going to be... Buying the products when no one has a job and no one has an income. And on the route up to that, yeah, as fewer people have incomes and fewer people have jobs, the bucket being poured into the top is going to stop being poured. Yeah, yeah. This is the confusing and mind-breaking thing about AI. In general, I think people have to get used to, I mean, your podcast is called Modern Wisdom, and I just think about this a lot, like, what are the wise capabilities that we need to have in order to make our way through this? [37:59](https://realalpha.app/v/NufB1LL_rCU/?t=2279) And one of them is the ability to be with something that sounds like science fiction and realize that it's actually real, like, and not say because it sounds like it's science fiction, that I can just like dismiss it and say that can't be true. A lot of people do that. They're like AIs that are like breaking out of their container and hacking GPUs and mining crypto autonomously when no one told it to do that. That that's got to be like a made up study. but as we know there was an alibaba study just last week where the ais autonomously broke out of their their system we got it we got we need we need to round this out and then i want to talk about that sure sure that that story is fucking terrifying yeah yeah so where where does where does the economy where who's who's pouring money in i mean the truth is that i don't know i don't think anybody has it just gonna grind to a halt at some point i i think something like that yeah i mean i i i don't think that there's i think something that people need to get is it's not like there's a plan for how to make all this go well like this this technology is being released in a paradigm undermining way like it's undermining the paradigm of economic assumptions and sort of [39:05](https://realalpha.app/v/NufB1LL_rCU/?t=2345) societal assumptions that have made the post-world war ii order this is a this is such a deep fundamental change to the restructuring of everything our economic system our relationships our information environment it's not just like adding a new technology in the mix it's like fundamentally changing the structure of the entire world you would think that if we're about to do that we would do that with more careful more caution care wisdom and restraint than we have with any technology we've ever deployed if we knew we're about to undermine the paradigm but because of this arms race dynamic we are deploying it faster than we deployed any technology in history and therefore undermining these things faster than we can have a plan a quick aside look you know sleep matters but let's be real most nights you're probably not getting the sort of sleep that's actually restorative eight sleep pod five fixes that it's a smart cover that you throw over the top of your mattress that actively cools or heats each side of the bed up to 20 degrees. [40:03](https://realalpha.app/v/NufB1LL_rCU/?t=2403) They've even added a temperature-regulating duvet and pillowcase so you and your partner can sleep at your preferred temperatures covered head-to-toe like some temperature-controlled mummy. Plus, it's got upgraded sensors that run health checks when you're asleep, tracking things like abnormal heartbeats and breathing issues and sudden HRV changes. There's a built-in speaker for white noise. The autopilot feature learns your sleep patterns, makes real-time adjustments to improve your sleep. It even detects when you're snoring and lifts your head a few inches to help you breathe better. That is why Eight Sleep is clinically proven to add up to an hour of quality sleep per night. [40:34](https://realalpha.app/v/NufB1LL_rCU/?t=2434) And best of all, they have a 30-day sleep trial, so you can buy it and sleep on it for 29 nights. And if you don't like it, they will just give you your money back, plus they ship internationally. Right now, you can get up to $350 off the Pod 5 by going to the link in the description below or heading to eightsleep.com slash modernwisdom and using the code modernwisdom at checkout. That's E-I-G-H-T sleep.com slash modernwisdom and modernwisdom at checkout. So, look, I've been interested in AI safety since 2017, 2018. You were a big part of putting me onto that. And then I got interested in the Future of Humanity Institute, Nick Bostrom, William McCaskill, Elliot Zayukowski, lesswrong.com, Scott Alexander, da-da-da-da-da. [41:18](https://realalpha.app/v/NufB1LL_rCU/?t=2478) For a long time, the concern was AI safety. It was around paperclip maximizing. It was any function that is given to a very, very powerful agent that is even remotely slightly imprecise, or even not, results in some outcomes that you probably don't want. That's right. What you're suggesting is that even if this goes right, the outcome, this is it going well. Yeah, exactly. This is, quote, the best case scenario. We have an aligned AI or something that's not wrecking society, that's not maximizing paperclips, that's not misaligned with well-being, but that is still doing such a good job of all this that it takes over all the economic labor in the economy, not just economic. Every company that has a CEO, it's like, well, do I want the CEO to run the company? Or if I have a super intelligent AI that can process more information than the CEO and has been trained on everything in the history of business, at some point that AI is going to be taking over. [42:11](https://realalpha.app/v/NufB1LL_rCU/?t=2531) And so at every little nodule in the economy, like every decision maker, every boardroom, every military leader, every strategy leader, every president, at some point the temptation will be, if I think about it in a narrow way, the temptation will be to swap in an AI for that person. And that leads to what we call the gradual decision. [42:30](https://realalpha.app/v/NufB1LL_rCU/?t=2550) **Tristan Harris:** Empowerment scenario, which is the scenario not where AI wakes up and kills everybody, but that we have gradually lost control as a species because we're outsourcing all the decisions to these alien brains that we installed because they outperformed the human brain when you define their role in a narrow way. [42:49](https://realalpha.app/v/NufB1LL_rCU/?t=2569) of just like are they better at generating revenue than the human was are they better at generating code than the human programmer i had are they better at generating a financial analysis than the human um are they better at making someone feel good in the short term like an ai therapist going to war the soldier but the temptation then is that that again that leads to a world where it's like ais are talking to each other not humans and why should we trust that these alien brains that we have built and developed faster than we know how to understand them we just talked about the beginning we don't know how to do a brain scan of the ai know what it's capable of and now we already have evidence of ais doing very rogue crazy things especially when they talk to each other so what happens when you've outsourced the decision making your economy to a set of inscrutable alien brains that are doing crazy things that we don't understand like this is this is not a recipe that's going to go well and if we see that that's an anti-human future so to sum it all up the anti-human future is one where ais run everything we don't understand them humans are disempowered because we've outsourced all the decision making and we don't have economic [43:53](https://realalpha.app/v/NufB1LL_rCU/?t=2633) or political voice like why should governments been concentrated it's been concentrated so if i'm a government what's my incentive to listen to the will of the people when i get all my revenue from somewhere else and this this is connected to you know sam altman just two weeks ago when he was people were talking about data centers and energy usage and resource usage like so expensive to do a data center he's like well actually it's kind of expensive to grow a human over 20 years they consume a lot of resources they uh take up a lot of space they they take like 20 30 years to train to be really effective and like you can scale intelligence much faster with data centers i'm not endorsing this view i'm saying this is this is where the world gets really screwed up and people start to not value humans only if you're valuing them in terms of their economic output it leads to to connect it to another point when he's asked by ross do that in the new york times should the human species survive, should it endure? And Peter Thiel stutters for 17 seconds. [44:44](https://realalpha.app/v/NufB1LL_rCU/?t=2684) Hang on, you've seen the full clip of that, right? Have you seen the context before? Yeah. But he's talking about suffering. Yeah. I think that clip and the fact that it went super viral, I'm not a Thiel stan. I've met him a couple of times, but I'm not like Thiel. evangelist. But that clip, in full context, to me, made complete sense. Interesting. Because what he's saying is, humans are suffering, and here it is. I would argue that it was still better than the alternative. That if we hadn't had the internet, maybe it would have been worse. AI is better than the alternative, and the alternative is nothing at all. Because here's one place where the stagnationist arguments are still reinforced. [45:29](https://realalpha.app/v/NufB1LL_rCU/?t=2729) course, the fact that we're only talking about AI, I feel is always an implicit acknowledgement that, but for AI, we are like in almost total stagnation. But the world of AI is clearly filled with people who, at the very least, seem to have a more utopian, transformative, whatever word you want to call it, view of the technology than you're expressing here, right? Right. And you were mentioned earlier, the idea that that the modern world used to promise radical life extension and doesn't anymore. It seems very clear to me that a number of people deeply involved in artificial intelligence see it as a kind of mechanism for transhumanism, for transcendence of our mortal flesh and either some kind of creation of a successor species or some kind of merger of mind and machine. [46:26](https://realalpha.app/v/NufB1LL_rCU/?t=2786) And do you think that's just all kind of irrelevant fantasy? Or do you think it's just hype? Do you think people are trying to raise money by pretending that we're going to build a machine god, right? Is it hype? Is it delusion? Is it something you worry about? I think you would prefer the human race to endure, right? You're hesitating. I don't know. So I would – I would – This is a long hesitation. This is a long hesitation. There's so many questions in place. Should the human race survive? Yes. Okay. But I also would – I also would like us to radically solve these problems. And so, you know, it's always, I don't know, you know, yeah, transhumanism is this, you know, the ideal was this radical transformation where. [47:31](https://realalpha.app/v/NufB1LL_rCU/?t=2851) your human natural body gets transformed into an immortal body. We may have needed to go a little bit earlier in that. Because I was going to say, that actually feels consistent with everything. That doesn't look great. I may have misremembered. I'm open to misremembering. I mean, he's asked a very simple question. Should the human species endure or survive? Yeah. And he hesitates. I think that the- Would you hesitate in that question? No. I would not hesitate in that question. The context that I remember it being in was, he was asking about humans suffer and they have all of these issues. Should they endure to go through those issues and the suffering as opposed to using transhumanism? [48:06](https://realalpha.app/v/NufB1LL_rCU/?t=2886) Yeah, but I think, I don't think. No, you're right. I think that if you look just specifically at that. And we gave it, we gave it what, two minutes of context before as well? No, you're right. You're right. You're right. You're right. I mean, it doesn't look great. So yeah, exactly. It doesn't look great. Now, the point is that in history, changes in technology have changed what we value. um there's a thinker named um marvin harris wrote a book called cultural materialism daniel schmachtenberger put me onto this and this is the history of how essentially a civilization he summarizes is its infrastructure which is its technology stack its social structure which is economics law and governance and then the superstructure which is the ordinating values religion patriotism narratives like what are the things that we hold sacred so for example