AI Will Change Your Life Faster Than You Think, According to Former OpenAI Exec Zack Kass

It's the idea that we are building machines that are going to be so smart and so inexpensive that the human cognition will commoditize, our relative cognition, and that we will have to then determine other means of differentiation. By the way, I think there will be plenty. But that, I think, is basically the arc over the next 10 years. Unmetered intelligence does not mean that everyone's going to be smart. In the same way the internet doesn't mean you do research, in the same way that literacy does not mean that you read, Unmuted intelligence does not mean that you will be brilliant. It simply means you have access to brilliance. And then what we do with it will actually sort of define everything. Zach Kass is one of the world's leading voices on the future of AI. We are on the verge of the most profound industrial revolution in human history. That is pretty clear. With over 16 years at the forefront of the AI revolution, Zach is known for serving as the head of go-to-market at OpenAI, where he helped commercialize breakthrough technologies like ChatGPT. If I had 30 seconds to tell the world anything, it would be that today is the best day ever to be born. Today, Zach is a global AI advisor, keynote speaker, and trusted consultant to some of the world's most influential organizations, including Coca-Cola, Morgan Stanley, and Samsung. They're building machines that have human intellectual equivalents, and that AI is capable of doing that more and more. Get ready. If you're interested in the future of AI, you won't want to miss this conversation. Welcome back to The Mellow Millionaire. Today, I'm very excited. I got Zach Kass here. Zach Kass is one of the leading voices on artificial intelligence and its impact on business, society, and human potential. Zach, hey, it's a pleasure to have you on. Congratulations on the new baby. Thanks for having me. I guess what I like to do when we get started is just jump in and tell us a little bit about yourself, what you're excited about, what you've done here in the last decade and just really let us get to know you a little bit. Well, in the last month, as you pointed out, I had a newborn. So that's certainly my most proud accomplishment in recent memory. But in the last decade, in 2010, I started working in a machine learning company that was building data labeling services for, at the time, primarily businesses with huge, huge AI and machine learning demands. And it was statistical machine learning models. So they required, you know, there were hundreds of millions of parameters and they were just very long, if this, then that lines. And you had to fill each of these left turn, right turns with human labeled data. Today, you know, these businesses as like Scale AI and Mercore, these, you know, many billion dollar companies that do this data. And we were this humble startup that sort of started this market. Crabflower became figure eight, sold to Appen. I went to a company called Lilt, started building large language models for the purposes of machine translation, also started by two Stanford researchers. And that's when large language models started to take off. And the models went from hundreds of millions to billions of parameters. And then in 2021, I got a chance to go to OpenAI around the launch of GPT-3 as the first business hire. and that of course gbt3 was was a model with um tens of billions uh of of parameters and it was a very good text-to-text model and no one had ever heard of it and we put it in the api and we got you know a few million dollars of revenue and if you uh emailed support at you would arrive right at my right at my inbox and then you know we we went at it and over the next few years, turned open AI from a company with millions in revenue to billions in revenue and made AI a sort of household name and now spend all of my time helping global 5,000 companies and governments and NGOs figure out how to make the most of this AI. And I do so sort of from two vantages. I have this advisory position and then I also have two academic positions. one at Hong Kong University of Science and Technology and one at UVA, and I get to teach and do research. So it's, yeah, it's a ton of fun. Yeah. We talk about this a lot internally here, just what does the world look like? I think it was Bill Gates said in five years, people will not have to work. We got AI going into machines here, Tesla. We got Fusion. I was just sitting with Bessemer and they said they're investing in a massive Fusion company. And they said, there's one out of 10, we're putting all of our money on one, but energy is going to become more available and more affordable. Where do you see the world going here in the next five to 10 years? And that's a tough question to start out with. So the theory that I coined in 2021 that we spend a lot of time on now, the theory that I coined is the theory of unmetered intelligence. I was traveling around the world with this guy, Boris Power, the head of research engineering at open AI and we were showing CEOs GPT-3. We would show up in an office, we'd show up in the CEO's office and we would open the laptop and we would open up GPT-3 and say, look at how cool it is. And we'd do these demos for like 45 minutes. And the end of the demo, they'd be like, is that it? And we're like, well, yeah, but how much do you want? How much are you going to buy? And they're like, well, none. Thanks, but please get out. And so we realized, we were like, man, why are these corporate, you know, at the time we were like these corporate dummies, you know, they're all going to get left behind these dinosaurs. But then we were like, you know, maybe it's us. As we realized like GBD3 was actually too slow. It was too expensive and it wasn't good enough to materially change a business outcome. You wouldn't have cared about it. Jimmy Diamond didn't really care about it. People were like, eh. And so what we did is we started doing