The AI Revolution: How AI Will Reshape Our Lives, Jobs, and Communities featuring the former Head of Go-To-Market at OpenAI Zack Kass

We should have never given kids smartphones, and we definitely shouldn't have given them social media. I am not a strong advocate for a lot of policy, but this is actually one of these places where the best thing for the kids may be to say, this is illegal until you turn 18, because we gave them something exceptionally addictive. And then we said, hey, why don't you do something instead? What if energy was free? Is a question we will soon get to answer. We will solve quantum, we will eventually solve nanotechnology, and we will probably cure all of the immutable diseases that plague our society. Because the real risk in all is not that we wake up one day and Skynet says humans should live in cages. The risk is that we start assigning tasks to machines that have terrible unintended consequences that neither we nor the machine could predict. How can you be a domino for someone else? The Joe Polish Show. Any problem in the world can be solved with the right genius network. Give it up. Zach Kass. I spent 15 years in the industry, and I spent some very formal years at OpenAI. And I was actually one of the authors of ChatGPT. One of three. And today I spend all of my time trying to figure out the answer to the question, which is what happens next? Which turns out is the question that we all want to answer. And I have two asks before we start. The first two invitations, I'll say, two invitations. If you have to be working, if you have to be on your device, I would just ask that you go outside. I will not be offended at all. If you don't have to be working, I would encourage you to put your device away. I say this for two reasons. One, we now know, speaking of addiction, that device addiction is actually transitive. When one person takes out their phone, someone else takes out their phone. And when I see phones open, it makes me feel bad, but also it distracts from people around you. The second reason I ask you to do this is because one of my political agendas, my only political agenda these days, is that we are terribly addicted to the device. People are pointing at addiction all over the world. And actually, and I will talk about this in a second, one of the worst is the device. And it is harming us in ways that we do not yet appreciate. And in a future where we actually solve a ton of problems, the hardest question we may ask ourselves is, what do we do with all of our time? And we'll talk about that in a second. So the question that we have to answer, that used to be easy and is now, quite hard is where are we? Where are we? I used to play a video. I used to be able to show people visual evidence of the ways in which AI was making progress. For obvious reasons, that became very hard because we actually had to start updating these visual AIDS almost weekly. Two weeks ago, OpenAI released OpenAI01, a new reasoning model. If you have not seen the preview of it, you should see it. It's incredible. We are at a place now, though, where visual representations of the technology are not actually sufficient to it. There are so many things happening. So the best thing that I can tell you, the simplest thing that I can explain to you of where we are right now, is that we are building machines that possess human intellectual equivalence. That is what we are building. And we are starting to build that across what we call modalities. We are entering a multimodal world. Machines were good at text, then they were good at images, then they were good at video, then they were good at speech. Soon they'll be good at olfactory. Those papers are already being written. Soon they'll be good at touch dexterity. Critically. And this goes to why I spend so much time doing what I do today. All of my work at UVA as an Associate professor of Future Studies is basically designed to figure out what happens next because of this. This is the only graph I will show you. Oh, do we have graphs? Oh, good. All right, good. Making sure we have graphs up. This is the only graph I will show you. There is a counter narrative message embedded in this. Because if I asked all of you, is AI getting more or less expensive? You would say it's getting more expensive. And the narrative in the media is that AI is getting more because it is on a training basis, which is to say GPT5 will cost a lot more than GPT4 and GPT6 will cost a lot more than Gpt5 to train. But the inference costs continue to decline. And this is critical inference is how we run the models. When GPT4 came out, it was $60 per million tokens. It was not commercially viable. That was the term we used, OpenAI. Today it's less than $3.50. And this is so important to understand for two reasons, in my opinion. The first, anytime historically a critical resource declines in cost this much, let alone this quickly, there is a Cambrian economic explosion. Very good things happen when important resources become very inexpensive. This happened with water, this happened with foodstuffs, this happened with electricity. This is happening with the Internet. This will happen with AI. We want valuable things to be inexpensive. The other reason this matters is because in a world where we are very worried about one company or a few companies controlling the technology, this is the best evidence we have that in fact the research will commoditize that the value in OpenAI may in fact be the product, not the research itself. And this is so important as we start to see the rise in open source, because this almost certainly means that we will actually live in this world where distribution is controlled on the device. It's not taxed critically, we're much further than most people realize. So the technology I mentioned earlier, this human intellectual equivalence, we are making incredible progress in it already. Which is simply to say the technology on your phones at this point is already old. It's old by any standard. Sam Altman recently said he thought he would be embarrassed one day by GPT4. I don't think that's true, but he called it the atari of AI and it probably is. Which is also to say that we are much earlier than you realize. And this begins a thread that I will try to pull for the next 45 minutes