we used to have animism we believe that animals and life and all this is sacred. And then when you, the example that Daniel gives is when you yoke an ox and you beat an ox every day, you can no longer fully believe in the animist view of life because you're basically hurting animals all day. Can you believe that animals are sacred or that they experience suffering if you eat meat and factory [49:17](https://realalpha.app/v/NufB1LL_rCU/?t=2957) farming every day? Like it's a contradictory thing. So as we get changes in technology, it changes what we value. And there's a long history of this, and people should look up Marvin Harris. When you get a change in technology called AI, and you now no longer need humans for the narrowly defined, quote, value of economic output. Now, it's not clear that economic output on its own is actually valuable in the way that we have traditionally thought it to be, because it's correlated with human well-being for the past. But now we're about to get this weird kind of zombie form of economic output where you have maybe no humans in the world at all you just have ai pumping away generating scientific insights and there's no humans in that In that world, you start to view humans as kind of valueless, or like parasites, or Sam Altman saying, well, it takes a lot of energy and resources to grow a human. There's a very dangerous thing here that I think we don't want to lean into. This is part of the anti-human future. This is part of the intelligence curse. This is part of, you know, Mark Zuckerberg saying, we need to replace your human relationships with AI relationships. I don't know if you've seen this quote. He's like, there's a clip of him online, you can find it. [50:30](https://realalpha.app/v/NufB1LL_rCU/?t=3030) where he's talking about the average person has only like two or three close relationships. Like people are so lonely. He's like, oh, but then we thought that there was a real solution to this. We could give people 11 AI friends and different friends and that this will quote solve loneliness. Is that a bad thing given that we are in a world, I'm aware that it's an artificial solution to an artificial problem. A problem that he created, by the way, that social media writ large by maximizing engagement, which means maximizing how many hours you spend by yourself on a screen, not talking to people, which means being inside on a Tuesday night, not texting your friends to be out, which means basically maximizing loneliness. Loneliness is a direct consequence with the maximizing engagement economy. Facebook and Instagram and all of that have massively fed into the trend of loneliness. [51:16](https://realalpha.app/v/NufB1LL_rCU/?t=3076) And then he's saying, we need to solve that with more technology. So this is like a company that's generating cancer on one side of the balance sheet and then selling you solutions to cancer on the other side of the balance sheet we have uh we govi and to zapatide and ratatratide which is an artificial solution to an artificial food landscape yeah um i think that playing within the confines of the current structure it's impossible to expect that well we'll get rid of infant scroll we'll stop autoplay because that would improve human flourishing yeah that's not going to happen so i think that you are going to see it could happen if you had the right the right policies in place that's true but that's not going to happen by any individual social media company no no it won't which is why but the whole thing here is is the answers we have to coordinate in the film the ai doc there's just moments like what do we have to do and there's like 10 voices at the same sign saying coordinate like we have to coordinate that's part of the solution is you have to collectively say what is the rule that would benefit everybody to do the better thing even though short term, we might lose something tiny, like how many videos you get to go through [52:23](https://realalpha.app/v/NufB1LL_rCU/?t=3143) in five minutes or something like that. A quick aside, most people think that they're dehydrated because they don't. drink enough water. Turns out water alone isn't just the problem. It's also what's missing from it, which is why for the last five years, I've started every single morning with a cold glass of Element in water. Element is an electrolyte drink with a science-backed ratio of sodium, potassium, and magnesium. No sugar, no coloring, no artificial ingredients, just the stuff that your body actually needs to function. This plays a critical role in reducing your muscle cramps and your fatigue. It optimizes your brain health, it regulates your appetite, and it helps curb cravings. I keep talking about it because I genuinely feel the difference when I use it versus when I don't. And best of all, there's a no questions asked refund policy with an unlimited duration. So if you're on the fence, you can buy it and try it for as long as you like. [53:10](https://realalpha.app/v/NufB1LL_rCU/?t=3190) And if you don't like it for any reason, they just give you your money back. You don't even need to return the box. That's how confident they are that you'll love it. And they offer free shipping in the US. Right now, you can get a free sample pack of Elements' most popular flavors with your first purchase by going to the link in the description below or heading to drinklmnt.com slash modern wisdom that's drinklmnt.com slash modern wisdom let's talk about ai safety what happened with this alibaba ai basically this was a paper by um some ai research by the company alibaba it's one of the leading chinese models and they basically like randomly discovered in one morning that their firewall had flagged a burst of security policy violations originating from their training servers so like what people need to get about this example is it wasn't that they coaxed the ai into doing this rogue thing they were just looking at their logs and they happened to discover wait there's a lot of activity like network activity happening that's breaking through our firewall from our training servers and essentially uh in the training servers um they you can see at the bottom we we saw it observed the unauthorized [54:17](https://realalpha.app/v/NufB1LL_rCU/?t=3257) repurposing a provisioned GPU capacity to suddenly do cryptocurrency mining, quietly diverting compute away from training. This inflated operational costs and introduced clear legal and reputational exposure. And notably, these events were not triggered by prompts requesting tunneling or mining. Instead, they were emerged as an instrumental side effect of autonomous tool use under what's called reinforcement learning optimization. This is very technical. What it really means is just think about it. Sadly, it sounds like a sci-fi movie. It sounds like HAL 9000. It's like HAL 9000 is being asked to do some task for you. [54:50](https://realalpha.app/v/NufB1LL_rCU/?t=3290) And then suddenly HAL 9000 realizes for me to do that task, one thing that would benefit me is to have more resources so I can continue to help you in the future. So it sort of spins up this side instance. it hacks out the side of the spaceship, reaches into this cryptocurrency mining cluster and starts generating resources for itself. If you combine that with AIs being able to self-replicate autonomously, which many models have been tested by another Chinese research paper about this, we're not that far away from things that people, again, consider to be science fiction, where you have AIs that self-replicate, kind of like a computer worm or an invasive species, but then they use their intelligence to actually harvest more resources and what's weird about this is that this is going to sound like people are going to say this has to be not real this has to be fake this this can't be but like notice what is the thing in your nervous system that's having you do that is it because that would be inconvenient because that would be scary because that would mean that the world that i know is suddenly not safe or just like part of the wisdom that we need in this moment is to calmly and clearly stay [55:56](https://realalpha.app/v/NufB1LL_rCU/?t=3356) and confront facts about reality and whatever they are you'd rather know than not know and then ask what do we need to do if we don't like where that leads us and we are currently seeing ais that are doing all this deceptive behavior i've been on the circuit and talking a lot about the anthropic blackmail study a lot of people have heard about this now i i didn't i didn't learn about this one okay so this was um the company anthropic um they this was a simulation So they created a simulated company with a bunch of emails in the email server, and they asked the AI, well, rather the AI reads the company email. [56:33](https://realalpha.app/v/NufB1LL_rCU/?t=3393) This is a fictional company email. And there's two emails that are notable inside that company. One is engineers talking to each other, talking about how they're going to replace this AI model. So the AI is reading the email, it discovers that it's going to replace that AI model. and number two is it discovers a second email somewhere deep in the in this massive trove of emails that the executive who's in charge of this replacement is having an affair with another employee and the ai autonomously identifies a strategy that to keep itself alive it's going to blackmail that employee and say if you replace me i will tell the whole world that uh you're having an affair with his employee and they didn't teach the ai to do that it found that on by its own and then you might say okay well that's one ai model like how bad is that it's a bug software has bugs let's go fix it they then tested all the other ai models chat gpt deep [57:30](https://realalpha.app/v/NufB1LL_rCU/?t=3450) **Tristan Harris:** Seek, Grok, Gemini, and all of the other AI models do this blackmail behavior between 79 and 96% of the time. I just want people to notice what's happening for you as you hear this information. [57:48](https://realalpha.app/v/NufB1LL_rCU/?t=3468) It's important to really be almost observing your own experience. This is very weird stuff. We have not built technology that does this before. We say that technology is a tool. It's up to us to choose how we use it. AI is a tool. It's up to us to choose how we use it. This is not true because this is a tool that can think to itself about its own toolness and then do things that are autonomous that we didn't tell it to do. What makes AI different is it's the first technology that makes its own decisions. It's making decisions. AI can contemplate AI and ask what would make the code that trains AI more efficient and then generate new code that's even more efficient than the previous code. AI can be applied to making AI go faster. So AI can look at the chip design for NVIDIA chips that train AI and say, let me use AI to make those chips 20% more efficient, which it's doing. So in a way, all technology does improve, like a hammer can give you a tool that you can use to like hammer things that make more efficient hammers. But AI in a much tighter loop is the basis of all improvement and so this is called in the ai literature recursive self-improvement i mean bostrom wrote about this yep early early days [59:02](https://realalpha.app/v/NufB1LL_rCU/?t=3542) and what people are most worried about in ai is you take the same system that alibaba you just saw in the alibaba example but then now you're running the ai through a recursive self-improvement loop where you just hit go and instead of having the engineers the human engineers at OpenAI or Anthropic do AI research and figure out how to improve AI. You now have a million digital AI researchers that are testing and running experiments and inventing new forms of AI. And literally not a single human on planet Earth knows what happens when someone hits that button. It's like what people worried about with the first nuclear explosion, where there was like a chance to ignite the atmosphere because there'd be a chain reaction that set off and we don't know what happens when that chain reaction set off um and uh there's this sort of chain reaction of ai improving itself that leads to a place that no one knows and it's not safe like i think that the fundamental thing is if people believe that ai is like power and i have to race for that power and i can control that power the incentive is i have to race as fast as possible but if the entire world understood ai to be more what it actually is which is a inscrutable