“AI and Our Economic Future” with Professor Chad Jones

Speaker 0:00 [on-hold music]

Speaker 0:05 Really, really happy to have a chance to, to share this talk with you. This is something that's, um, been on all of our minds. Uh, my research for the last fifteen years and, and longer has been on economic growth, and I feel like, uh, you know, AI is this incredible new technology and how it shapes the future and, and, and what kind of, uh, consequences it may have for us and, and our children, um, is something I think all of us are thinking about every day. So this is based on, I think, four or five research papers that, that, uh, I've been working on over the last couple of years. So I, I think it's an easy statement that AI is likely to be the most transformative technology of, of our lifetime. Um, importantly, it's the latest in a line of transformative technologies, right? So electricity, uh, the transistor, semiconductors, uh, information technology, the Internet, are all transformative technologies. And a question I've been pondering, and I'll get to this i-in a little bit, is, you know, to what extent is AI different, and to what extent does it share features with these earlier transformative technologies? And I think in, in terms of the difference, the next bullet point, uh, which Sarah echoed a little bit, I think really gets to the heart of it. What if machines, AI for cognitive work, and then AI running robots for physical work, what if they can perform every task a human can do? What does it look like to live in that world? Um, and, and to start out, I wanna lay out for you two scenarios that I think of as, as two extremes. Ne-neither one is, is probably what's gonna happen. They're kinda caricatures in a way, but I think we learn something by thinking about these two scenarios, and then I'll, I'll, I'll sort of show you some research I've been doing to help me think about where in between these two extremes we, we might end up. Um, okay, so the, the first scenario is gonna be AI dramatically accelerates economic growth, the foom scenario of Silicon Valley, which, you know, we, we read about, uh, practically every day. Uh, the second scenario is where AI is just a normal technology. AI is business as usual. Um, uh, and so, you know, maybe it's normal the way electricity and semiconductors and the Internet were normal. They were transformative technologies, but, uh, yeah, so, so we'll see. So, so let me, let me dive into these two scenarios. So, so in the first scenario-- And I should say, w- my plan is to talk until five forty-five and then take fifteen minutes of questions.

Speaker 2:36 Um, so hopefully that'll, that'll work well. I got lots of clocks here to keep me on time. Um, so AI dramatically accelerating growth. I think this one we're in the middle of watching now. So, you know, Dario Amodei, Sam Altman, Demis Hassabis, Geoff Hinton, sort of the luminaries of AI, have been saying for the last ten years that these things are coming, and we're kinda marching along the schedule that they laid out for us ten years ago. The first part of that, that schedule, I think, is AI automating software. Right? AI automating software engineering. And we saw back in November, I think, when, uh, Claude Opus 4.5 was released, Anthropic took this model, and whenever they're trying to hire a software engineer, they give the software engineers a, a, a two-hour take-home exam. And they see how they do, and that's part of how they decide who to hire. They gave this same two-hour exam to Opus 4.5, and it scored higher than any human in history. Right? That was already, you know, uh, you know, s-seven months ago. Um, and the models have only gotten better. We're up to Claude Opus 4.7 now, right? Two, two generations later. Um, in the next decade, is it plausible that we'll have AI agents that can automate most coding? Yeah, maybe that's next week. I mean, this seems like it's coming very soon, but I mean, you know, maybe it's a decade. Um, okay, when you have AI agents that can do everything a software engineer can do, well, you put them to work doing more things. In particular, you put them to work on AI research, build better algorithms to improve the AI itself, and build agents that can use a computer the way a human can use a computer. Right? And so again, shortly after that, it seems plausible that we'll have, uh, agents that can function as virtual remote workers. Anything you could call up a colleague on a, on a, on a virtual Zoom call and ask them to do, the colleague could be an AI rather than a human. Right? Once you have that, and again, shortly after maybe, well, you can scale these things up on millions of GPUs, right? And you can end up with billions of virtual research assistants, each running a hundred times faster than we run. And you put these to work, these, these country of geniuses in a data center, as Dario Amodei called them, you put them to work discovering new ideas. Right? You tell them, uh, "Help me design better computer chips. Simulate the real world and help me design better robots so we can finally get the grippers that are just sort of the, the, the bottleneck for robotics. Um, help us design better technologies, new pharmaceuticals." AlphaFold, you know, was, was, you know, nearly ten years ago now, and it's transforming pharmaceuticals, but there's so much more that can be done. Right? If you've got this country of geniuses in a data center, what virtual cognitive tasks can they not do? Well, once they're designing better robots in virtual reality, we test them out in the real world, and again, eventually, maybe it takes a decade, but eventually, we have robots run by these super geniuses in a data center, and then we've automated physical tasks as well. And once you automate, you know, cognitive tasks and physical tasks, yeah, in the growth models that I like, that I wrote down, that I taught many of you, uh, growth explodes. So this explosive growth is something that absolutely can happen if this story is right, and it's not obvious where this story breaks down. So this is a scenario