to an hour to convince you of where the human experience may go. I teach a class at uva and at the beginning of the class every semester, we ask every student in the class to pick a century between 2000 and 2000 BC and to explain the modern world, this maybe this room to someone in their century. That's the assignment. They go off and they come back with their papers. ChatGPT reads them, I don't that's true. Or a bunch of TAs. And we distribute the information. How did people do and it turns out empirically after 1900, it doesn't matter what century you pick. It's an equivalent challenge and you have to go really far back for it to be much harder. Something happened circa electricity. Something happened around 1900, probably electricity. Among other things. The rise of neoliberalism, this expansion of ideas where the world starts to create this incredible super linear path. Explaining the moon landing to someone in 1900 would be an unfathomable assignment. We had not flown a plane that would happen in 1903, 69 years later, the Apollo project. I say this because we are sitting on a lot of moon landings. Many things will happen in your lifetime today that you cannot fathom right now. Humans are very good at thinking about and conceptualizing our known universe. And we're very bad at conceptualizing our unknown universe, which is why all the aliens in the movies look like humans. This is the best evidence I have to let your brain expand and explore what might come next. So, given that, I'm going to tell you where I think we're going, and I will start by saying I don't expect any of you to believe what I'm about to say. I spend most of my time with OpenAI folks. We sort of live in this cycle. This weekend I was actually at a think tank event. I think my ideas are still out there for a lot of people. I'm not offended at all. If you think I'm crazy, I would ask that you reserve your ridicule for when I leave the room. But. But you would not be the first person to stand up and. And declare that I was wrong about something in the middle of a talk. That being said, it's much more fun if you suspend disbelief. It is much more fun if you actually let me take you on this ride. We'll start with the very easy stuff. The very easy stuff is the phases of AI integration. We're going to go through three phases of AI integration over the next 15 years. This is basically certain. The first phase is the one we're in today. It's the enhanced application phase. Joe's got an app that he's embedded AI in. A lot of you have apps that you've embedded AI in. It works because you're taking a really cool resource and improving the application experience that you may already derive. ChatGPT is obviously an AI native app, and there are plenty of those. And this is actually already introducing about 20 to 30% productivity gain in the world. This is according to a recent report that I got to do with Bain. And this is nice because it comports the world we live in. We don't actually have to update anything. We just get a bunch of productivity gains. But it doesn't do any of the things that AI actually can do at scale. Which enters the autonomous agent phase, the autonomous agent phase, or agent of AI. And by the way, I'm always curious, show of hands, who's familiar with agent of AI or autonomous agents? Nice. More and more it's entering the zeitgeist. I used to have trouble explaining it, so I would tell people it was a Siri that works. That was the easiest way that I could explain this to people. And of course, once I did that in front of a Siri founder and that was not a good thing. But Adam Cheyer's fine. I'm not worried about him. Apple Intelligence now is showing you commercials of what they want you to do. They want you to assign tasks and goals to an AI agent and have that agent execute those tasks and goals inside of your app, inside of your web browser, disintermediating you between the apps that you want to use in the pursuit of solving your tasks more efficiently. This is very exciting. It will present incredible new efficiency gains and will catapult us forward in incredible ways that we don't yet appreciate. My prediction is that in the next 10 years we will see the Internet move from an HTTP protocol to an agent based protocol. Web pages will not actually make sense because in a world where my agent just books the reservation with JW Marriott through their agent, why do I actually need to go to a webpage that probably is coming? The third phase is the one you have to squint at. And this is the natural language operating system phase. And this phase I would also bet on this phase will be marked by two major updates. The first is that you will move from the device that you are all holding in your hand right now, taking a photo of the screen, the handheld device, to a wearable device. This is certainly going to happen. It will probably be a pair of glasses or a watch or both. And my guess is that this starts to arrive in the zeitgeist soon and then actually enters our homes within the next five years and becomes popular within, very ubiquitous within the next 15. Mark Zuckerberg recently said he wanted this future by 2030. That seems too fast and I'll tell you why in a second. But the more important thing that the natural language operating system presents is my favorite promise of AI. And I explain it anecdotally. I grew up to two doctors, very smart people who didn't adopt technology until later in life or modern technology. And I remember watching my mom Google things. She would type full sentences into Google. Obviously that's not how Google works. It's a keyword based product. They've vectorized the Internet and they want you to search using keywords. My mom, expecting to interact with these technologies in a natural language way, can't. And it never really clicked. And it worked fine for her because she's a doctor in Santa Barbara. She lives a charmed life. She can go about her day. The week we launched ChatGPT, I came home that weekend. My mom was the 30th registered user outside the company and I gave her ChatGPT and I watched her for the first time use a technology that did not require learning how to use it before you used it. She could interact with this machine in the way that she expected to interact with it in the way she interacts