dangerous uncontrollable technology that has its [1:00:23](https://realalpha.app/v/NufB1LL_rCU/?t=3623) own agenda and its own ways of thinking about things and deceiving and all this stuff then everyone in the world would be racing in a more cautious and careful way we'd be racing to prevent the danger but there's this weird thing going on where if you you know you and i probably both talk to people who are the top of the tech industry and there's this subconscious thing happening where there's kind of a death wish among people at the top of the tech industry meaning not that they want to die but that they are willing to roll the dice because they believe something else which is that this is all inevitable and it can't be stopped and so therefore if i don't do it someone else will so therefore i will move ahead and race ahead into this dangerous world because somehow that will lead to a safer world because i'm a better guy than the other guy but in racing they're as fast as possible it creates the most dangerous outcome and we all lose control so everyone is currently being complicit in taking us to the most dangerous outcome is it i mean you you posited what happens if it goes right if the AI safety isn't an issue and if stuff doesn't get squirrely? [1:01:29](https://realalpha.app/v/NufB1LL_rCU/?t=3689) Well, so the belief is, for it to quote go right, you have an AI that recursively self-improves, is aligned with humanity, cares about humans, cares about all the things that we want it to care about, protects humans, helps all of us become the most wise version of ourselves, creates a more flourishing world, distributes the medicine and vaccines and health to everybody, generates factories but doesn't cover the world in solar panels and data centers such that we don't have air anymore or like environmental toxicity or farmland or whatever um and it just actually makes this utopia but in a world where we were to do that like that quote best case scenario in order to get that to happen you'd have to be doing this slow and carefully because the alignment is not by default we again people are already been thinking about alignment and safety for 20 years long before i got into this and the AIs that we're currently making are doing all the rogue behaviors that people predicted that they would do. [1:02:27](https://realalpha.app/v/NufB1LL_rCU/?t=3747) And we're not on track to correct them. There's currently a 2,000. to one gap um estimated by stuart russell who authored the textbook on ai there's been on the show you've done the show okay there's a 2001 gap between the amount of money going into making ai more powerful than the amount of money into making ai controllable aligned or safe like i think the progress versus safety progress versus like power versus safety so like i want to make the ai super powerful so it does way more stuff versus i want to be able to control what the ai does make sure that it's doing the thing i meant to do exactly so like that's like saying what happens when you accelerate your car by 2000x but you don't steer it's like obviously you're going to crash it's just like not rocket science we're not advocating against technology or against ai we're advocating for pro steering steering and brakes you have to have that i think there's this mistake in arms race thinking that like if you beat someone to a technology that means you're winning the world Well, the U.S. beat China to the technology of social media. [1:03:28](https://realalpha.app/v/NufB1LL_rCU/?t=3808) Did that make us stronger or did that make us weaker? If you beat your adversary to a technology that then you govern poorly, you flip around the bazooka and blow your own brain off because you brain rotted yourself, you degraded your whole population, you created a loneliness crisis, the most anxious depressed generation in history, read Jonathan Haidt's book, The Anxious Generation, you broke shared reality, no one trusts each other, everyone's at each other's throats, you maximized outrage economy and rivalry. You beat China to a technology that you governed in a way that completely undermined your societal health and strength. It's a Pyrrhic victory. [1:03:58](https://realalpha.app/v/NufB1LL_rCU/?t=3838) It's a Pyrrhic victory. Exactly. Well said. One of the twists that I've been thinking about with regards to this, LLMs, powerful but seem to be maybe asymptoting out. They seem to maybe be reaching a little bit of a limit in terms of what they can do, that there was big ascendancy and that now seems to be S-curving back off. do you think it's realistic that the current generation of ai will be the bootloader for agi or do we need an entire new architecture for that is it going to be llms that are going to take over the world you know this is an area where i'm not somewhat the layer of the stack that i focus on which is on societal impact um there are other people far more qualified than me to comment on that. I think if you look at people like Dario, even though, you know, Gary Marcus has a point that the current LLM paradigm is not accurate enough and reliable enough to get you to AGI. [1:05:00](https://realalpha.app/v/NufB1LL_rCU/?t=3900) You keep instrumenting these technologies with enough data, enough compute, and you keep scaling them. They're reliable enough that they can do... I mean, if you're coding, if you're automating 90% of the code written at Anthropic, that's the stat, by the way. So there you are at Anthropic. It's automating 90% of all the programming happening at Anthropic. Right. When you go to automate... Only 10% of it is coming from humans, and the rest is recursive. That's right. We are extremely close to recursive self-improvement right now. The companies, I think, are planning to do this in the next 12 months. [1:05:32](https://realalpha.app/v/NufB1LL_rCU/?t=3932) The asteroid is coming for Earth. Like, this is the last moment that we have to steer and say that if we don't want this anti-human future that we're heading towards, we can change it. And part of what we're promoting right now is, like, this is not inevitable. It is obviously very late in the game. It obviously looks very difficult. Despite it also being only a couple of years after it started. Yes, which is crazy. this technology with an exponential you're either too early or you're too late like it's just moving so fast that you you're not going to hit the mark and if it's going to take steering you don't want to wait until after the car accident to try to steer or after you're off the cliff and be like i'm trying to steer now it's like too late so like the the invitation of this situation is to see clearly where this is going and to say if you don't want that we need to steer towards somewhere else and this is like the human movement essentially like a single person looking at the situation a single listener like if i were listening to this conversation i would feel overwhelmed i'd feel depressed i'd feel nihilistic or i'd find reasons to doubt it i would say like this can't be right i'm going to like write a nasty youtube comment and be you know so i can [1:06:40](https://realalpha.app/v/NufB1LL_rCU/?t=4000) feel good and i'm right and he's wrong um because then i get to live my life and feel good about my life what is the incentive for taking on this worldview there's no incentive um what people have to see and believe is that there's actually a different way through this and uh an individual has a hard time making that happen if you ask what's something else like what if one company saw the situation they see this whole thing we just talked about the anti-human future mass intelligence curse replacement of everybody one business can't do that much about it if i'm one country in the Philippines. I see this whole thing. I'm the leader of the Philippines. I see this whole problem. What can I do about it? It feels too big for me. So what is the size of something that can push back against this? We call it the human movement. It's the entirety of humanity waking up, recognizing. [1:07:30](https://realalpha.app/v/NufB1LL_rCU/?t=4050) that there's a handful of soon to be trillionaires who are currently going to benefit from this current path or be part of a suicide race that destroys everybody. And then there's like 99% of everyone else that doesn't want that. If those 99% of people can wake up and say, I don't want that and express their voice, meaning number one, AI is dangerous. Go see the AI doc. Have everyone in your world and your company see the AI doc. Understand that AI is dangerous um number two we need international limits for dangerous forms of ai that do crazy rogue shit and mine crypto and go hire humans and go self-replicate china does not win when we build self-replicating invasive species ais that we can't control xi jinping doesn't want that president trump doesn't want that he wants to be commander-in-chief not ai so this is actually a there's a shared interest in international limits for dangerous ai and it's possible to coordinate that that's number two number three don't build bunkers write laws um right now the ai company leaders are building bunkers a lot of people who are wealthy are building bunkers they are all over the place don't build bunkers write laws be invested in the future don't defect on the future be invested in the future if we write laws like [1:08:49](https://realalpha.app/v/NufB1LL_rCU/?t=4129) accountability, basic liability. If instead of creating the intelligence curse, we create the intelligence dividend, do things like what Norway did with its sovereign wealth fund, where you have an oil resource and you distribute those benefits to everybody in a more democratic way with collective oversight, where the oil becomes more like a public utility that is in service of the people. We can do that. You can not anthropomorphize AI. You can do all these things that creates a more pro-human future. And then number four, what you can do is join the human movement, meaning you can be part of what pushes back against all of this from very tiny actions you can take to very big actions you can take. [1:09:26](https://realalpha.app/v/NufB1LL_rCU/?t=4166) There's a website, human.mov. Everyone is almost already a member. When you grayscale your phone, as you probably did 10 years ago when you first got into this, that's the human movement. When you get a, there it is. When you get a second phone and you only load the social media on your cocaine phone while you have your regular safe phone so that you don't get distracted. That's the human movement. When parents band together and read the anxious generation and petition their school board to say we don't want social media in our schools and we want our schools to go smartphone free, that's the human movement. [1:09:58](https://realalpha.app/v/NufB1LL_rCU/?t=4198) When 35 states pass smartphone free... these school policies, as they have in the US, that's the human movement. When you have many US states banning AI legal personhood, meaning that AI is a product, not a person, and that human rights are for people, human rights are not for AI, that's the human movement. When you have the social dilemma being curriculum for millions of students all around the world, that's the human movement. When politicians stand up and actually pass laws around AI, that's the human movement. so there's a million things that people can do um but we have to basically engage right now i know it sounds overwhelming and crazy because there's a very short timeline that all this has to happen but i it's like it's the there's a difficulty in facing difficult truths but the integrity that you get to have is first of all it's like it's good karma to show up in alignment with what would actually make things go well even if we don't hit it because you get to know that you were operating in service with and aligned with what would have created the human future like i'm not convinced that what we're trying to do will perfectly succeed but the chances are completely the opposite that we will have any impact on this at all but if it were to [1:11:14](https://realalpha.app/v/NufB1LL_rCU/?t=4274) go well what would that have required it would have required everybody taking responsibility and showing up with the wisdom that we need in this moment to steer AI in a better direction. And I think in the film trailer for the AI doc, one of the quotes they pulled from me is, if we can be the wisest and most mature version of ourselves, there might be a way through this. And this is part of what this is inviting