Speaker 6:08 that I think, um, is one that's very popular in Silicon Valley now. It's totally plausible. There's a question of the horizon. Does it happen in three years or five years, like the AI 2027 people, or, uh, Leopold Aschenbrenner, situational awareness, um, or does it happen in twenty-five years? Well, in some sense, if growth is exploding and accelerating, you know, five years versus twenty-five years is still transforming, uh, the world. Okay, so that's a scenario that's very familiar in Silicon Valley. I do think it's plausible

Speaker 6:38 and has some merits, but it's certainly not guaranteed that this is the way things happen. Let me give you a scenario at the other extreme. AI is just sort of a business-as-usual technology. And here, I think the story, uh, goes, goes the following way. Um, let me show you a graph. And again, if, if you ever took my class, you saw this graph thirty-seven times, so apologies, but now you get to see the updated version. It's changed so much relative to when you took the class, right? [laughing] Okay, so what is this? This is average living standards in the United States. Real income per person, hundred and fifty years on a logarithmic or a ratio scale, right? And what you see is you never get too far away from this straight line with a slope of two percent per year. Living standards in the United States for a hundred and fifty years have risen at two percent per year plus or minus a little bit, okay? Now, what's, what's especially interesting about that is when you reflect on the transformative technologies that entered and diffused throughout the US economy during this period, right? In 1870, the electrical transformation of the US economy was just a, a, a glimmer in Thomas Edison's eye. Over the next fifty years, this transformation happened. Internal combustion engines, jet airplanes, right? Um, antibiotics, vacuum tubes, transistors, semiconductors, information technology, the Internet, all of these amazing technologies that absolutely transformed living standards and yet the growth rate was always two percent per year. So this raises a big puzzle, I think. How is it-- How can it simultaneously be true that these technologies were wildly transformative and yet growth rate's still two percent? Well, the answer, if you think about it, is that what is the counterfactual, right? I think the way technology works is that within any technology class, ideas get harder to find. The steam engine runs out of steam, and so that if all you had was the steam engine and you didn't discover electricity, growth would have slowed. Or if all you had is electricity and you didn't discover internal combustion engines and semiconductors, growth would have slowed. And so the straight line would have bent. The arc would have bent here. And instead, what each of these technologies did is they allowed two percent growth to continue for another fifty years, right? And sort of the, the pessimistic in quotes scenario about AI is maybe it's the latest transformative technology that lets two percent growth continue for another fifty years. Okay? So you can see how these are two very different extremes. I think they both have a lot of merit. O-one other point I'll say about, about sort