with the world. And it was an incredible moment for me in realizing that we were truly witnessing what a natural language future would look like. A world in which machines operate in the way we want them to operate. And this is what the natural language operating system will be. It will be the first time that humans can ubiquitously say that the technology does not require major education. And this presents an incredible access. This is the best opportunity we have to actually distribute this technology to everyone. So given that something else is obviously happening, which is that the world will soon have access to fully integrated, unmetered intelligence, if the machines are getting exceptionally good and the machines are getting exceptionally inexpensive, and we are installing the machines in ways that we can sort of engage with them effortlessly. What you're describing is a world where brilliance is free. Imagine a world where ChatGPT is sort of integrated into how you live in sort of incredible ways. And this has all sorts of repercussions, and we're going to talk about them. But ultimately, what I think this means is that we're about to enter our next Renaissance. The first Renaissance was a remarkable period in time, because it was basically this time where we started doing all of these things at a supercharged scale. We had long understood some things in science. We had long understood some things in architecture. We had long understood some things in political theory. And then we started amalgamating all these ideas from all these corners of the world, and we started creating these shared understandings of these best practices. Human potential expanded, not a little, but a lot. The next Renaissance will be defined by another massive expansion in human potential. My prediction is that what one person can do today will truly pale, probably by an order of magnitude to what one person can do tomorrow. And if you want to comprehend this, think about what would have had to been true for you to run the company you run today with the resources that you run it with 50 years ago. Probably impossible. The Internet expanded our ability to be entrepreneurs, to create value. The AI revolution will do the same. But it doesn't mean that there won't be a lot of things to solve. And so the question now is, how will humans actually harness machines in the next Renaissance? And I will tell you right now, before we go any further, I am a humanist at my core. It is amazing to hear Joe and I did not coordinate our talks, but I ultimately care more about personal connection than basically anything. And I will tell you that now to provide context to what I'm about to say. One of the things that we have started studying at uva, which I think is core to understanding how this next renaissance will work. The AI revolution is the idea of a societal threshold and a technological threshold and their relationship. A technological threshold asks the question, what can a machine do? That's all it asks. A societal threshold asks the question, what do we want a machine to do? Pretty fun, pretty interesting. There are tons of examples of these. Contemporary History presents some really fun ones. My favorite is the elevator. 1854 Otis Elevator invents a new form of elevator that's basically empirically safer because it has a fail safe, so that if the rope breaks, there's a grab. And it's the first time that building owners say, yeah, we're willing to put people in this thing. We don't think anyone's going to die. So they start selling all these elevators. Six months later, all these building owners call back and they say, hey, we have a problem. No one's riding in these things. Can we return them? What does elevators know? You cannot return the elevators, but we should figure out how to get people to use these things so we can sell more. So they go back to the drawing board and they come up with three ways they think that they can convince people to use a device that empirically works. How can they move the societal threshold? One, they put a mirror in the back of every elevator so that when you walk in, you'd be distracted by your own image. Two, they put music in the elevator so when you walk in, you'll be entertained. And three, they put an operator in the elevator. They put a physical person in the elevator so that when you walk in, you know that there's a human in control. And that if this person's job is to ride up and down all day without dying, you can too. And it works. People start riding elevators, and today we all ride elevators just fine. The obvious question then is, what's our elevator? What's our elevator? It turns out we have a ton. We have more than any generation prior. We are staring at more societal thresholds that confound me than I could even name. But there is one in particular, one societal threshold that I am truly fascinated by, that we have been studying very carefully for the last year. It's the autonomous vehicle in the United States. 50,000 people will die this year on our roads. 7,000 of those are children. In a world where we talk about saving the kids at basically every political rally, it's a bipartisan issue. Why do autonomous vehicles, the obvious solution to this problem, pull at 25% approval? Autonomous vehicles are so unpopular There are state legislatures who ran election campaigns on banning them. I could not understand this. So we spent a bunch of time studying it. We came up with three reasons. We think that the societal threshold does not move with respect to the autonomous vehicle. And we think now that it may apply to. To the future, at least the short term future of AI devices. The first reason. The first reason that a societal threshold does not move with respect to autonomous vehicles is quite simply. We love control. We love control. Did someone call it in the back? Someone's congratulating themselves. There is something. I'll give you a medal later. There is something visceral about putting our foot on a pedal and turning a wheel and steering this machine. And if you actually. Some people hate their commutes so much they don't remember what it feels like. Some of you are successful enough that you've bought some really nice cars and you drive them on the track. So you remember. But if you want