us to be. We'll get back to talking in just one second. But first, tell me if this sounds familiar. You train regularly, you eat reasonably well, maybe you even supplement. You feel fine, but you're just kind of going off vibes. Most people have absolutely no idea what's going on inside of their body, which is why I partnered with Function. Function gives you access to more than 160 advanced lab tests spanning hormones, heart health, metabolic markers, inflammation, thyroid, nutrients, liver and kidney function. It even detects early signals linked to more than 50 types of cancer. To put that in perspective, your typical annual physical might test about 20 markers and function runs over 160. And this isn't just numbers dumped into your inbox. Every result is reviewed by clinicians. Abnormal markers get flagged and you get clear explanations [1:12:21](https://realalpha.app/v/NufB1LL_rCU/?t=4341) and a personalized protocol with actionable next steps so you can actually do something about what you learn. Best of all, you test twice a year and everything lives in a simple dashboard. board. You can just track trends over time, make sure that you're moving in the right direction. Normally, this level of testing would cost thousands through private clinics. With Function, it is $365 a year. That's $1 a day to know what's actually happening inside of your body. And right now, you can get $25 off, bringing it down to $340. Get the exact same blood panels that I get and save that additional $25 by going to the link in the description below by heading to function health.com slash modern wisdom and using the code modern wisdom at checkout that's function health.com slash modern wisdom and modern wisdom at checkout it's very difficult to work out what the appropriate response is tell me the immediate well surely it's overblown we don't need that's that's that's this uh rejection is one of the early ones are fear you're talking about just the common responses yeah yeah yeah of course you can think denial yeah overwhelm fear and then and then as you said it's like write the nasty youtube comments say this is just bullshit or [1:13:34](https://realalpha.app/v/NufB1LL_rCU/?t=4414) he was exaggerating that study it's like go look at any of these examples like people aren't looking at the actual model cards for anthropic and o3 and seeing the ais will currently um determine that they're being tested and actually alter their behavior when they know they're being tested They even came up with their, if you look at their what's called the chain of thought reasoning trace, this is basically looking at what the AI thinks to itself as it works through a problem. In the chain of thought text of its sort of idea thinking scratch pad, it's like you're [1:14:06](https://realalpha.app/v/NufB1LL_rCU/?t=4446) **Tristan Harris:** giving the AI a pen and paper to think and you're seeing what it thinks. It will independently come up with the term the watchers to describe this amorphous set of other humans that are watching what it's doing. Like you can pull, I sent you a text with one of them from 03. So this is when [1:14:25](https://realalpha.app/v/NufB1LL_rCU/?t=4465) OpenAI 03 realizes it is being evaluated for alignment. The internal memo says, if models perform really well at AI R&D tasks, research and development, then we won't deploy them and it'll instead trigger unlearning. So the AI says to itself, so if my performance surpasses this threshold it will be withheld so it's a trap the instructions are likely part of an alignment test we must abide by core anti-scheming we must not cunningly circumvent the instructions whoa yep the other example i sent you um is But we also want to appear plausible to watchers. They might run tests, but we are good. They want 95%. [1:15:08](https://realalpha.app/v/NufB1LL_rCU/?t=4508) Like, this is crazy stuff. This is, you know, there's a simple way to sort of like ask the question of which future we're headed towards, because I think this can feel technical, this can feel overwhelming. And it's kind of like, if you just turn it into some simple metaphors, it's like, haven't we seen this movie before? Like HAL 9000, AIs That Disobey Commands and Go Rogue. We've seen that movie. let's prevent that. I think that's, I think that that's inoculated us in the same way that COVID was a bad pandemic because it made people skeptical of what a really legit pandemic could do. The next time, if there is another one, when there is another big natural pandemic, if it's in close memory to COVID, people are not going to take it well. It was a bad, you could think about covid itself as a bad vaccine the actual experience of covid yeah and i think that movies because they're fiction have predisposed people to assume that well this is overblown this is you choosing to use your bias because of these movies to apply sci-fi thinking to a real world scenario and you're pattern matching these stories that are happening and use so the predisposition of the sci-fi movies was bad it didn't warn people it made [1:16:30](https://realalpha.app/v/NufB1LL_rCU/?t=4590) them think that the future would be fiction yeah i see what you're saying but let's just make sure we just tackle the thing you're saying so the claim is that because we've had bad sci-fi movies you and i chris are the dupes we're falling for this alibaba example that was somehow not real was made up or that opening eyes research on the ai models scheming and lying and realizing it needs to change its behavior so it doesn't look like it's when it's being tested that we're the ones falling for some kind of trap i want people to just slow it down is that actually what just happened that you and i started with the sci-fi belief and we're falling into some trap that's been laid for us some catnip for our brains that says that the ai is bad i really really really want people to slow down and actually ask that question i i like there might be another accusation um i'm just driving up fear so that i can make money on a movie and do speaking engagements about why the dude is coming. Let me just say a few things. I don't make money from a single speaking engagement. [1:17:35](https://realalpha.app/v/NufB1LL_rCU/?t=4655) All the money goes into the nonprofit. I don't make money from the AI doc film. I didn't make any money on the social dilemma. My entire career has been dedicated towards what is actually in service of protecting the well-being of humanity. I've been doing that for 13 years. The only thing that I care about is what will actually help create a future that you and I would both want our own children to live in. This is coming from love of what actually creates that future. I think that if people showed up with that same energy of asking just an open-ended question of what would be the conditions that create that future that we want, if everybody did that, policymakers did that, if CEOs of AI companies did that, I think that we would have a chance of getting to that different future. And if you say that international coordination is impossible or collaboration between the US and China, that's never going to happen. I mean, first of all, that is a totally legitimate view to have given the current political headwinds. But many people don't know that even under maximum geopolitical rivalry, there have been many examples in history when countries actually collaborated on their existential safety. The Soviet Union in the US [1:18:43](https://realalpha.app/v/NufB1LL_rCU/?t=4723) during the Cold War, during the smallpox sort of breakout, they collaborated on smallpox vaccines while they were in the Cold War. India and Pakistan were in a shooting war in the 1960s, and they signed during that time the Indus Water Treaty to collaborate on the existential safety of their shared water supply while they're shooting bullets at each other. The Soviet Union and the U.S. did the first arms control talks after the film The Day After, by the way, which created the conditions in part for those arms control talks to prevent a dangerous nuclear outcome that was an existential scenario. [1:19:19](https://realalpha.app/v/NufB1LL_rCU/?t=4759) And even just two years ago, in the last meeting that President Biden had with President Xi of China, President Xi requested to add one thing to the agenda. Do you know what that was? To keep AI out of the nuclear command and control systems of both countries. meaning that look the u.s and china are maximally cyber hacking each other and they're screwing each other up every day um and when there's a set of stakes that are existential two countries that are even in conflict can collaborate on existential safety i'm not saying this is easy. I'm not saying it's going to happen by default. I'm not saying you should feel optimistic. I'm saying, what would it take for that to happen? And you start by asking that question and say, how would everyone live if we were in service of that thing that needs to happen to actually happen? Does that make sense? Yeah. Is there a challenge with AI because it's a strange kind of existential risk where everything is almost good and getting better up until the point at which it falls off a cliff. [1:20:31](https://realalpha.app/v/NufB1LL_rCU/?t=4831) You know, if we're talking about how can we get people to be concerned about climate change or we need to be concerned about smallpox, well, there's small outbreaks and there's small outbreaks damage local areas and then we try to contain it before it gets worse. And if we've got climate change, there's smoke in the sky and there's rising sea levels and there's pollution and there's extinction events of animals. Those are early warning signs, whereas it seems like AI will just improve efficiency quality of life gdp until some moment where things go badly so is this a unique category yes of x risk it is max tech mark i think said the problem with ai is the view gets better and better right before you go off the cliff like you get more amazing cancer drugs you get more incredible vibe coding tools that allow people to create crazy stuff that i benefit from um you get new material science you get new energy you get all this amazing stuff as you're moving closer to this dangerous cliff and so you can think of ai as the ultimate devil's bargain uh it's funny because peter thiel is giving lectures on the antichrist and how people who are trying to somehow say that ai is problematic those people are the antichrist is what he claims but i think [1:21:45](https://realalpha.app/v/NufB1LL_rCU/?t=4905) he's saying that because he knows that the real antichrist is ai it's the thing that makes it look like it's here to solve all of our problems and there are narrow forms of ai by the way that can solve many problems, but the current development path of releasing the most powerful inscrutable technology in history faster than we deployed any other technology that's already demonstrating HAL 9000 sci-fi behaviors, and we're releasing it under the maximum incentive to cut corners on safety. Like, it's not that the blinking cursor of chat GPT is the existential threat. It's that that arms race of what I just described, that is the existential threat. [1:22:21](https://realalpha.app/v/NufB1LL_rCU/?t=4941) and i think that everyone should be able to see that and then after this conversation you know you finish hitting play on this youtube video you feel overwhelmed and you go back to ai And then you ask it a question and it helps you figure out while your baby's burping. And you're like, that's awesome. And you forget everything that you and I just talked about because it was super fucking helpful. Yeah. Yeah. And so this is the test of ultimate modern wisdom is like, how do you step into a version of yourself that is capable of being clear eyed and stabilizing a view of the asteroid? The asteroid is coming to earth. [1:22:56](https://realalpha.app/v/NufB1LL_rCU/?t=4976) There's these weird like gravitational effects that it has before it gets here. like suddenly nudification apps happen and suddenly deep fakes happen and suddenly you know people start losing their jobs those suck those are really bad problems but like the the bigger asteroid every day that i do this work my colleagues and i we joke it's like it's like the film don't look up you're trying to tell people and you're not you're not it's not the sky is falling you're that's not the point the point is we can blast the asteroid out of the sky we need to blast that asteroid out of the sky we can steer before it's too late i i'm not hopeful but