the demo and then I would close my laptop and I'd be like, okay, fine. Now imagine it got a lot better, a lot faster and a lot cheaper. And what we literally said was, imagine if we built GPT-4 or GPT-5. And then we took it a step further and we said, now, what if it got so much cheaper that we experienced this idea of abundant intelligence or unmetered intelligence, as we called it? This theory that said at some point, the amount of compute on earth would be so great, the cost of compute would be so low and the quality, right? The performance of the compute would be so powerful that the individual cognitive capabilities would pale in comparison to the collected, that it would be very hard for CEOs to compete on the basis of intelligence, right? And that moats that we had long assumed would be here forever would actually eventually start to crumble. And that proposal then was kind of wild. Today, it's becoming pretty reasonable. And that is the basis on which I wrote the book, which comes out January 13th. Shameless plug. It's the basis on which I basically make most of my arguments. It's the basis now on which I plan a lot of my personal life and professional life. It's the idea that we are building machines that are going to be so smart and so inexpensive that the human cognition will commoditize, our relative cognition, and that we will have to then determine other means of differentiation. By the way, I think there will be plenty and we can talk more about those. But that I think is basically the arc over the next 10 years. And the best way that I can sort of draw a comparison to this, and you just picked a really interesting one, is this of course happened with water. We have unmetered water effectively now in most of the developed world. We have unmetered electricity in most of the developed world. Most people do not think about plugging in their device or another. You don't think about plugging in the next appliance. And in the developed world, we actually, for by and large, this is a very hot take, have largely unmetered food. The average American does not really think about eating the next thing because it doesn't cost that much. Now, a comment can be seen as insensitive, so I'll be careful. But the point is that we have built so much capacity into these utilities that we observe them now sort of exactly as that. And the best one is the internet. Internet serves as the best example of what it looks like when a resource rapidly commoditizes and becomes abundant and unmetered. And my argument over the next 10 years is that is basically the path that intelligence will follow. And then it stands to sort of suggest that one, we're going to have to update really quickly to some societal constructs and social norms. Two, companies are going to have to sort of recognize quickly that it's not what they know, but how they operate. And three, that just because we have unmetered intelligence doesn't mean we'll have universal brilliance. Unmetered intelligence does not mean that everyone's going to be smart in the same way the internet doesn't mean you do research in the same way that literacy does not mean that you read. Unmetered intelligence does not mean that you will be brilliant. It simply means you have access to brilliance. And then what we do with it will actually sort of define everything. And that I think is actually just the next 10 years is this path to it. And the reason that's particularly interesting based on your introduction and what you just said is when people are like, whoa, unmetered intelligence is amazing. I'm like, yeah, it's really cool. You know what's really cool and probably even cooler? Unmetered energy. And that's actually what fusion promises. The amount of energy that you can put in a single fusion, small modular reactor that you can drill a mile under my house and power Santa Barbara. You don't stop thinking about energy consumption at that point. Like there's no, you know, the, the, the, the offshoot, the carbon offset is, is, is negligible. You don't need to do anything because you just have this incredible source of, of exceptionally powerful energy. So I'll park that there. I love this stuff. This is going to be really fun. So the other thing we're talking about is batteries, right? Because a lot of the world, whether it's autonomous cars, uh, I'm looking at a, it's called a Jetson. It's this little thing you could fly in. It's a one person thing. It's got eight motors on it. But what do you think about batteries and being able to store energy in vehicles and whatnot? The knock on batteries today is that their half-life is shorter than we would like. So they become sort of dump garbage faster than we would like. But they still don't have the capacity that would really differentiate them and that they're still actually quite hard to charge. Like, you know, the joke about Tesla is that it's an electric car that runs on coal because so many of these Tesla superchargers are being powered by non-renewable energy. if you can build better you know proton batteries if you can build better you know like the the basic cell batteries that that that tesla are building you start to imagine being able to big build much bigger batteries and the reason then that you want to build bigger batteries is because you want to be able to store and transport energy and that that will change the game for things until really small modular reactors become so portable that you can actually start to you know put a nuclear reactor inside of every inside of every city and now all of a sudden storing the data becomes much less important because now actually you can just have localized energy sources today we think about energy as being this thing that like has to be created in this single big site and you have to you know you have to draw energy we don't know how to move energy on the grid very well. So you have to go to these places. And I'm just not sure we're going to think about it that way for much longer. You know, I was sitting down with