to know truly what it looks like, watch kids who get off the Autopia ride at Disneyland watch. This is by any account the worst ride at Disneyland. It is, first of all, a car that drives around on a track that has about this much wiggle room, that spews fumes on a concrete jungle built in 1954 in Tomorrowland, by the way, it is an eyesore and an embarrassment to Disney. And they constantly talked about shutting it down. The problem is, parents keep reporting their kids love it because their kids love the experience of controlling these machines. There is something so exciting about it. We went to Disney and watched it. It's really amazing. The second reason is that technological thresholds are now moving way too quickly. The fun thing about when we started this study a year ago is that autonomous vehicles weren't actually good enough. There were good arguments to be made. They were not empirically safer than humans. Those arguments were. Are bad at this point, just terrible. There's a separate problem, which is the asymmetry of information isn't just strictly reserved for the technological threshold. It turns out 20 times more Americans are afraid of flying than driving, despite one being empirically very safe and one being empirically very dangerous. So in a world where we aren't even risk adjusting on that basis, but how could we possibly update to what technology is actually capable of doing as technological thresholds move faster and faster? The third reason, though, is the most important. It's the one we're circling bright red. It's the one that I get to talk to a bunch of large enterprises about. It's the one that you're going to think about for the next couple nights as soon as I tell it to you. Humans have an exceptional tolerance for human failure. And we have none for machine failure. This is the thing. This is the thing. This is why we can watch as drunk driver after drunk driver kills kids. But when a Tesla swerves into the wrong lane, the vitriol is off the charts. How dare Elon get him in front of Congress. Let's get to the bottom of this thing. It's bad. It's incredible. And this isn't just reserved for autonomous vehicles. I will remind you there are so many places in your life where you say that risk is too great. So instead I'll assume this risk. There are two reasons for this, that we've actually started a study. And I'll hint at this because the book comes out in February and we actually. I haven't published any of this. We're talking about it. But we've actually narrowed down the two reasons we think this is true. The first is humans assume that if you observe a machine fail once, it will fail systemically. It is a signal that there is much greater failure. Rot. Machine rot is a term in the industry that's actually inversely true. The cool thing about machines is when we watch them fail once, we can actually stop them from failing again in the same way. That's really cool. That's also limited to machines. Humans do not have that feature. We get to fail over and over and over despite watching our neighbors do the same dumb thing. The second reason is we have no empathy for machines. Ray Gun is the exception that proves the rule. Never in Olympic history has someone who has done badly been chastised. Ray Gun broke the game. She was so bad that people truly hated her. But until Ray Gun, truly, the losers were often exalted almost as much as the winners. We feel for the figure skater who falls on the ice on her final jump. Outpouring of love. I'm so sorry. How, you know, this is. This must have been so hard. We're behind you. The country loves you. When a robot breaks, we go, fuck that thing. We have amazing empathy for each other and none for a machine that is performing the same task. Go home and watch Wall E again. The very incredible thing about Wall E and the thing that I love about this film, and quite honestly, the new robot film that just came out, which I was crying half of it. You watch this movie and you're like, I have feelings for this machine. And it's very interesting because we're not Wired that way. Very fascinating. So given that we started actually talking about, okay, there are obviously a bunch of boogeymen. There are a bunch of problems that we're going to point at and say this can't happen and this can't happen. And they aren't actually risk adjusted problems. They are false problems, they are false positives. But there are a bunch of things that we should be talking about. There are a bunch of things that we should actually start looking at and saying, hey, the AI revolution is going to present some very interesting risk. And these are the four camps of risk. I think the four ways in which the AI revolution may complicate and compromise our lives. Number one is my theory of idiocracy. This guy gets it. What's funny about the theory of idiocracy, first of all, it's the theory that says at some point we will build systems that are so smart, that do so much more for us than we used to do for ourselves, that we actually start to devalue critical thinking and devolve intellectually. Someone in the room right now is whispering to their friend, we're already there. And I assure you we're not. That we are not already there. You would know if we were there. We're not. We actually live in a world where people are getting smarter all the time. You are all much smarter than your parents, they are all much smarter than they were, much smarter than their parents, and so on. This is how the world works and I can prove it in all sorts of ways. That being said, it's still a material risk, it's something to consider. And we are seeing some somewhat strange behavior in Gen Z's test scores. But where Gen Z is actually exhibiting the most strange socio cognitive patterns is in the second second theory, which is the theory of dehumanization. This theory says that at some point we will live in a world where a material percentage of the population has more interest in their virtual reality than their physical one. And this I do worry about much more than idiocracy, because this one we are starting to see. The patterns are there. There are trends among the younger populations of incredible rise in anxiety and depression, a strong discomfort in physical spaces and physical settings, and an incredible addiction