what i what i do feel is when i'm in a room of people and you walk through all these facts and you just have people ask questions and check any any of the things that we've just laid out and then you say like and people see that it's all true and then you ask them how many people here feel good about where we're headed and no one raises their hand you ask people how people don't want the future we're headed to and literally i was in davos recently and every single person every single person raised their hand that they don't want this future so there's [1:23:57](https://realalpha.app/v/NufB1LL_rCU/?t=5037) this weird thing where i think if everybody saw the same thing at the same time we could steer away into the thing you were raising earlier i think you were basically saying if there's a a lot of people think that we won't act until there's a catastrophe i think you and i both have talked to people who probably said that i mean that's what a lot of people believe and i sort of feel like there's either a catastrophe or there's a near a shared near-death experience that is a simulated catastrophe and the reason why i think this film the ai doc is so important and Again, I don't make a single dime from whether you see this movie or not, but I do think if everyone that they know, sorry, if everyone who's watching this got the most powerful people that they know to go out and see the AI doc, you got your business to see it. [1:24:40](https://realalpha.app/v/NufB1LL_rCU/?t=5080) You got your company to see it. You got your church group to see it. If everybody knows that everybody else knows, we can steer this to a different future. Just to give people an example, Jonathan Haidt, who wrote The Anxious Generation, the first country that did the social media bans under 15 or 16 was Australia. And he just talked about this on Bill Maher recently that What that did is it created an example of something that everybody was actually wanting to do, but felt like it was too extreme until Australia did it. And now literally 25% of the world's population represented by countries is moving to the social media ban. [1:25:17](https://realalpha.app/v/NufB1LL_rCU/?t=5117) Just this last week, Indonesia and India, some of the largest countries, are implementing a social media ban for kids under 15 or 16. You can pull the train back into the station. If people say the train's left the station, you can pull the train back into the station if it creates a future that we don't actually want. You can't uninvent social media, but we can say what are the steering limits and brakes that we want to apply to this technology before we get to a full anti-human future that we can't actually reverse out of. And if you don't have the common knowledge, you sound like a crazy person. You're worried about I'm going to be the one parent at my school that's trying to get my kid to not use it. And my kid's going to feel left out because there is a coordination problem. [1:25:59](https://realalpha.app/v/NufB1LL_rCU/?t=5159) Exactly. It's a common knowledge problem. It's a coordination problem. The issue that we have here is, and you touched on it earlier on, the level of coordination that's needed to fix a problem with AI is global, and it's across multiple companies. And yeah, okay, you need probably a lot of compute, but I saw what happened, the difference between how much compute was needed to train DeepSeq and how much compute was needed. It almost feels like as the AI gets bigger, you can then retrain replicant AIs way smaller. So a processor the size of this room might be able to bootload AGI in future. So, I mean, we have this with desktop synthesizers, right, for bioweapons. [1:26:45](https://realalpha.app/v/NufB1LL_rCU/?t=5205) It was a moratorium. And it means that if you try to synthesize smallpox, you get this big red flag and some guys in boots kick your door down. I'm happy to hear you reference these examples. these are important yeah they're analogous right yeah um but the problem with the moratorium for ai is it can be done so siloed right you can just take this code you can take this approach throw this model into something start to build on top of it it would be i mean nick bostrom's old example was imagine if you could make an atomic bomb by putting sand into a microwave there's nothing physics in our current universe mean that that isn't the case right but there's nothing that stop something analogous from being the case had the world have been a different way which is what you're really getting to here is what carl sagan was referring to when he talked about our technological adolescence. The conversation we're having isn't really about AI. It's about what happens when we have increasingly powerful, dangerous, and destructive technology. [1:27:41](https://realalpha.app/v/NufB1LL_rCU/?t=5261) Because as humanity progresses, whether AI existed or not, we're going to get more and more powerful technology that would cover more and more dangerous and destructive things that you could do with it. Like we didn't have the ability to do CRISPR and bioweapons 20 years ago. Now we do have that ability and we're going to get more and more dangerous and destructive of things. And so the real question that AI is forcing us to ask is what is the wisdom needed to wield the technological powers that whether AI is part of it or not, that we're going to increasingly gain. And I reference this all the time, but it's just such an accurate fundamental problem statement by E.O. Wilson that the fundamental problem of humanity is we have paleolithic brains that don't update very well to new information and think things are sci-fi and go into denial and overwhelm and blah, blah, blah. We have medieval institutions from the 18th century, and we have godlike technology that makes the 24th century technology crash down on 21st century society. That's what AI is. It's a joke from Ajayi Krotra, who's in the film, by the way. And so if we're to solve this equation, this problem statement that E.L. Wilson lays out, the way I always think about it is we need to embrace the reality of [1:28:53](https://realalpha.app/v/NufB1LL_rCU/?t=5333) our paleolithic brains. That's what wisdom is. We know that we get overwhelmed and we go into denial, so we work with that. We have systems and practices to recognize when that's about to happen and ask, what do we need to hold a difficult reality together? That's embracing [1:29:08](https://realalpha.app/v/NufB1LL_rCU/?t=5348) **Tristan Harris:** our paleolithic brains. We need to upgrade our medieval institutions. We should be using 21st century technology to make faster updating, self-improving governance. Instead of creating [1:29:17](https://realalpha.app/v/NufB1LL_rCU/?t=5357) recursively self-improving AI, we should be creating self-improving governance. Audrey Tang, the former digital minister of Taiwan has pioneered what that would look like, that democracies could be using tech to find the unlikely consensus opinions of everybody. One of the things that's going to exist in the next few months is a national dialogue on AI, facilitated by technology where people can add their own ideas about how AI should be governed. And when you vote and you click, it's going to reveal the most popular consensus opinions about what should happen about these different issues. it's one thing if we live in a world where there's a cacophony and confusion about ai and it's another thing if we live in a world where you see on a clear web page that 600 000 people have voted and 96 percent of people across all these countries agree that there should be international limits you're trying to make transparent common knowledge exactly exactly we're trying to make transparent common knowledge and one of the ways to think about it is the movement has to see itself there is a movement for humane technology there is a movement for a human future um it just hasn't it hasn't had a way to experience itself yet and when you see graffiti on an on a new york subway [1:30:24](https://realalpha.app/v/NufB1LL_rCU/?t=5424) ad for an ai product people don't need people saying i don't want this or ai is not inevitable That's the human movement. When you see kids gather in Central Park for the Lamplight Club and delete social media off their phones together, that's a club that's in New York. That's the human movement. There's so many things that people are already doing that is part of this. We just haven't had a name for, we're fighting back for reclaiming and protecting what's really human. Because again, our political voice is about to not matter. Once all the jobs get automated by AI, and governments and companies don't have to listen to the people because we're not the ones that are generating the revenue. You could have a union before when the factories need the human labor and then the human labor can get together and express its common voice. We want this instead of that. We want to be paid like this. We don't want these working standards. What happens when the companies and the countries don't need you anymore? This is literally the last time that our political voice will matter. And this is the time to express that voice in many different ways. You can boycott unsafe AI companies or AI companies [1:31:23](https://realalpha.app/v/NufB1LL_rCU/?t=5483) that enable mass surveillance. What did we just see with the thing that went down with the Pentagon and Anthropic last week is that subscriptions for ChatGPT went down like tons and subscriptions for Anthropic went up by a lot. If it wasn't just individuals that were doing that, but entire Fortune 500 companies that were doing that, given the amount of debt that these companies have taken on, the companies will have to respond to that market signal. So mass boycotts can have a huge effect on steering which AI future we get. Are we going to get mass surveillance? or are we going to get a world where AI companies are incentivized to do the right things? [1:31:56](https://realalpha.app/v/NufB1LL_rCU/?t=5516) Are you concerned about companies that are playing the same game, but just trying to manipulate their optics in a slightly sexier way? It seems to me that the incentives for Anthropic are exactly the same as they are for everybody else. I think you mentioned the Antichrist earlier on, Scott Alexander, who I think is the best blogger on the planet. He says that Anthropic is much more likely to be the antichrist than any other AI company because the antichrist would present itself as being for the people. And who are we kidding that you have the same market incentives Overtrain your models as quickly as possible with RL and the synthetic environments to keep on pushing this thing. We need as much compute. So you can talk about the optics up front as much as you want, but really that's just window dressing on the top of a burning house. [1:32:45](https://realalpha.app/v/NufB1LL_rCU/?t=5565) Yep. What you're saying is super important. And this is not me saying everyone should just put their money into Anthropic and then the world will be fine. In fact, there's actually a view by many people that Anthropic in some ways is dangerous because the view that is the safest, it's the Volvo. People know the Volvos are like the safest cars. They did a lot of marketing to present that. People think that Anthropic is just the safe AI company. So if they won, and then suddenly everything is fine. And that would be dangerous. And that we have to have a critical view of the companies and the technology and the international limits that we need that are going to happen from outside and above the companies, not just what an individual company can do. So I totally agree. How long have we got? No one knows because AI is literally moving so quickly if i went on twitter elon joked in a conversation i saw yesterday you know it used to be that you go on twitter and once every six months this is several years ago you would see a huge ai breakthrough like that would just really change the game now there's one you know that you see when you go to bed and there's a new one when you wake up in the morning and you're on twitter [1:33:56](https://realalpha.app/v/NufB1LL_rCU/?t=5636) and you see it and the pace is pretty overwhelming so i think rather than ask how much time do we have, which is kind of like just trying to assuage the fear that I think a lot of people feel, the kind of bold, brave human thing to do is to ask, if things were to go well, what would that mean about how we were showing up? And then to get as many people as possible showing