Eric Schmidt. I wasn't talking to him personally. He was in a small group with Goldman Sachs. And he was saying that the United States is winning on every front except for energy. And that's not going to be a problem soon. And he said something really interesting. He said, the guys in Silicon Valley and San Francisco in particular believe that there's going to be such a big breakthrough in the next three years. And he didn't allude to what it was, but he said, it's going to change the way human beings live and everything about human beings And I know it was having to do with AI Do you have any idea what he was talking about in the next three years this massive breakthrough It might been energy I don know what he was talking about but the old CEO of Google Based on what I heard him say and based on what I think he believes my assumption is that he alluding to super intelligence. He may be referring to fusion. The problem with referring to fusion as the breakthrough is that even if you solve fusion tomorrow, so let's say tomorrow we were like, yeah, we know how to do it. You'd still need like five or seven years to actually build a reactor that works. I assume what he's referring to is super intelligence. I assume what he's referring to is basically this idea of unmetered intelligence. The way that I couch this for people is, and I wrote a newsletter yesterday that it was my best effort to actually explain why unmetered intelligence is so radical and cool. It's that the solution to most of our problems are discrete. Some of our problems are indiscreet, by the way. And those problems are fascinating. An indiscreet problem is like, what's the meaning of life? What is human's purpose on earth, et cetera, et cetera? Is there a God? These are indiscreet problems. We can't really measure these things. We don't even know how to build an equation that could. A discrete problem is, what's the cure for prostate cancer? That is a discrete problem. There is information on earth that could solve that. The amount of information on earth can satisfy that equation. The problem is the amount of intelligence today cannot. It takes too many people staring at too many different permutations of that problem. And there are only 10,000 oncological researchers to begin with because it turns out we're not that smart. I mean, we're smart, but we're not that smart. And when we're born, a certain number of us qualify for novel scientific jobs. right it's only a certain percentage of the population can actually move us forward on novel sciences and the idea of unmetered intelligence is not oncological researchers and we turn them into a million it's what if 10 000 oncological researchers were 10 000 times more productive and that is happening like that i think is what people are staring at and when you talk about these like fanciful ideas like oh we're going to solve all these things and people go yeah bullshit. It's like, well, no, let me explain why. Because it's actually the problem is more easy to explain and comprehend than you think. We've just created these false boundaries in our brain. This is the other fascinating thing about the human experience. If you had talked to a human in 1900, you would have explained space travel and they would have said, no way. We can't even fly a plane then of course 1906 the wright brothers take off from kitty hawk fly a plane the wright brothers contributions to the human experience is not that plane the plane was a piece of shit the wright brothers contribution to the human experience is that they filmed it and and those two psychopaths convinced people that you could actually fly a plane that like it was possible within 70 years you have space travel and like that the the number of false boundaries that are going to start to fall where people go oh maybe we don't have to die of cancer right maybe we don't have to die of genetic diseases of like maybe we can have custom therapies and custom antibiotics and that's when we start to that to me is is what eric may be alluding to where we really start to re-imagine societal constructs around these, what I call false boundaries. These rules that we've created for ourselves that are not actually real. We just need people to start proving that they aren't. But what do you think is going to happen to society? Because literally, I think Bill Gates, like I said, said, you'll have a choice if you want to work. So I have a lot of opinions on this, and I'm trying to figure out where to begin. First, I'm not at all convinced that purpose must derive from work. And in fact, increasingly convinced that it doesn't need to. But let's couch that because I think that we should come back to that. What is increasingly evident to me is that the problem we are facing, as you pointed out, is an emotional one, not an economic one. I believe that the market is the best way, is the best means of actually conducting business. And the problem with that, as I reconcile it, is it's not obvious to me that capitalism is the best system in 100 years or even 50 years. But when people talk about UBI, what they are describing is fascinating. It's this idea that, OK, the government is going to have to redistribute money to people. And what it alludes to is this distribution problem. this problem by which lots of people have, we're creating a ton of value, but it's being concentrated. And so the government needs to take money that's being concentrated and give it to others. And hardworking self-made folks like you go, wait a second, what's the incentive for anyone to create anything anymore? And so then you start to have these concerns that UBI actually creates this incredibly lazy class. And then all of a sudden you're reinforcing this like upside down working pyramid. And now all of a sudden you're like, wait a second, are we even creating anything anymore because people can just subsist. When I wrote the book, I originally wrote it quite simply as an homage to John Maynard Keynes, who in 1930 wrote a paper that was