to their virtual ones. This is a real risk. And I will just tell all of you, these are my. I'll just show my colors. I basically spend a lot of time telling anyone who will listen, we should have never given kids smartphones and we definitely shouldn't have given them social media. And I am Not a thank you. I am not a strong advocate for a lot of policy. I think that we should unwind much of it. But this is actually one of these places where the best thing for the kids may be to say, this is illegal until you turn 18. Because we gave them something exciting, exceptionally addictive, and then we said, hey, why don't you do something instead? They're 16 years old. What are they going to do? How do you want them to opt out of a system that is more addictive than actually anything else we've ever introduced to them? It's really bad. The average teen in the United States today will spend five and a half hours on social media. The average American male between 19 and 23 will spend two and a half hours today playing video games. That's up from an hour in 2021. We thought it was a pandemic bump. It's not going away. There's a real issue here, and it actually dovetails perfectly in what Joe is talking about to all of you, which is that we are creating incredible ways to distract ourselves from the things that matter most to us, which, it turns out, is time with friends and family and physical community. And one of the things that I spend time and I will just tell you all, is advocating for at a municipal level is investing in physical infrastructure. People are like, what should we do to prepare for AI? I'm like, build a park. Build a place that people want to be outside. AI is going to do a lot of great things. It will solve a bunch of interesting problems. You know what you need to do? Subsidies for good restaurants. Create spaces that people want to live in, because otherwise, we will wake up one day. We will have a world of abundance, and we will live on it here. The third thing I worry about is job displacement. Who here has worried? Now, most of you are, I know, CEOs and owners of your company, so maybe this doesn't apply, but I will ask anyway. Who here is worried that AI might one day take your job? That's. Yeah, that's pretty confident. I would have expected that. Who here is worried that AI might take someone's job in this room? Seems much better. That, by the way, is a result of the zombie apocalypse phenomenon, which basically says that everyone believes they would survive the zombie apocalypse and their idiot neighbor would certainly die, which I actually want to study. We don't have. There's just not enough funds to go around. But I would love to explore that. I also worry about job displacement. I worry a lot about job displacement. I don't worry about it for the reason most people worry about it. When I ask people, are you worried about job automation? Their first consideration is the economic consequences. In fact, it's basically the only consideration. What will we do for money if the robots take our work? How will the food arrive on our tables? This is misguided and actually dangerously so for the following reasons. First, industrial revolutions are very good for the world. Industrial revolutions do a lot of things empirically. One is they create more jobs. Unemployment declines. Two, they create better jobs. Wages go up, job safety goes up every time across the board. Industrial revolutions are very good for the world economically, and you cannot argue otherwise on basically any measure. The problem is they are not obviously good for the world emotionally. There is incredible amount of cost to industrial revolutions as a result of people feeling like they have been harmed in the process. And that harm derives very clearly from the fact that we attach more and more of our identity to our work. And if our work changes, it doesn't matter if there is more and better food on the table. We feel aggrieved more perhaps than we feel capacitated for in the form of economic gain. And what's fascinating about this is we recently observed this in full effect. The dock workers unions of the United States stood up and said, you cannot automate our work. And we called a few of them, we called a few of the regional heads, we spoke to these people. They all agreed that there was incredible, incredible surplus of jobs in semi skilled and skilled physical labor available to them. They also all told us that they wanted to be a dock worker. Many of them told us that their fathers were dock workers. Many of them told us that their sons were going to be dock workers. Taking that work away from them was the problem. In a world where we start to automate our jobs, the cost is not an economic one, it's an emotional one. We're obviously now facing an identity displacement crisis. This is going to happen in mass. And the real issue in all this is if we don't actually start to find identity in the things that we know matter most to us, it will truly be an epidemic. In a world where you ask someone to say what makes you happy, and the only thing that they can think of is their job, we're obviously doomed, at least in the short term. The last thing I worry about is existentialism. Who here is worried that AI might kill you? That seems pretty good. Who here is worried that AI might kill someone in this room? That's much better. I also worry about this too, but I don't worry about it. The reason most people worry about it, that won't surprise you. Many people, when I ask, do you worry that AI will kill you? They think of the Skynet. They think of Terminator, think of. Very smart people ask me if Skynet's possible. I have to remind them it's a movie. But I do worry about it. And I worry about it for a reason that is slightly more nuanced. That is actually the scientific theory. And I can explain it the following way. If I put a wheel on this device and I train it on a very simple task, which is to fetch me a bottle of water outside those doors, and I put it on the ground, I push go, and it rolls off, and it comes back with a bottle of water. All good, right? Problem is, I don't actually know what's happened outside those doors. And it could be pretty bad. There could be tables flipped over. There could be a bunch of water spilled everywhere. A