Speaker 9:41 of a lesson from economic history, um, and you, you see this a, a little bit in the discussion I just gave. Economists who've studied, um, how steam power got replaced by electric power or how information technology diffused throughout the economy have highlighted repeatedly that these transformations take decades. You have to reorganize the factory when you go from steam power to electric motors, right? Instead of one shaft through the middle of the factory, you can now put motors everywhere, right? Or information technology, right? You, you have all these complementary innovations that need to be made, right? You invent the spreadsheet, you invent the word processor, you invent the database, you invent SQL, right? And so you have lots of complementary innovations that need to be integrated and, and production needs to be org-reorganized. That takes a time that's measured in decades. So, uh, again, this sort of lesson from economic history is don't be so quick to assume transformative technologies bend the curve here. They're bending it relative to growth slowing down, and it may take a while. Those are the lessons. Okay. But still, we've got this question. Okay, both of these scenarios have plausible aspects to them. Where are we gonna be in between, and how should we think about that? And that's the question I've been thinking about for, for the last year and a half or so. And the, the concept that I've come up with that I think is really helpful is

Speaker 11:13 weak links. Right? A chain is only as strong as its weakest link. And let me give you the, the, the application of that to, to business. Right? Business success requires completing many, many tasks successfully. Right? So if you wanna release a new version of the iPhone, you have to design it. You have to, uh, source all the parts. You have to get it manufactured. You have to make sure the manufacturing is within very, very exact tolerances in every place. You get one thing out of place, the whole thing falls apart, right? Once you've manufactured it, you have to figure out how to get it delivered in a timely fashion. You know, hundreds of millions of iPhones delivered, you know, every fall. Um, you have to handle the retail, the advertising, et cetera, et cetera. If any of these tasks, uh, falls down, then a lot of the value gets lost, at least in the short term. Okay? And that's what a weak link model is, right? And so, uh, the Space Shuttle Challenger- Right? Explodes because a $25 rubber seal, the rubber O-ring, fails. One small part fails. When you think about ASML and TSMC, the, the, the manufacturer of these computer chips, with machines that are just incredibly complicated, with precisions that are just incredibly tight, right? One little error there and, and your chips don't work, right? And so, so, so how does this help us think about these two scenarios? Well, again, back to the chain. If you've got a chain with 20 links, and you make 17 of them really, really strong, that helps, but it doesn't in the end really change the overall strength of the chain because there are always three more weakest links. And so le-let me give you an example of that, that I find really, really helpful. In your pocket, in my pocket, we have a computer with 100 million times the transistors than the equivalent of us had in the 1970s. Okay, so, so i-in your pocket, you have 100 million times the transistors than the equivalent of you had in the 1970s, right? I'm not 100 million times more productive at research, right? Why not? Well, my computer can invert matrices like nobody's business, but I have to figure out what data to put in those matrices, or I have to figure out what questions to ask, what theory to test, right? It's a weak link problem, right? You could be really good at one thing. Instead of 100 million times more productive, maybe I'm two or three times more productive. That's great, but I'm limited by the other weak links that haven't changed, right? So this is, I, I, I think, a really important insight about how the world works, and it's gonna help us understand where we are in these two scenarios. Let me highlight one other thing that I'll come back to in, in, in shortly. Weak links are the source of scarcity. One of the key lessons of economics is scarcity is what gives rise to high returns, right? What is the scarce factor? What is the weak link, is a question we should always be asking when we're studying some economic phenomenon. And so when we come back to what's gonna happen to the income of our kids, thinking about this, what, you know, what is scarce, I think is a very helpful, uh, framing for that question. Okay, um, ah, one, one more, one more graph that I like a lot. So, uh, again, i-if you, if you took my class, you may remember this, economists are infatuated