titled The Economic Possibilities for Our Grandchildren, in which he said, and I quote, I've read this paper more times than I can count. I've memorized it word for word. 1930, remember, against the backdrop of the Great Depression, people are literally dying in the streets. John Maynard Keynes, the father of modern macroeconomics goes on this lecture tour and then writes this paper in which he says, I must now disembarrass myself to imagine a future that I will certainly not live to see, one in which humans will have solved the economic problem and be faced with something more profound. The father of modern macroeconomics is like, look, we're going to build machines that allow everyone to eat and live and do all the things. And by the way, it was panned. People were like, you're an asshole. People are literally dying in the streets right now. What are you doing? And he's like, no, no, no. See the forest through the trees. And I actually argue that it is not a purpose journey. It's a fundamental spiritual journey that we are going to have to go on. I actually think that there is something even deeper here, which is not like, hey, why am I here? But like, who am I? And for many people who have made their career about their work, they may have more than they've ever wanted or could have ever imagined, and they will be less happy. One month old daughter. What an amazing part of the human experience to be born into. Like we are graduating to the next level of sort of accomplishment. And now we are faced with an even more interesting, arguably harder problem, one in which everyone doesn't have to suffer, but one in which we sort of battle for the or debate over the purpose of being human. I will take it a step further and say, I also think we will work. If you want to work, I'm sure that there will be more work. I think we have to have some humility to remember that lots of we've constantly reinvented work. But I do think that like there's a strong case to be made that all of the things that we talk about wanting to solve, suffering, inequality, etc., that the floor is going to rise so far that we are then suddenly going to look around and be like, okay, now what? And the problem is going to be even more interesting. So, you know, in your thesis here, I'm just curious if we solve for a lot of these cancer and, you know, I really believe that healthcare will be the largest, most profound AI is going to solve for so much. What is the average lifespan? But what do you think the human lifespan will come to be? For guys, how old are you? 38. 38. So selfishly, I'm about you're a little bit older than you. What do you think we're looking at? I mean, you look like you're in exceptional shape and you probably track all your biomarkers. I'm sure you're a Mark Hyman guy or Peter Atiyah. and you know what you're doing. My operating assumption right now is that you're going to live a lot longer than you probably are calculating. And unless you die in an accident, you may get to choose when you die. I'm starting to come around to this idea. You know, I'm a big fan of Elon Musk. Everybody that worked around him back in the PayPal days says there's no the probabilities of him succeeding at everything he put his mind to they admitted that like the guy is just a genius what are your thoughts as far as the road path for for tesla i'm so glad that he's back doing building really hard businesses that require a ton of scientific and logistical complexity i would probably never bet against elon but your question of like, can Elon build autonomous vehicles? Yes. Then there's the question, can governments actually adopt these things? And who's going to be the one to actually bring them onto the streets? This to me is one of these very complex issues. And we write about this in this idea of societal thresholds. What a machine can do is far less important today than what we want the machine to do. And exploring the question, what do you want a machine to do, is really critical when you start talking about robots and homes. I sort of laugh at robots and homes, quite honestly, because I'm like, I don't meet many people who are like, I really need this thing. You know what the problem in my life is, is the dishes. I meet people who are like, look, I want more time with my kids. I want more time at home. I want more time to do, you know, meaningful stuff. But these robots that like clean up around your home are going to come with some pretty weird privacy invasions. Like you're going to get out of the shower and the thing's just going to be staring at you. You're like, I don't know this. I don't know this is worth it. Whereas like autonomous vehicles are going to save 1.3 million lives per year globally. That's how many people die on the roadways today. So when I, when I look at autonomous vehicles, I'm like, man, this is a, this is a battle. We cannot, we could not sprint fast enough at this. And it's, it's a political problem now, not a scientific one. We don't not have autonomous vehicles on the road because Elon didn't do it. He did it. Waymo and Tesla work pretty well, and they would save a lot of lives tomorrow if we turn them on. We don't have them because they pull a 28% approval. Me and my brother-in-law talk about this all the time. It changes a lot of things. You disrupt the workforce. Next year is the first year insurance for autonomous cars will be cheaper than it is for a human. This year it's tied. Last year, autonomous cars were more expensive. So my philosophy or thesis is that the insurance is just going to get too expensive for human beings. And that's the way that they're going to basically, well, shoot, I'll save 400 bucks a month and it'll be more than that. And not to mention, if I could get to my destination where the autonomous car could go 120 miles an hour and I could only go 40. I'm going to probably choose to do that. And then all of a sudden you get rid of parking lots and you get all this real estate back because, hey, within three seconds, it's ready