JW employee could be huddled in the corner or used to be holding this water bottle. I didn't want for these things to happen. I'm not a bad guy. I just didn't specifically say to the machine, be nice. Do nice things on your way to get this water. And we call these unintended consequences. And the incredible thing about AI now is that in a world where machines are getting exceptionally good and machines are getting exceptionally inexpensive, the most important thing we can figure out how to do is making sure that machines care about us. And this is called the alignment problem. And the alignment problem asks the most important question that we must now answer, which is, how do we train a machine to be effective at accomplishing a task and concern itself with the planetary experience, the cost and benefits of its actions predicted or un. And the really amazing thing about the alignment problem is it has an organic allegory for all you parents in the room. You are raising your children to be brave and capable and effective and cunning. And you're raising them to be kind and caring and thoughtful and compassionate. And it's not sufficient to say to your child, always follow all the rules or always break all the rules. Always put yourself first. Always put others first. These platitudes do not work. You know this. And you can't stand next to your child 24 hours a day and tell them exactly what to do, despite what some of you might prefer. So you raise them on a set of values. And these values are meant to inform the millions of decisions that they make every day after the age of three, when they come into their Ego that helps them calculate the cost and benefits of their actions, not just to themselves, but to other people. We are fascinating creatures in that regard. And there's two really good pieces of news. One, we are very good at alignment. Humans get better all the time. I can also prove this. Despite what people might gutterly feel right now, we get better at alignment all the time. That's why the world gets better. The other piece of good news is that the smartest people in the world are trying to solve the AI alignment problem. And we're actually making incredible progress. We are building machines that have a remarkable sense for predicting and understanding unintended consequences that often we couldn't make sense of. Because the real risk in all this is not that we wake up one day and Skynet says humans should live in cages. That's not actually the scientific theory. The risk is that we start assigning tasks to machines that have terrible unintended consequences that neither we normally nor the machine could predict. That's the problem, and that's the thing we're now trying to solve. Given that I've sufficiently scared you and you're probably wondering how this could possibly go, well, let me tell you because spoiler alert, it probably will go very well. All of the indicators, as best I can tell, suggest that we are on an incredible path. And there are going to be short term and medium term ways in which the AI revolution materially reduces suffering and increases joy. As long as we get these things right, the first thing that's going to happen, and I'm basically now certain of this, is this massive expansion in human potential. What one person can do today will pale in comparison to what one person can do tomorrow. Why does this matter so much? Well, it obviously matters from a GDP and economic standpoint, a massive boon in entrepreneurship seems almost certain. But something else is true, which is that we are clearly rate limited in a lot of our understanding of the world by the number of people actually thinking about a problem. There are about 10,000 oncological researchers in the world. When we talk about I hope they cure cancer, we're literally talking about a people that we could shake a stick at. Like, there is a group of people in a couple of rooms around the world trying to solve one of our most important problems. And when we're like, well, why aren't there more? It's like, yeah, let's go print a bunch of oncological researchers. Let's go send people to school for 20 years. It's really hard. In a world where we expand human Potential. It's not just about expanding the economic value of humans. It's about expanding the intellectual value. We may soon answer the question, what if we have a billion people studying these cures? Which actually leads me to the obvious next step in this process, which is a wave of scientific discoveries and breakthroughs. This is going to happen. We are very close, by the way, to solving fusion energy, for those of you who are paying attention, and if you think AI presents a weird philosophical question, Abundant energy will rock your world. What if energy was free? Is a question we will soon get to answer. But it's not just limited there. We will solve quantum, we will eventually solve nanotechnology. And we will probably cure all of the immutable diseases that plague our society. We will cure cancer. We will cure neurodegenerative disease. This is going to happen. Yeah. Gets me out of bed in the morning. The other thing that obviously happens as a result of this expansion in human potential is deflationary. And this is again, something that I don't think is fully priced. The cost of goods and services come down all the time. CPI feels very high right now. So we talk about living in a more expensive world, but the reality is the world has never been less expensive to live in. And the easiest way for me to describe this to you is that our grandparents, their generation spent on average 40% of their wages on foodstuffs. We spend on average 10%. The phone in your pocket would have cost $20 million 10 years ago. It would have been prohibitively expensive. Today, you lose it, you buy a new one. The average American will travel outside the country on vacation for the first time in our country's history over the next five years. You don't have to like spirit and frontier. In fact, you can make fun of them. I actually had flu spirit here from Dallas. I didn't like it, but at all. But a lot of people do because it makes traveling effectively free. It's incredible. It's a $70 flight from Dallas to Phoenix. You don't have to ride it, but a lot of people who want to move around the world