Speaker 14:45 with the question: Who gets GDP? Who gets paid, right? What share of GDP is paid to labor versus is paid as a return to capital? Turns out, for 75 years, it was remarkably stable. Two-thirds went to labor, one-third went to capital, okay? In the last 25 years, interestingly, the share paid to labor has fallen by about 10%, share paid to capital has gone up, and economists are, you know, obsessed with understanding this. Um, we have two theories. One is automation, I'll come back to that in a second, um, and the other is market power, concentration. Maybe big firms are bigger, and they're exercising more market power. They're able to capture more of the GDP as profits and pay less to labor, okay? We don't know the answers to that. Uh, I'm not gonna answer that today. Instead, I wanna, I wanna take that in a different direction, which is, okay, two-thirds is paid to labor, one-third is paid to capital. But you can break labor and capital into their components. How much of the GDP that's paid to labor is paid to people with more than a college education versus less than a college education? Or look at the capital income. How much is a return to buildings and structures versus equipment? Or within equipment, how much is a return to computers, okay? So what share of GDP is paid as a return to computing power, and how has that changed over time? Okay, turns out if you look in the right spreadsheet, you could just look this number up on, you know, the, the, the BEA, BLS website. But you have to, you have to look, you have to dig, okay? What does it look like? Um, on the one hand, computers are everywhere. On the other hand, the price is falling like crazy, right? It's a P times Q, where Q is going up and P is going down. So which effect dominates, okay? Here's the data. During the 1990s, during the dot-com boom, the share of GDP pays a return, a return to computers went up. It peaks in 2000 at just under four and a half percent, and since 2000, it's fallen by a third to 3%, right? Computers are indeed everywhere, and yet they're paid less as a share of GDP rather than more. The price decline dominates the quantity increase. This is exactly what a weak link model predicts, right? Computers are plentiful. They're the most plentiful thing in the economy. Everything else, humans, are scarce, right? And so the computer share has gone down. So when we think AI might automate everything, and we worry, "What are all of us gonna do?" This graph is at least something to take into account, right? Computers are getting less of GDP, not more of GDP, even though there are, you know, 100 million times more transistors in your pocket. Okay, uh, so, so what do we do next?

Speaker 17:46 We write down a model to try to study this, these two scenarios, okay? And the model features ideas as the source of long-run growth, Paul Romer's Nobel Prize. Um, it features a production function for goods and ideas that both involve weak links, right? The production of everything, in my sort of view, involves weak links, so goods and idea production functions have weak links. But they get automated away. It used to be inverting matrices was done with a pencil and paper by hand. Now the computer does it, right? It used to be that driving your car was done by hand. Now in San Francisco, the computer does it, right? Hopefully, hopefully, uh, throughout the world, the computer will do it someday soon. Um, okay, and then we, we-- That, that automation process occurs endogenously over time. You invent better machines, you invent faster computers, and that lets you automate more and more things, okay? And then we calibrate the model

Speaker 18:47 to fit the historical US data back to the nineteen fifties, okay? And then we run it forward to ask what's gonna happen, okay? And none of the numbers I'm gonna show you should you take that seriously, but it's just helpful to see what happens and try to think about what the economic forces are at play and how important they might be. Okay. Before I show you the simulation going forward, let me give you another thought experiment that you can do in this model that's really interesting. So as I mentioned, Dario Amodei, Anthropic, you know, OpenAI, um, everyone is expecting we're gonna automate software engineers very soon. One can ask, how much richer would we be if we automated all software engineers? We can ask that in the model, and in fact, I'm gonna change the question slightly. What if we had infinite amounts of software rather than finite amounts of software? So anything that uses software, push it to infinity. How much richer would we be, okay? Remember the computer example. We have a hundred million times the transistors, and we're not a hundred million times richer, so it's gonna be much less because of weak links. How much less? Turns out there's a very simple, elegant formula. Because of weak links, infinite amounts of some task raises GDP by that task's share of GDP. Software is about two percent of GDP. If we had infinite software, we'd be two percent richer, right? Why only two percent? Because all the other weak links are bottlenecking us, okay? And so this, this, this sort of makes you realize automating one thing really well is not enough. What you need to do is you need to keep automating the weak links, right? We need a dynamic model, not just this static model, and that's what the, that's what the model does. Okay, um, yeah. So our, our model features the two key ingredients that I think one takes away from those scenarios that I outlined. In the first scenario, you've got automation gives you new ideas, gives you more automation, gives you more new ideas. There's a flywheel effect there with positive feedback. And positive feedback wants to explode, right? On the other hand, based on the business as usual scenario, we've got weak links, right? Weak links tell you, okay, automating some stuff but not the rest, the chain is still weak because of the weak links. Okay. So what if you put both these ingredients in a model, calibrate it to what we've observed in the past, and run it forward? And I'm gonna show you two more simulations, right? Two sets of simulations. Um, in one, AI is just a continuation of the historical patterns of automation that we've seen, okay? So it just continues... You know, we, we, we've been automating the economy. This is an important point. We've been automating production for two hundred years. The Industrial Revolution, right? Textile looms, um, tractors and railroads, right? Computers. We've been automating for a long time, and yet two percent growth, right? But it-- You know, if you continue doing that, something may change. And then the second scenario recognizes, look, AI may be different. AI might not just be a continuation of historical patterns. Um, a, a weak link of our paper is we don't have a, a micro-founded model of AI that explains exactly how it's different. So instead what we're gonna do is we're gonna take a very aggressive calibration. One of the sectors we look at in our, in our history is the computing sector, right? Moore's Law is incredibly fast. So what if, as our aggressive scenario, where AI is a great-- a, a break with the past, we say the entire economy starts out today getting automated just like Moore's Law. Machines get better at ten percent per year throughout the economy starting today,