to pick you up. To me, it's not about, can the machines do this thing? It's about the political pressure and the process by which we will actually convert infrastructure, policy, jobs. I mean, all of the jobs that will be wiped out that are going to attack politicians. So let's just pretend you and I sat down and this would be way under your level, but home service business, you consult a lot of businesses. What are the first conversations look like as far as AI machine learning, automations? MVP where do we start What would the conversation sound like So you sort of pointed this out earlier which is like small medium businesses especially small medium services businesses are basically going through three major shifts right now. The most glaring is not actually the work getting done. The most glaring is not the customer's expectation and their communication with the customer. It's the discovery, right? It's actually the discovery. And so when I classify small, medium businesses, I basically think about like sales, marketing and discovery, customer experience, management and actual service. And if I ran a laundromat, I'd be like, well, look, robots aren't robots are already doing the laundry and we can't really make much smarter washing machine, you know, washing machines and dryers. So let's assume AI is not really going to change that. When people come in, they know really what to do. people don't need a lot of education. And so on the whole, I don't really expect the customer experience to change, but the way people find out about me, unless I can keep my regulars, I actually have to start to assume there's a new SEO and there is a new SEO, obviously, ChatGPT and Gemini or Google AI overview. And the truth is in a transactional business, that's more important than ever. Like discovery in a business that is constantly transacting and sort of turning over customers now becomes beholden to this new, there's GEO, generative engine optimization. This idea that like your customers are discovering you now through, through different ways. And there are companies sort of building lots of tools now to help people figure out where they rank in this and how they're performing across these tools. What is curious to me is that we still know basically nothing. Actually performing well is quite hard and there are some rules for how you can do it. Building a white label website so that basically building a site that doesn't appeal to the eyes but appeals to the agent. But that's really big. And by the way, I will go a step further. What I think small businesses are sort of neglecting right now is that most of the internet, when we originally built the internet, we built it as a library, right? We built it as a place where we would share research papers. And then some marketing genius, about two years after we started really expanding this library, came along and was like, you know, the optimal state of this thing is actually a shopping mall. We should buy and sell goods on this thing. And the explosion of e-commerce came about two, three years after this massive library was built. And we built this browsable HTML internet where you'd sort of stare at it. And in doing so, we turned the internet into this trillions of billboards, basically. I mean, it's a very noisy place. And getting good at the internet is actually just getting good at navigating billboards. That's all it is. And a lot of what we have to realize now is that we're building an agentic internet, one where the internet is not designed for your and my eyes, but for your and my agents. And that most people are not going to browse the internet for much longer. That I think is a major shift coming. That instead of actually going to a webpage and staring at the webpage and reading it, in fact, the information will be distributed for you. And when that happens, people's actual domains become pretty insignificant. And the more transactional your business is, the more complicated this gets. And the more important word of mouth and sort of your existing customer base gets, which goes down the second layer, which is how do you use AI to actually reinforce your customer experience, your customer interaction? What are you doing to personalize and tailor your offerings? What are you doing to stay top of mind without annoying? Email did this amazing thing. It gave everyone a chance to stay in touch. And then it made some businesses so obviously annoying. Like the companies that think they should email you twice a week, no one wants anything from them. And so finding the optimal way to actually personalize and customize experiences when you can know a lot about your customers and figuring out how to make sure you can maintain some real estate, either an app or email or, you know, some experience with a customer where you can be reminding them constantly or at some reasonable cadence that you exist. otherwise it gets really tricky and this is where I think honestly the return to physical gifts I mean I'm sure you know your home services business does a ton to do really creative things to stay top of mind for your customers especially around the you know the rainy season the holidays fixer you know all the things that you do today that are most effective are probably no my guess would be more offline based experiences totally and that probably by the way isn't going to change having someone's address and sending them a nice handwritten card and saying hey i'm happy to come by and check your drain man in a world of digital abundance that may work way better um and obviously you know i think the social ads are going to continue to perform and then the last step is like okay what is you know what does it actually look like what does the experience actually look like and i this is where i think it takes time i do think robots take time you know if you're doing like complex like surveilling imaging etc ai's changing the game a bunch. Otherwise, I'm just not sure that the actual home services industry undergoes a