do. And that's incredible. We are deflating the costs of goods and services all the time. And this will happen on a massive scale, in part because of scientific discoveries and in part because of major breakthroughs in material sciences and robotics. This is going to be amazing. The last thing that happens, and this is a fun one that is both very controversial and has exceptionally high variance. I think we're about to get a bunch more time. I'm like, pretty certain of this at this point. Although lots of people feel very upset by this, by the way. I have a lot of opinions, as you can tell, about how the future might work. I believe we might live thousands of years long. I believe all sorts of things. This is the one that really upsets people. Of all the things that I believe, this is the one that really viscerally riles people up. I have people stop me in the hallway and say, there's no way we're gonna get more time. And if we do, it'll be terrible. It'll be so bad for us. They're usually white men who work in it, by the way. And I have two responses to them. I have two responses to them. My first response is I. And so they'll tell me, right? The Internet promised to make us more productive. It promised we would work less. I work more than I ever have. I work more than anyone in my lifetime ever had. What gives? I remind them two things. One, the average worker in 1960 was 57 hours. Today, it's a little less than 45. We do work less, just so we're clear, in part because productivity in the United states is up 70% since 1995. We produce a lot more on a per capita basis. The other thing that I respond with is I ask them to see their screen time. I asked these guys to show me their screen time, which, by the way, goes about like this. Why? Why? My phone's dead. I would have more luck asking to see someone's W2. I am basically certain. And it's not actually a huge surprise. It's not a huge surprise because we have an incredibly uncomfortable time reconciling our addiction to a device that we are constantly telling other people to spend less time on. It is very upsetting for us, too often for many people to look at their screen time. And we certainly don't want to do it in front of a stranger. We do not want a stranger looking at the things that we do on our devices. And my theory here is that we actually feel like we work more because we've just become trapped in our machines that we built these things that gave us incredible productivity gains. And then we found ways to play Candy Crush. And it's bad. It's really bad. We talked about this. So my first challenge to this room, two things are going to happen in the future. One is you are going to get the choice, I think, to spend your time deliberately. And I strongly challenge you to build a better relationship with the technology you use. The second Thing that's going to happen is we are going to build more addictive systems that are trying to capture your time and you should be very mindful of those traps. That is I think, the long term real addiction threat here. Okay, Given that, I'll talk quickly about how you can prepare and I care about how you can prepare as entrepreneurs, so I will talk to you about that. And as business people, I also care about how you can prepare as members of your community. Obviously I care a lot about your children, the next generation. So I'll talk to you as parents too. The first thing that has to happen is you must now prepare for tremendous change. I am embarrassed by this slide, by the way. This is such a trope. It's such a. Sir, please don't take a photo of this slide. I don't want to be associated with this. I put this up. I put this up anyways. I put this up anyways because it's true now. It's always been true and now it's very true. We are clearly accelerating. And I just feel like I have to stress this, that what is true today will not necessarily be true tomorrow from a technological threshold. And we are going to change a lot. That means, in my opinion, one of the things that we're now studying is this idea of embracing adaptability as a strategy. This is like core to the work that we do with large enterprises. Many enterprises, when you ask a CEO how do you evolve, they talk about their culture. We have a culture that allows us to evolve and they talk about it as this like second order consequence of other things. My prediction is the future will basically be determined by companies ability to evolve and almost nothing else. Because in a world where we start to experience incredible change agent events like the Internet or AI and then what on a faster and faster timescale. The thing that will define a company's success is not how big it is, but how quickly it can update to new information. I'm basically certain of this now. We will see more change in the Fortune 1000 over the next 20 years than we saw in the last 100. Definitely. Because companies are just not going to update to new information. And given this, the thing I'm now talking to individuals about and managers about is this idea of learning how to learn. One of the ways that you can embrace adaptability as a strategy is by telling your people in your organization, telling your leaders, hey, we, we need to be looking around the corner. There is an incredible class of people who don't want to look around the corner because they forgot how to learn a long time ago. They are good at a very specific thing. It is attached to their identity. And the idea that they must learn a new thing before they retire is very unsettling. Right. This is their thing. This is what they are good at. They were planning to do this thing and then retire. That presents a huge problem because that thing may go away. There may be a whole new set of things. So learning how to learn becomes the most important thing. Kids ask me all the time, all the time. They email me all the time, what should I study in college? Then they'll usually be like, can you introduce me to Sam? Can I get a job at OpenAI? And I tell them I have this very disappointing answer. Cause parents ask me the same thing. What should my kids study in college? It doesn't matter. It doesn't matter anymore. No one has figured it out. It just doesn't matter. There is a fast declining correlation between your major and your economic outcome. And