Speaker 22:52 and then that automation gives you more ideas, gives you more automation, okay? So that's gonna be a very aggressive scenario. I think of the second scenario as, as too aggressive, and the first scenario is probably not aggressive enough. So again, maybe the truth is somewhere in between. All right. So, so what do we see when we do this? So I'm gonna show you, um, three sets of graphs for each scenario and a lot of lines on the graphs, but hopefully it'll make sense. This first graph is the capital share versus the labor share. Remember I said one-third is paid to capital, two-thirds is paid to labor? What share is paid to capital in our simulations? You can see going forward, it's thirty-eight point two percent for the, the next eighty years. And then it actually splits off, you know, a hundred years from now. What's going on? We, we have three scenarios. The purple scenario says there's nothing special about humans and everything gets automated in finite time. If that happens, the share of GDP paid to capital goes to a hundred percent, share paid to labor goes to zero, okay? The green scenario says almost that, but there's three percent of tasks that are reserved for humans, that can't be automated away. Magnus Carlsen playing chess, Leo Messi playing soccer, right? There's some things we're gonna reserve for humans. And in that case, you get infinitely good at the ninety-seven percent, but the three percent done by humans bottlenecks you. And in fact, the weak link is the humans. The human share goes to a hundred percent, the labor share goes to a hundred percent, the capital share goes to zero. This is like that computer example I gave you, where the share paid to computers is falling. Okay, and then interestingly, there's a scenario in between that I call the baseline, the blue scenario, that says actually the capital shares remain stable, one-third, two-thirds, right? And that one's interesting because you say, okay, what's gonna happen to growth in the intermediate scenario where the capital share-- You know, labor still gets two-thirds all the way through. What happens to automation and, and economic growth? Okay, here's the growth rates over time. The baseline is two percent growth, like in the, the graph that I showed you. And you see a couple of different things here. First purple line, growth goes off to infinity. Full automation, if machines can do everything, including produce ideas, and machines keep getting better, growth explodes. The green scenario says, "No, no, no. We're bottlenecked by the three percent of things that can only be done by humans," so the growth rate in the end is limited to how fast humans get better. And I'm not that much better at inverting matrices by hand than, you know, um, my mom was when she was young, right? So yeah, that, that rate is pretty slow. And then interestingly, the baseline scenario, the blue scenario, if you look closely, growth is speeding up. And in fact, even in the baseline scenario, growth speeds up, and it ends up settling down at fifty percent per year, okay? But it takes centuries to get there, okay? So even in the baseline scenario, growth accelerates. You can see the acceleration, two to two point three to two point six to three, but look at the axis. The acceleration is stunningly slow, right? By, by twenty fifty, instead of two percent, we're growing at two point three percent, right? This is a model that eventually explodes, but it takes a long time. Why? Because of weak links, right? This is again like the computer example. We have a hundred million times the transistors, but yeah, we're just limited by all the things that humans still have to do, okay? I can also show you this in levels. So this is the growth rate graph. Here's the level, income per person. So remember the two percent line, that straight line is the dashed line in orange now. And then the numbers are how much richer are you relative to the orange line if growth hadn't accelerated. And you see, yeah, by twenty fifty, we're four percent richer. By twenty seventy-five, we're fifteen percent richer. And so growth is accelerating. Growth is even exploding, but the explosion is slow because of weak links, okay? So now you can see why-- Oh, and, and, and one other thing that I find interesting about these graphs, notice the three different scenarios, may- maybe look here, are very different for the future. Two hundred years from now, the world looks very different, whether in green or purple. And yet, for the next seventy-five years, it's really hard to tell which scenario you're in, okay? Now, you, you can see why I wanted a second set of scenarios, because this growth explodes, but it takes a hundred years. You might say, "Okay, well, you're being too conservative with respect to AI. AI is not just a continuation of the automation that we've seen for the last two hundred years."