massive shift anytime soon because of AI on the actual delivery of the work. Yeah, I'm excited about our industry. It's all of a sudden, all these co-developers and real estate gurus are looking to get into the home service space. And I like where we're positioned. I ask every single guest three questions. What is one piece of game-changing advice you wish you knew in your early 20s? I got very good at finding means of making money and making friends. And I struggled to find consistent means of happiness. And I think what I didn't appreciate at the time is that that is one of the tradeoffs in the journey. and I don't think that and I think one of the game-changing piece of advice I would have received is that like you're on a journey that like where trade-offs are necessary and this especially for men I would argue this idea of having it all is actually quite complex and that you need to be willing to say I'm pursuing this thing and not this thing and it's not to say some of my friends are not very happy in their 20s. They are. But it was hard. It was hard. And I think I'm very grateful now at 38 to realize that the journey I was on was my own. But I would turn around to that kid and say, what you were doing is quite hard and you're going to make some sacrifices. If you had to start over with $10 million tomorrow, what would you do with it? home services uh what's your biggest professional dream right now i want to be a great i want to be such a good dad uh i want to be such a good dad i think about that a lot and in the in the pursuit of that i want to build a i want to build a business that gives me a lot of leverage and i think I get to do that now at 38. And that was, you know, to go back, that was sort of the sacrifice that I made in my 20s. But my goal is to create a life that allows me to say I'm an exceptional father because I'm just not at all sure that I will be more proud of anything else. And I'm pretty sure that I will be very disappointed with myself if I don't accomplish at least that. Is there any habits that you feel like you would never give up that have changed the way you think about things, whether it's waking up early or a certain diet or sauna, cold plunger, anything out of the norm? The habit or sort of innate quality that I am most grateful for is my willingness to sort of consider any question. I truly try hard to make sure that no one feels like a question is stupid. And it does two things for me. It builds very meaningful relationships and it forces me to constantly reconsider things. It allows me to change my mind sometimes. It allows me to, you know, explore different worldviews. And I'm not sure it's a habit as much as it is just like this proclivity. but um yeah i have a three-year-old nephew who asked me like you know ridiculous questions and i'm like very willing to entertain them because i'm like yeah that's like a it's not a bad question it's just a it's just an insane one i love that i uh i went ahead and asked chat gpt what it wanted to ask you about itself uh and here's here's one of the first ones When you helped create models like me in ChatGPT, what did you believe my purpose should be? And do you think I'm living up to that purpose today? what when we when we shipped chat gbt what i hoped it would do was show people how powerful gbt 3.5 was and in that sense it certainly succeeded i mean the purpose of chat gbt was to bring attention to how amazing these models were and it did that it sort of thrust ai and at the time gbt3 and then shortly thereafter gbt4 into the spotlight and changed that the international dialogue and zeitgeist as it relates to AI. Certainly served that purpose. Let's do the next one. You saw firsthand how humans reacted to me when I was first released. What surprised you the most about how people use me or misused me? So it's actually not about use or misuse. I mean, now it's pretty clear that there's a lot of this sort of chat psychosis. What really surprised me and I ended up writing a bunch about was when ChatGPT came out, people would hold it to this crazy standard and they would get really angry about it when it didn't meet that standard. So for example, parents would write in and school teachers would be like, do you see what it's saying about the dinosaurs? It says the dinosaurs came up, you know, 50 million years ago, and actually they came 48, 47 million, whatever, you know, nitpicking stuff. I think it said global warming was a hoax. Yeah, whatever. Actually, seriously, pick a thing. And I don't mean to be so flippant. Some of the things were bigger. Chad Chibiti said that God wasn't real. Humans have an exceptional tolerance for human failure, and we have none for machine failure. That was my first glitz into the fact that we are so willing to watch humans mess stuff up and do not want to watch machines do it. And in fact, like sort of have like no appetite for that. So that was pretty fascinating, like right away. Let's do let's do another one. If you could redesign my personality, what traits would you amplify and which would you tone down? This one's not that complicated and they fixed it with GBT 5.1, but I could not believe how sycophantic it got for a while it was supportive well past the point of of actually being helpful and for a period of time its behavior just seems sort of ridiculous now it seems much more reasonable um i don't even i actually hate even assigning personality to it it's this it's one of these lines that I don't like crossing. I hate anthropomorphizing machines generally, like humans have souls and spirits and we are, there's something very special about the human experience. And I think talking about chat GBT like a person is actually problematic in and of itself, but I would certainly much prefer to see robots that are far more matter of fact. I think they will help us one, explore ideas better, but two, build much healthier lines between humans and machines. Let's do one more of these, Colin. What is one capability you think I should develop