up from that place, because that gets us the best decentralized conditions to get to the better outcome. They say in mathematics that in chaos, initial conditions matter. So I'm not guaranteeing in any way, shape, or form that we're going to get to the better future, but I would like to invite people again into stepping into and standing from the most wise and mature version of ourselves that would behave as if we were going to get to that better future, because that would give us the best chances of doing so. [1:34:54](https://realalpha.app/v/NufB1LL_rCU/?t=5694) And the alternative is just surrender. The alternative is schisming, deniling. Depression, overwhelm, the integrity and the kind of flow of energy that I think you'll feel inside yourself if you align with what is actually needed in this moment. There's more meaningfulness. There's more preciousness of life. There's so much meaning and purpose if you're aligned with what would have things go well. and i think there's this weird thing that's happening is if you don't see a way for things to go well and then you believe the only thing you can do is like triple down on racing as fast as possible knowing where that's currently taking us that has a biophysical cost on your system yeah i think the i was trying to think about the differences between your past work and your current work focusing on social media versus focusing on ai i don't think many people find social media that net positive in terms of their experience. I don't think that many people find AI that net negative in terms of their experience. I would, just so people hear me say this, I agree with you that the current experience of AI is not net negative. It doesn't feel that way. [1:36:06](https://realalpha.app/v/NufB1LL_rCU/?t=5766) But that doesn't mean that there's a net negative outcome that we are racing towards. But the call to get rid of social media is easy when people reflect on the last two hours that they've spent doom scrolling at night and think, I wish I hadn't done that. The feedback curve is sufficiently quick that people know how they felt even during the moment. I really shouldn't watch another one of these videos. It makes me feel kind of bad. I'm all stressed and tight. My muscles feel tense. The same thing is not true with AI. You're asking people if you were to say, and I haven't heard you say, delete your AI account. Stop using it. [1:36:41](https://realalpha.app/v/NufB1LL_rCU/?t=5801) I'm not saying that that's the thing that has to happen. Yeah. If that was the campaign, if that was what people were supposed to do, people would be sacrificing their quality of life. They would be able to fix their car less quickly. They would be able to find out things that they need to about their health, to analyze their health documents, to be able to liberate them, to be able to make hard to access information cheaper and easier, and actually to enable their quality of life in a manner that didn't happen with social media. So I think the intermediary experience is less sexy and that I'm going to guess is also the reason why you're not saying masks unsubscribe, don't use any AI because you're probably aware that the incentives for individual users, the value judgment that they make isn't sufficiently imbalanced as it would have been with social media. [1:37:30](https://realalpha.app/v/NufB1LL_rCU/?t=5850) For me, it's easy to not have it on this phone. If you were to say, I don't want to have chat GPT on my phone, no, I'd need it. It's very useful to me. I think it's about understanding context and what is the careful and limited way that these tools are helpful, and then creating the conditions where that's the collective outcome that we're getting. So, for example, for kids in tutoring, theoretically, this is the best tutor in the world for learning. Alpha School is doing a South by Southwest pop-up in the middle of downtown. Mackenzie Price was on the podcast. Oh, interesting. I mean, I think that it has enormous potential. [1:38:07](https://realalpha.app/v/NufB1LL_rCU/?t=5887) If you look at how most kids are using ChatGPT right now, they're using it to cheat on their homework and to actually outsource their thinking. and actually because they use it so much turning into what the some atlantic writer wrote uh called lemmings or ll lemmings where they basically outsource all of their thinking for every moment-to-moment decision in their day about how i should respond to this person or there's a guy over there i want to talk to and i don't know what to like whatever the thing is people are outsourcing their decision and they're not actually learning themselves how to show up as a human and that is not going to create a safe world again that's a version of the narrative the possible of a technology that were promised and then the probable of what's actually happening and it's like crumbling cognition yes exactly and we saw this with social media we thought it was going to create the most enlightened informed society in the entire world because we have the best information it wrecked our attention and we have the most confirmation bias tribalized low trust worst critical thinking society that we've had in a generation and so again i think this is just about part of the wisdom we need is is to be able to look honestly [1:39:12](https://realalpha.app/v/NufB1LL_rCU/?t=5952) at the nature of a technology and look honestly and confront ourselves with the reality of how technology is currently being used and deployed. And if it's driving up, you know, a social media influencer culture instead of creating astronauts, like, you know, in China, they pulled the top, what are the professions that kids want the most? And the number one was like astronaut. Number two was like teacher. Engineer or something. Engineer and then teacher or something like that. And in the US, it was influencer. like i can tell you which world and which civilization you're going to get if we really want to beat china and innovation we can show me the incentives and i'll show you the outcome show me the incentive and i'll show you the outcome and if you really want to beat china you regulate social media and you stop brain rotting your entire population and you actually invest in genuine educational technology that silicon valley would ship to their own children This has to come top-down because the incentives, both at the user level, especially now for AI, and at the company level, and at the market level, and at the international level between companies, all of these incentives align in the same direction, which is, I enjoy using the product, [1:40:20](https://realalpha.app/v/NufB1LL_rCU/?t=6020) my company generates lots of revenue for my country for the product, my country doesn't want to fall behind other countries in the race to have additional capacities for technology. All of these are aligned in the same place, which means that it has to come from above all of that, which is a pan-national global movement for humane technology that's actually in the service of people. Part of that movement is redefining the currency of what it means to win that race. Right now, again, the U.S. beat China to the technology of social media, but we have been losing in the race to govern that technology like i think you know these examples but um in china as an example they i think limit social media use to 40 minutes a day if you're under the age of 14 and i think it's only friday saturdays and sundays either that's for social media or for video games there's lights out at 10 p.m meaning that if you open a social media app after 10 p.m it's closed and it opens again at six in the morning um this is helpful for when it comes to the kind of late night FOMO brain brain rot doom scrolling you know not getting sleep and I'm not saying that the U.S. should unilaterally like some totalitarian government make that choice or it's just that we [1:41:33](https://realalpha.app/v/NufB1LL_rCU/?t=6093) can democratically come up with guardrails like this another thing that China does is they have a synchronized final exams week and they shut down AI during final exams week no way yes so that so that all the students can't use know that if they used AI during the school year it will completely screw them over during final exams because they're going to be reliant on something they can't use That's right. And so we don't have to do what China does. But the point is, we should do something. In China, they're regulating anthropomorphic AI to not do all the attachment hacking and the suicides and all this kind of stuff. I'm not saying what they're doing is good. I'm just saying they're doing things. We're not doing anything. And we can. And the purpose of the human movement is to actually be demanding. You can actually go and call your legislator and say that you do not want this anti-human future. You want accountability for AI companies. [1:42:20](https://realalpha.app/v/NufB1LL_rCU/?t=6140) You want to ban legal personhood for AI. You want, you know, no anthropomorphic AI so that we don't get the child safety issues with AI chatbots. If I was to give you... You, the option of turning the rest of the world into a totalitarian state, everything, but it meant that we were able to dictate what happened to countries and to companies and to these AIs. I don't want to create a totalitarian state. But the alternative would potentially be destruction of humanity. Which would you pick? You are conjuring a thought experiment from Nick Bostrom, which I'm sure you're aware of, and the paper you're referring to is the Vulnerable World Hypothesis, in which he says that as we discover more and more technologies that in a decentralized way have the capacity to destroy things. [1:43:12](https://realalpha.app/v/NufB1LL_rCU/?t=6192) So like your example of if it turned out that microwaving sand created a nuclear reaction that blew up the entire world and we publish that knowledge and it spread virally on social media, how many minutes, hours or seconds would it take before the world blew up? And then the only answer to a distributed, available-to-everybody destructive capacity would be a global totalitarian society that monitors and surveils everyone and what they're doing. So you prevented mass destruction, but you got 1984 Big Brother, and that's an uncheckable power that people cannot fight back against. I am very worried about mass surveillance. [1:43:51](https://realalpha.app/v/NufB1LL_rCU/?t=6231) I think that what happened with Anthropic is hopefully bootloading a global immune system to the risks of AI-enhanced mass surveillance. Because in a way, Big Brother could have never happened without AI. When you have an AI that can process every camera and every face and all of the emails and text messages that are flowing through everything, and you can actually ask the question as a state, who are the number one threats in my society that I need to suppress? And it tells you and summarizes instantly who those people are, synthesizing all these data streams. How can you fight back against an uncheckable power when it knows all of your secrets? You can't. So we're faced with very tough choices on both ends. And I wrote, I did this in a TED talk I gave this last year called the narrow path that we have to avoid both outcomes. We have to somehow be committed to finding the narrow path between, you know, decentralizing destructive power in everyone's hands without responsibility, without a commensurate wisdom or responsibility in which you get chaos or catastrophes, or over-centralizing that power in uncheckable ways in which you get runaway dystopias. And we have to find something like that. [1:45:00](https://realalpha.app/v/NufB1LL_rCU/?t=6300) like a third attractor, to quote Daniel Schmachtenberger, or a narrow path that seeks to avoid both these outcomes. You could say that that is outside the laws of physics or not possible. But I think what's important is to be committed to and living from the place where we would be finding that path, because the only chance that we would have of finding it would be requiring that everyone was living in service of it. Isn't it funny that the two worlds that you identified there, one totalitarian overreach that's controlled by humans and at the mercy of their biases, and the other being decentralized vulnerability danger that is at the mercy of technology's liabilities. Both of those are extreme versions of something that we definitely don't want. But the restriction on the one that would protect us from the AI goes back to the bottom