Speaker 27:56 Maybe it is, but seeing this, you wanna say, "What if it's not?" And so now we say for our next set of scenarios, no, suppose AI starts out today with Moore's law everywhere. It's not just Moore's law in the computing sector. Machines get better throughout the economy at ten percent per year instead of three percent per year, which was our baseline number for the, for the aggregate economy, okay? Okay, capital share, labor share looks the same. This graph, you might say, is-- isn't that the same graph? Well, look at the axis. The years have changed, okay? And so if I show you the growth rates now, okay, now this is different, right? You can see in the purple line and even the blue line, growth is exploding, right? By twenty fifty, we're well on the way. We're above twenty-five percent growth per year. And even today, the scenario starts in twenty twenty. If you have Moore's law everywhere today, instead of growing at two percent, the economy grows at four point seven percent. So we certainly didn't do that for the last six years, and so we kinda-- that's another version of this example's too aggressive. But, but what I take away from here, let me show you... And, you know, growth goes to seven percent and thirteen percent by twenty forty. This is, this is an explosion that happens relatively quickly. And yet, relative to the AI twenty twenty-seven folks or the people in Silicon Valley who think it's happening in four years, no, it's not until twenty sixty that the explosion, you know, is fully complete, right? Even in this case, in this incredibly aggressive case, where we're fifty percent richer in twenty thirty than we would have been without the, the accelerating growth, even in this case, the explosion takes thirty years. Why? Again, weak links. And so I think this weak link phenomenon, um, is really slowing things down, and that's one of the lessons I take away, uh, from the scenario. So, so, so two lessons here. You know, I, I, I'm a person who's made my career off of that straight line graph that I showed you. The whole reason I'm at Stanford, the whole reason I got tenure, you know, is that, that one picture. [audience laughing] Right? That picture says, bet on growth not changing, right? Two percent per year for a hundred and fifty years, maybe for the next thirty years. And yet, I'm telling you, in a world of AI, all the scenarios I run say growth explodes over the next fifty or a hundred years. Right? Now, what happens in the long run depends on these details, the Leo Messi example. Um, growth explodes, but still the explosion is not nearly as fast as you would have thought when I s- said the word explosion, right? And that's the weak link thing. So, so those are the, the things that I think are going on. Notice, why does it explode even with weak links? Well, eventually we automate away all the weak links, and then this flywheel