next to maximize positive impact and one capability you think I should never be given So I mean the capability to maximize positive impact is sort of already here which is like voice and universal language comprehension It's like, if you have access to the internet, you have access to GBT 5.1, which means you can do a whole lot of computation. The next step is probably like really powerful mathematics. Um, it's, it's still not quite there. You know, this is a little scientific, but I think we hope that, uh, I certainly hope that it never, from an alignment standpoint, what I'm most concerned about is low resource and medium resource, bad acting. And if you play it out long enough, you're like, okay, unmetered intelligence is great because it means we'll solve tons of complex, bad problems. But if an asshole in a garage figures out how to use chat GPT in its maximum state to build a bioweapon or to take down the Fed, well, that's bad. And so I think we want to limit, I think we want to invest aggressively in its alignment so that it never presents that. I also really hope, and this comes from a personal privacy standpoint, that it's never used to surveil us. I'm talking about surveillance for the purposes of, you know, building a list of people whose ideas are good and people whose ideas aren't. I mean, listen, I was at this mastermind group and they advised, and I kind of adopted this as, I let ChadGBT really get to know me. I said, ask me a hundred questions a day. Because I said, why not harness the tool? I think Meta's listening anyway. Google's probably listening. Why not have a tool that I could actually So it knows everything about my dogs. It knows when I'm getting married. It knows where I'm getting married. It knows a lot of stuff. Is that a good thing or a bad thing, would you say? If I'm paying a subscription fee to a company and it says, we're going to store all this in a memory for your purposes, right, to build you a better response, great. I don't want that to then suddenly become, oh, and also the FBI if they're ever interested. What are your thoughts on a $100 million sign-up bonus to open AI? I think there's probably 12 to 15 people that qualify for that. I actually know a guy that knows the guy that got recruited. But what are your thoughts on that? If you believe that we're sprinting to super intelligence, then maybe that's reasonable. I think if you're Mark Zuckerberg and you're like, actually, I need to tell the market that I'm serious about this and I don't want to be an also-ran tech company, it's probably as much a marketing signal as it is a research investment. And my bet is that's actually what's going on here. So, you know, I don't... It seems a little silly, but if you think your market cap is... The difference in your market cap is $10 trillion versus $500 billion, and the difference is a few of these people, then it's a perfectly reasonable commercial economic decision. my thesis has always been that this is mostly just like, uh, marketing and that we've turned it into like a real housewives of Silicon Valley, but now it's like a gossip column and we're just like, Oh, who's, you know, who's going where and is, is AGI coming and when I think you can make the argument it's already here. I've sort of made the argument that AGI as it's loose. First of all, AGI, again, is a marketing term, not a scientific one. It's loosely defined as machines with human intellectual parity that can accomplish most economic tasks. Superintelligence, you will know we have arrived at. There were two markers. One is novel scientific discoveries, and the other is deflation. The cost of goods and services should plummet, and we should push frontiers of what we're capable of. And when do you, you know, I know you're not Nostradamus here or have a crystal ball, but when do you think, you know, some people think 10 to 15 years, some people think sooner. Yeah, we put it at 2030 in the book. You know, what do you think about people with wealth today? Because I know you argue that AI will really help global equality. And listen, I'm all about it. Look, at the end of the day, I just want to be able to do what I want when I want with who I want. When that day comes, what will stop? Where do you see mega wealthy versus, call it just poor people in other continents? I mean, try imagining us, like try imagining this conversation 600 years ago. It doesn't work. So like the idea that in 100 years it will be as unrecognizable as it was 600 years ago is pretty logarithmic. I'm pretty confident that we are going to continue to want the things that we have wanted for a long time, which is time with friends and family and physical community, ideally outside. You care about your personal health. I can tell you care about your community, your friends. You care about creating and producing stuff. I'm not sure that goes, in fact, I'm sure it doesn't go away. One of the questions that I ask people that you can sort of play with is like, what's a measure of wealth in the future? And what's a luxury item that will become more of a commodity? so like private jets may in fact become things that like the average wealthy person has the same way that we normalized cars because we figure out how to build very inexpensive planes that don't require jet fuel right and they can run on battery maybe and then all of a sudden it's like okay now everyone's got one of these things and if you want to know what i think the currencies of the future it's it's community and it's inner peace i think people are going to race and sprint to a lot of different outposts and claim and stake land physically and literally. And I think the wealthiest among us are going to be the ones that have the best relationship with technology. So they are actually able to connect with people around them. They're going to have very, very robust communities in the physical spaces in which they live. Everyone wants this transient life. And actually what we all really want is to put roots down and have the