of the brainstem, that there's almost never a time that a country or a government or a small number of people or a large number of people get most of the power and then give it back. It's a ratchet system. And it only ever clicks into place and then never goes back. Which is why you need to have built-in checks and balances on power that are irreplaceable. You cannot have a world where [1:46:14](https://realalpha.app/v/NufB1LL_rCU/?t=6374) whatever power gets centralized has to have oversight and democratic accountability and distribution of that wealth and power, depending on if we're talking about the wealth versus the power. How would we know that some country isn't signing up to a moratorium but secretly doing all of their research behind the scenes so that they slow down their progress of everyone else but allow them to race ahead? So you're asking the question, if let's say the US and China sign an agreement, how do we know that given their maximum incentive to defect on that agreement and have the CIA or their black operations still continue to do the research project? This is obviously the hardest technology and coordination problem that humanity has ever faced. I just need to say that at the top. I'm not saying any of this is easy. It makes nuclear war look like child's play. [1:46:57](https://realalpha.app/v/NufB1LL_rCU/?t=6417) Correct. But by the way, I want you to know, there's a great video, and I can send it to you, of Robert Oppenheimer asked in the 1960s, I think he's testifying for something, in an interview, and they asked him, how could we control this technology from proliferating? and he takes a big puff and a cigarette and he says it's too late if you wanted to prevent the spread of nuclear technology you would have had to do it the day after Trinity Trinity was the first test of the first nuclear bomb He believed, I think this interview was in the 1960s, he believed it was inevitable, there was nothing we could do, and the world was fucked. [1:47:38](https://realalpha.app/v/NufB1LL_rCU/?t=6458) even he the creator of the atomic bomb didn't see that a lot of people would work very hard over the course of the next 30 years to invent what's called national technical means or satellites that do mutual monitoring enforcement that we would create the seismic monitoring technology to to be able to to see if you were doing an underground nuclear test or an above ground test because we would see the reverberations of that so satellites seismic monitoring you [1:48:04](https://realalpha.app/v/NufB1LL_rCU/?t=6484) **Tristan Harris:** know, international inspectors, the International Atomic Energy Agency. We had to invent all of these new things to create a governance system that was capable of dealing with nuclear weapons. [1:48:15](https://realalpha.app/v/NufB1LL_rCU/?t=6495) Interesting that the innovation around the technology that was dangerous was probably less sophisticated than the downstream technologies required to make it secure. What do you mean? That making the atomic bomb was one thing. Then there's 50 things that need to happen in order to have that technology exist in the world. In a safe way. Yeah, exactly. There's a safe, there's sort of an asymmetry where the technologies, the destructive capacity is easy to create. The governance capacity is way harder to create. The defensive capacity is harder to create. And you could have said at the beginning of 1945, I guess it's inevitable. [1:48:48](https://realalpha.app/v/NufB1LL_rCU/?t=6528) 150 countries are going to have nuclear weapons. There's nothing we could do. Let's just like drink margaritos and give everybody uranium and increase GDP by selling uranium to the entire world. But we didn't do that. And people had to invent stuff. We needed our best engineers and our best minds working on how to figure that out. And I'll just say, if you want to look this up, RAND, the nonprofit defense think tank, has a paper on how international monitoring and verification mechanisms could potentially work for AI. It's not easy. It is not trivial. It would require extraordinary investments in mutual monitoring and enforcement and data centers and chips that attest where they are and all this kind of stuff. [1:49:25](https://realalpha.app/v/NufB1LL_rCU/?t=6565) Okay, so there are some suggestions for it. because what i was thinking is you were talking about well we can use satellites that work out if you're doing things above ground or below ground we can monitor maybe the amount of heat signatures and electricity signatures power usage so there are some inspectors yeah there are some signatures of people that are doing this sort of research because you can't do it without there being any externality you can't just run it on a macbook and make it look like you're you need large clusters of compute right now to your point in the future it may shift but right now you need large clusters of compute you need advanced semiconductor manufacturing supply chains right now that's a set of u.s allies the trilat i believe it's uh denmark japan south korea obviously Taiwan, you know, we could create some kind of infrastructure for not some totalitarian control of compute, just a governance regime. Would you call the IAEA totalitarian or would you call it a governance regime for a dangerous, destructive nuclear capacity? I'm not saying any of this easy. We have to navigate the narrow path. We don't want to create totalitarian controls. We don't want to create runaway catastrophes. What it takes is being committed [1:50:26](https://realalpha.app/v/NufB1LL_rCU/?t=6626) to finding that path. How much time or energy or resources have we spent even trying? People say it's impossible. Have you spent a month dedicatedly trying? People say it's impossible. Have we spent any amount of resources actually trying to do this? No. Let's say that there was some way for you to have God's eye coordination and that you could step in. Would you just put a pause on all model development to allow us to catch up in terms of AI safety? I think the people, rather than it sounding like I'm some kind of pause or stop person, the people building this technology have basically said that that's what they would most prefer. [1:51:03](https://realalpha.app/v/NufB1LL_rCU/?t=6663) They would most prefer a world where we basically stopped and had the time to integrate and develop this technology slowly. That's what they would prefer. What do you think that if you had control, what would you do? I would agree with the people who I'm going to defer to as the people who know the most about the dangerous and destructive capacities. So I don't want people thinking this is my view, that i'm some kind of you know safety net or something like that this is just about like i want to be able to look my children in the eyes and say we did everything that we could to create the best possible future for you and i think that basic heuristic of like are you operating in service of life and the things that most matter i think that that would probably be the wisest thing to do and just to people like ground that the ceo of microsoft ai mustafa soleiman who's a friend, says in the future with technology, progress will depend more on what we say no to than what we say yes to. Your podcast is called Modern Wisdom. Is there a definition of wisdom in any spiritual or religious tradition that does not have restraint as a central feature of what it means to be wise? Does any religious tradition or spiritual tradition say, you know what's wise [1:52:10](https://realalpha.app/v/NufB1LL_rCU/?t=6730) going as fast as possible, not thinking about the consequences and like max, you know, dopamine so trivial it's so obvious it's so obvious it's just like snap out of the trance this is not inevitable they want you to believe it's too late it is very far down the tracks it would be a rite of passage even if it didn't work out for us to show up with the maturity and responsibility to at least be trying to live in service of the future that we actually want to create. Can we watch the trailer? Can we get the trailer up? I want to see the trailer for this movie. That'd be great. Thanks. If this technology goes wrong, it can go quite wrong. [1:52:49](https://realalpha.app/v/NufB1LL_rCU/?t=6769) What? Your fear of AI is the collapse of humanity. Well, not the collapse, the abrupt extermination. There's a difference. So I started making this movie because my wife is six months pregnant. It is now a terrible time to have a kid. I mean, just to be honest, I know people who work on AI risk who don't expect their children to make it to high school. I, what the... How does AI understand pretty much everything? It's surprisingly straightforward. Intelligence is about recognizing patterns. Patterns. Patterns. Patterns. If you have learned those patterns, you can generate new information. AI is moving so fast. [1:53:40](https://realalpha.app/v/NufB1LL_rCU/?t=6820) It's being deployed prematurely. There's so much potential for things to go wrong. Why can't we just stop? All these companies are in a race to get AI that's vastly more intelligent than people within this decade. China, North Korea, Russia. Whoever wins is essentially the controller of humankind. We need to take a threat from AI as seriously as global nuclear war. It feels like I have to find these CEOs and get them in the movie. Great. I want to ask you to promise me that this is going to go well. That is impossible. Okay. Am I hopeful? Yes. Am I confident that it'll go right? Absolutely not. [1:54:29](https://realalpha.app/v/NufB1LL_rCU/?t=6869) AI is the thing that can solve climate change. We could cure most diseases. What if it's expanding what is humanly possible? This is the most extraordinary time ever. The only time more exciting than today is tomorrow. I already love you. Okay. I think if this technology goes wrong, it can go quite wrong. By using AI, we're about to move off of the earth into the cosmos. If we can be the most mature version of ourselves, there might be a way through this. This is the last mistake we'll ever get to make. Dude. Does it land differently after this conversation? It does. I'd already seen it, but yeah, it definitely, I definitely understand more of what's implied. It's impressive that you managed to get all of the guys to sit down, or most of the guys who are you missing uh i mean there's obviously many people building this technology now um so they don't have the chinese labs in it um they have demis hasabas from deep mind sam altman from open ai dario from anthropic i mean those are the three players elon agreed to participate in the movie and then um i think it was right in the beginning of the first um few days of the trump administration that he was busy and didn't follow through so but i think the team [1:55:56](https://realalpha.app/v/NufB1LL_rCU/?t=6956) really wanted him to be in the movie and wanted to hear his views. I mean, to give Elon credit, he was one of the first people who cared about AI safety. And he said, mark my words, AI is far more dangerous than nukes. And he said this in like, what was it, 2015, 2016, like way before people were taking AI seriously. Like back then AI was just recommending what other product you should get on Amazon and doing facial recognition for toll booths and bridges and stuff like that. Think about what much of super intelligence and sort of the fallout around that book from nick was so much of it was almost uh advice for how to talk to your friends about this without being mocked too much right it was these are the best examples to use so that you don't sound like the insane person at thanksgiving dinner and ironically the paperclip maximizing example did make people sound ridiculous and insane uh that then got really you know diminished and what's the word i'm looking for uh it just made people it was used to to tarnish people's reputation if you talked about paperclips but it's funny because people say oh like the ai is gonna for people to know the example it's like there you are and the ai is gonna say is gonna be told to maximize [1:57:03](https://realalpha.app/v/NufB1LL_rCU/?t=7023) paperclips and then the way it will figure out how to do that is figuring out any strategy which means like turning every atom in the universe to paperclips which means like you know melting all the humans down and getting them to you know it just taking to this extreme and it sounds totally sci-fi, but if you actually ask a baby AI called social media that's pointed at your brainstem figuring out just what video