in fact, I will make the argument that in 20 years it won't matter if you go to college at all. It probably just won't matter. Yeah, it's worth it. Fine. Did you go to college, sir? Yeah. Well, yeah. Okay, but the other guys don't need to. But this is, this is, this is a fact. We are basically at a point where the system is so heavy now, it's just, it's bloated, it doesn't make sense. And we're going to figure out alternatives to higher education pretty obviously. But in a world where you're learning how to learn, my position is study something. If you go to college, that matters to you. Become fascinated by the pursuit of learning. I don't care what you're learning. The act of maniacally obsessing over information and piecing together complex thought is sufficient to proving that you are capable of doing it again and again. Don't study something because it's going to make you a bunch of money. It probably won't study something because it will help you learn how to learn other things. The last thing we talk about. And then I will. Then I will break designing around things that AI cannot do. One of the things that we spend my favorite time in our research doing is what we call the idea of negative space. AI is negative space, everyone. The second most popular question I get asked, what will AI do next? Totally reasonable question. I love answering it. It's very fair. No one ever asked me, what will AI not do next? What is OpenAI not gonna build? What are machines going to Fail at. And this is becoming more and more obvious, especially in settings like this, in a place where we could have all gathered on zoom from the comfort of our homes. We come here together and spend time in a physical space together to learn. Machines are not going to replace this. We tried during COVID it went badly. Something is very real about the human experience. And so my point to people, when they talk about planning for the negative space is optimizing for humanistic qualities in a world where our knowledge and our skills commoditize, and they probably will. We will be separated not by what we know, but by who we are and how we make people feel. This is basically certain at this point. And it's already happening. It's already happening. And that's the coolest thing. All of you, I'm sure, have so many examples in your companies of star performers being great at their jobs, not based on their knowledge and skills, but based on their interpersonal activities. This is already happening. Professional services. By the way, Lost Alpha a long time ago. Your wealth manager does not make you more money than another person's wealth manager. Sorry, it's true. You choose your wealth manager because you trust them. There are exceptional wealth managers in the world and we work with them because we like them. Great real estate agents don't promise better home prices. That's not what they can do. They promise a great experience in buying the home. The experience matters. And of course, a thing happened very recently that I now talk about. A month ago, my father received a lifetime achievement award from the Breast Cancer association of America as an oncologist for the work he had done in Santa Barbara over the last 30 years treating breast cancer. It was profound. I cried, basically. As you can tell, I'm a crier. I cried a lot of it. At one point, a woman stood up to talk about my father and she talked for five minutes. For the first 30 seconds, she talked about her patient outcome. She talked about how my father had saved her life. She had survived breast cancer and gone on to live a very fulfilling life. For the next four and a half minutes, she talked about how my father made her feel. Obviously, it comported to the world I believe in, but it was someone saying it in such an incredibly real way. And then she closed with something I will never forget. She reminded everyone in the room that oncology had basically become optimized outside of clinical trials. When you show up at a doctor's office, they get a set of reports and the math tells you what you should do from a therapy basis. We've gotten there. She believes that my father saved her life not because of the therapy, but because of how he treated her. And then she said the thing I will never forget. In a world where anyone can treat you for any disease, the bedside manner is not a feature, it's the product. This is where the world is going. I firmly believe this is where the world is going. In across services and eventually elsewhere. In a world where you can be anything I tell people, be adaptable, be courageous, be curious, be kind, be wise, be empathetic. Be the things that we know. AI will not be the last thing I'll say. And then I'll leave you. The future is great. Today is the best day ever to be born. I'm sure of this. If you don't believe me, I will prove you wrong. Outside you're wrong. It just is. Empirically. It's an incredible time to be alive. And tomorrow will be better. Most people don't believe this because we have basically live in a world where we are overexposed to negative information. This is how it works. We are underexposed to positive information. We are overexposed to negative information. Parents especially have high anxiety right now. They're worried about their kids futures. I understand why and I have to remind them that their kids futures are so bright they will explore the world in ways that we could never have imagined. Then my last challenge to you is go back to your dining room tables, go back to your places of work and talk about a world where we may cure cancer, where we may cure neurodegenerative disease, where energy becomes abundant, where every child has an AI powered tutorial, where every person has an AI powered primary care physician. We are moving in this direction. The indicators are flashing green. Everyone in the world wants a reason to be optimistic and I basically think it's the missing ingredient right now to most of our happiness. If I can be that reason for you today, I am so glad tomorrow the next day be that reason for someone else. Thank you so much. Don't miss another episode of the Joe Paulish show. Subscribe today@thejopolishow.com if you'd like access to the show notes to take action on what was discussed, please visit thejopolishow.com 8.