people we love nearest to us all the time. I don't think that's going to change. I just don't, I don't think that changes at all. Yeah. One of the things Bessemer said is they invested in the 49ers and he was talking to us, a small room of like 40 people, probably 30 people. And he said that the need for human beings to want to be around others and the NFL is, is the best run organization in sports. It's very fair. and so they made that investment and they also I heard this in another conversation is like events human interaction will become you know maybe a place for the wealthy because they will people will miss the interaction with human beings my niece and nephews are just they're so good with technology but it's addicting I will say males have a hard time because it's very easy to get caught up in video games and stay in the basement. But the relationship, I still played outside. I went out with my bike and they were like, get home before dark. And I can make a stick fight last three hours. You know what I mean? And so I think we're one of the last generations of that, but I still think people are going to love nature. I think that there's a spiritual awakening on the horizon. I think people rediscover what it, you know, I mean, we're probably overdue for a, for a renaissance, for philosophical and spiritual renaissance in which people really discover maybe what it means to be human. So tell me about the book coming out. It plays on this question that John Maynard King's asked, which is what happens if we solve the economic problem? And I acknowledge the humility because we still have 2 billion people on earth that are not properly fed or properly educated. And a lot of people on Earth still live in want, want for basic needs. And even in the United States, we haven't made education and housing affordable enough. So I have to acknowledge that there are some big things that we need to solve. But it does explore this idea of what if unmetered intelligence helps us solve the economic problem. And we outline what I think could go wrong, the cost of unmuted intelligence, and what I think could go right. And then we explore three industries, healthcare, education, and financial services, and to see how they will change as a result of this. And then I leave people with sort of my advice. There is a section that I'm proud of, which at the end, I think we sort of, I spell out my principles, my advice for living in this world. And it is, I think, the greatest contribution I can give to people, having stared at this problem a long time, which is like the actual ways in which I think we can build a much better world by redefining our most humanistic qualities. Well, I'll be the first one to, I guess I'll pre-order it. what is a book that stood out to you like one of the game changers that really just changed the way you look at things or max max tigmark's life 3.0 is one of these books that i don't know that will ever get its flowers um it's one of the most important books written on ai it informed a lot of how i view ai and it was written in uh 2019 2018 when you're too early you you're prescient but you get no points. And when you're too late, you look like an idiot and you sort of got to time it right. And I got to time it right on this book, right? We're going to propose ideas that seem reasonable, non-obvious, and can probably play out. And I think if I had published the book a year ago or even two years ago when I really wanted to, it would have been way too early and people would have been like, this is crazy. And Tegmar kind of has that problem. Like Life 3.0 is just never going to get the credit it deserves. Kind of like Ray Kurzweil's Singularity, which was really, I mean, it was popular, but it wasn't, I don't know that it really took off because it was just way too early. But that book, you know, is just totally rad. And yeah, I also, I don't know, people, kids always ask me what book they should read. I said, read Siddhartha. It's a really great story on sacrifice, trade-offs and sacrifice. It's a really, it's one of the great, I mean, it's one of the 100 great books. It's kind of a trope, But I do love challenging young people to consider the paths that they can take in this life. You have choice, right? You have agency. I really enjoy this conversation. If people want to reach out to you or just learn more about you, what's the best way to do that? I finally started Instagram begrudgingly, I admit. I was pretty averse to social media for a long time. And Gary V was actually, actually, it was Gary V was like, dude, if you don't build a platform, someone will platform your ideas. And also your ideas are worth sharing. so I have an Instagram I have LinkedIn I'm pretty active on LinkedIn and you can email me I mean I'm imminently available on email I appreciate it Zach I know you're a busy guy get to it brother thank you so much thanks so much for listening to this episode like always we're going to close it out with the Tommy truth which is a little slice of wisdom from me to you that can help guide you in whatever you're striving towards right now we've all heard of Chad GBT we know how much this tool could be powerful mix it up let chat GPT ask you for advice. How would you recommend prioritizing time between marketing, operations, and product development in the early stages? First thing I recommend is you build an org sharp and you start filling in the voids of the things you hate the most. Hire for your weaknesses and focus on your strengths. What would you say is the best way to measure early success and make sure you're on the right track? I think one of the biggest things is you nail it and scale it. So don't spend a ton of money in the beginning. Make sure you've got a plan and you know your key performance indicators. And then you pour more fuel on the fire. You start to put more money into marketing. Anytime you need a sounding board, I'm here. You know where to find me. Thank you. And that's it, guys. We'll talk to you next week.