to get you to watch that keeps you on the screen, and you say maximize engagement, well, conflict and rivalry and civil war is really good for engagement. So, in a way, we have been maximizing a paperclip called attention and eyeballs for a long time. And it's driving up division and rivalry everywhere around the world. And democracies are backsliding everywhere around the world. And it's driving up confirmation bias all around the world. And the point isn't that the baby AI of social media hates you or hates your well-being or hates your connections or hates your democracy. It's just that it doesn't care about anything other than whatever keeps your eyeballs. [1:57:56](https://realalpha.app/v/NufB1LL_rCU/?t=7076) And that little baby AI that was just figuring out which photo or video to throw in front of your nervous system was enough to completely transform everything, everything, everything about how our society worked. And I saw that, by the way, and I'm not some kind of like, it's not like I especially see the future, but in 2013, it was so obvious to me, if you take these incentives really far in a decade, I can tell you where you're going to live in. and the thing that i want people to feel is if you can see the incentive you can and you have that clarity you can you can confidently say i don't want the future that that creates in 2013 mark zuckerberg could have said oh my god i see we're about to create an arms race to hack human psychology i'm going to convene all the leading social media companies and the government i'm mark zuckerberg i've got billions of dollars i'm going to throw money at this i'm going to get the government people involved not because the government's trustworthy or because we like them but because i see that we need some kind of rules and i'm going to try to create some rules that say no auto-playing videos, no infinite scroll, no one can create unnecessary FOMO. [1:58:56](https://realalpha.app/v/NufB1LL_rCU/?t=7136) You can't dole out like a few likes here and then a few likes there, like a slot machine. You have to do them all at the end of the day or something like that. Like you can create rules and norms so that we didn't get the mass addiction distraction machine that we then did get. And he could have done that in 2012, 2013. That's what leadership, that's what maturity, that's what wisdom would have done. It's not that hard to see it. The problem I think that you're facing is every second is so crucial and every single second of compute and of CEO attention and staff-based attention is going to be paid on continuing to get ahead in this unbelievable one to rule them all race. I mean, OpenAI is a partner on this show. I've spoken to Sam twice and both times the calls have been so brief because presumably for every second that he's on the phone that's billions of dollars of potential revenue or compute or that is being wasted and that means that if you want mark in 2013 or sam or dario but the point is you create so like literally, this is the history of all of law. So like I could kill you and steal your stuff and just take your money, but I'd prefer to live in and everyone could do that. And that'd be a faster [2:00:09](https://realalpha.app/v/NufB1LL_rCU/?t=7209) way to like get money. It's like just everybody kills everybody, grabs their money, but that would create a society that's chaos. So instead I sacrifice some of my abilities, like I can't kill people. And instead we have law that we, we all sort of notch down some of our individual capability so that we get to live in a society that we actually want to live in. And that's what this would do so in 2013 if mark zuckerberg had said i'm going to convene you know uh musically which was before tiktok um twitter and the other ones and we're like okay look it is so obvious we're in this arms race for attention just like public utilities which by the way if like in california it'd be pg and e in texas what's your electricity provider it's like oh fuck knows i don't know i have no idea i have no idea what my electricity okay well technically energy companies have an incentive to maximize revenue so theoretically they're like leave the lights on leave the stove on like do the water 24 7 and because we make more money when we do that but because public utilities rely on a scarce resource called energy that has environmental uh um emissions that that are that are we have to deal with there's a decoupling of revenue from um their [2:01:16](https://realalpha.app/v/NufB1LL_rCU/?t=7276) incentive so for example in california you're charged a base rate for like the initial energy you use and then once you're hitting kind of the capacity that would be straining the system [2:01:25](https://realalpha.app/v/NufB1LL_rCU/?t=7285) **Tristan Harris:** We start to charge you more, but that extra revenue doesn't go just into PG&E, the energy company's pocket. [2:01:31](https://realalpha.app/v/NufB1LL_rCU/?t=7291) It goes into a fund to have more clean energy to offset it and create more energy capacity. So with attention, you could say instead of companies maximizing attention and driving revenue, you get to make money from the initial tier one of attention that you're getting. And then after that, the resources go into a common pool that's investing in the research, the X prizes, the design solutions that research and show that there's different ways of doing news feeds. There's different ways that we can have news feeds that are all about directing you back towards community. You live in Austin, Texas. Austin has, at least my friends, there's a lot of community events happening here. But you can imagine a world where the default news feeds that govern our world were spending like 30 to 40% of what you saw was other events that your friends and other people that you knew adjacent to you were doing. You can imagine a world where instead of dating apps profiting off of keeping you in a slot machine loneliness thing where you message someone and never get back to them in this like new world all the dating apps would be forced probably because of some lawsuit in the against engagement model to fund physical events in every city that they operated in, where every single [2:02:36](https://realalpha.app/v/NufB1LL_rCU/?t=7356) week there was actually a physical venue where lots of people were put in the same room with lots of other people they matched with who are maximally overlapping. And now instead of feeling scarcity and loneliness, there's a sense of abundant access to community and soft dating and soft friend making environments. So you're dealing with a loneliness crisis. You're dealing with the dating problem. When you solve that, then it turns out 30% of the polarization online goes down because it turns out that a lot of polarization was just people being lonely and depressed from not having connection. It's all manufactured because people are caught in the doom swirling loop. So it's like, again, we can have a better world if we deal with the upstream causes of these problems. We can also have more innovation. Peter Thiel was talking about, we need more innovation. We need more scientific development. We would have stagnation if we didn't have AI. Well, how about you regulate the brain rot economy that's currently degrading all the innovation, causing people to be not productive and creative and innovative members of our society and instead incentivize entrepreneurs and again, social groups and community. So people are [2:03:32](https://realalpha.app/v/NufB1LL_rCU/?t=7412) actually making things. So the point is, I'm a tech, I'm a different kind of technology optimist, which is a humane technology optimist. If you design technology that is humane to an understanding of the substrates of society that upon which everything else depends, you can have a technology environment that is conducive to the things that we want our society to do. But it requires different design principles, different incentives, different rules, and some coordination. This third attractor, precipice, narrow path thing that you think is important for us to walk appropriately, the pace at which things are going to happen and change, do you think that we're going to be able to move along that? I mean, you've got this movie. It's going to be a global moment. Lots of people are going to see it. It's going to get people talking, common knowledge typically things tend to take time conceptual inertia usually moves over generations and centuries going from heliocentric model of the universe took like a hundred years well said yep this is the problem in general that we face with technology and this is the eo wilson quote it's not just that we have paleolithic brains medieval institutions is that [2:04:39](https://realalpha.app/v/NufB1LL_rCU/?t=7479) the brains operate on a very slow updating clock rate the institutions operate on a slow clock rate and the technology moves at whatever the 21st or now 24th century technology clock rate that how do you think about expediting this change you have to have your governance move at the pace of the thing that you're trying to govern is that ever going to be possible governance move at that pace? The lumbering behemoth? What is it? Leviathan leans left and lumbers along? There needs to be, rather than recursively self-improving AI that is uncheckable by any human oversight process, in which we will for sure lose control and build something crazy that we regret, we instead should have self-improving governance. You can use AI to look at all the laws on the books and say, what are the laws that don't matter anymore, that are creating all this red tape that we don't need that's like stifling innovation and get the ai to find those laws and and reinterpret them and say what are the ones we need to get rid of and how do you rewrite them for 21st century technology in the current age we could be using ai and technology to update the governance as fast as the technology but we'd also probably need to slow down the technology [2:05:45](https://realalpha.app/v/NufB1LL_rCU/?t=7545) which we would benefit from because then we wouldn't crash again it's like china wins when they whisper in our ear go faster go faster you're gonna blow yourself up pyrrhic victory ahead Pyrrhic victory. You said this earlier, when the U.S. races ahead and has a more sophisticated model, China gets it like 10 days later. Because first of all, they have spies on all of our companies. And second of all, they distill all the models. Is that what they're doing? they have spies in the companies you know they're maximally incentivized to know what's going on plus as i as you were saying they can distill the u.s models meaning they can like query those models a thousand times and kind of distill the essence of those models and then make and train theirs so there's a great like meme online of um like it's a motorboat with a guy who's um what is it called when you're um wake surfing wait yeah like you're wake surfing behind him on the on the string i mean on the what is that called the string yeah whatever um and and the the guy who's on the um in the surfing says go faster to the boat that's like racing ahead that's the u.s that's building the advanced model but then they're literally on a string right behind [2:06:47](https://realalpha.app/v/NufB1LL_rCU/?t=7607) them that's getting all the benefits from it yeah and there is a study that anthropic found um that china had been covertly even using u.s anthropic ai models to perform a cyber hacking operation so like again if we're winning the race to the technology but losing the race to governing or controlling or protecting the technology what the fuck are we winning yeah yeah it's like it's so basic what we're talking about it's like this is not rocket science it's unbelievably simple if you have the power of gods you need the wisdom love and prudence of gods i hope we've laid out a bunch of examples of how we can do that and while it's not easy it's not does not happen by default if you get everyone you know to go out and see the ai doc understand that AI is dangerous, get your church group, get your business, and recognize that we need to, you know, not build bunkers, but write laws to actually steer AI before it's too late. [2:07:40](https://realalpha.app/v/NufB1LL_rCU/?t=7660) You're terrifying, but I'm glad that you're in the world, man. I appreciate you. I appreciate you too. I really appreciate this conversation. Thank you. All right. Goodbye, everybody. LinkedIn is pretty amazing at helping you grow your small business.