Speaker 30:59 effect really takes over, and that's, that's what's giving you the explosion. Okay? Um, all right. So, so what I wanna do, uh, in the last 10 minutes is talk about several other things. And, and these things, um, at least the, the job and inequality thing, like, my kids ask me about this every day. Every time I give a talk, they ask me about this. I don't know nearly as much about this as I do about growth, but I can at least, having thought about it for a while and read other people's expert work, I can share some insights with you. So Jeff Hinton, Nobel Prize winner, inventor of, of, you know, deep neural networks, um, in 2016, I was actually at the conference in Toronto when, when he gave this remark. He said, "We should stop training radiologists." He said, "In five years," from 2016's perspective, "there'll be no more radiologists with jobs because the AI will be better than the radiologists." Okay? He wasn't wrong about AI being better than the radiologists. That's true on many dimensions today, but not on every dimension. What do I mean by that? Well, in fact, there's a nice article written in, uh, this magazine, Works in Progress, that documented we have more radiologists today than we did in 2016, and they're paid more than they were in 2016, right? Why is that? I think again it's weak links. So the insight economists have come up with is jobs are bundles of tasks. There are 100 different tasks that you do in your job. When the AI automates 75 of them, well, the weak links are the things that are now scarce and get the high return. And the automation, you know, the fact that I have a computer when I'm doing research makes me more productive as a researcher, so my wage goes up. The fact that the radiologist can consult with an AI model that helps them read scans and, and detect, uh, cancers and other problems makes them more valuable and more productive, and they're still needed to consult on surgeries or consult with other doctors or double-check the hardest scans, right? So automating 75% of tasks can raise ra- wages. On the other hand, if you're betting on Uber drivers 10 years from now, I think there's a good chance we won't have Uber drivers 10 years from now. Right now I'm gonna sound like Jeff Hinton, you know, in, in 10 years when we still have Uber drivers, but, you know, the Waymo, the self-driving cars are really automating everything that an Uber driver does. And, you know, it's, it takes time. It's actually fascinating how much time it takes. You know, the, the first self-driving car competition, DARPA had a competition in 2004. Uh, Carnegie Mellon, Stanford, other teams entered. No one won. No one completed the course. 2005, Stanford wins, Sebastian Thrun. And, you know, that was 2005. We're more than 20 years later, and yeah, in San Francisco you could take a self-driving car, but they're very rare outside of, uh, the Bay Area, and even in the Bay Area they're, um, not, one wouldn't call them common. Okay? And so again,

Speaker 34:02 why? Kind of the weak link view tells you things take a lot longer than you think. Uh, inequality and meaningful work. So, you know, historically labor's the main asset that people trade to get consumption. What happens when the machines can do things better than you? You might worry about the value of your labor, and will you be able to trade it for consumption? That's a valid worry. I, l- let me give you the, the optimistic take. The world where AI changes everything, as we saw in the simulations, is a world where GDP is incredibly high. We're, we're living in a world of abundance there, and there's plenty to go around. Rich countries already engage in lots of redistribution. An interesting question that I, that I wanna look at is, suppose we kept the US redistribution programs in place as they are now, and we run the model. What happens to the wages of the bottom 10%, you know, or consumption of the bottom 10%, not the wages? You know, I think it goes up, right? And so there, there's a chance to, to make everyone better off. Economists love making everyone better off, and we say things like that about trade, and yet then you get the China shock and, you know, people in North Carolina, uh, seeing their jobs disappear and communities being decimated. So it doesn't happen automatically. But at least it's a world of abundance where there are possibilities. Um, uh, uh, I will say one, one other thing here. Um, I, I worry, you know, I, I've been using the AI models to help me with research all the time now, and already GPT 5.2 Pro was as good as me at math, and 5.5 Pro is way better than me at math. And I wonder, how long is it gonna be before the AI writes better growth papers than I do? And what am I gonna do? I get all my meaning, half my meaning, family's great. The other half of my meaning- [laughing]

Speaker 35:58 [laughing]

Speaker 35:58 The other half of my meaning comes from, you know, developing growth models and the AI is gonna be better than me. What will we do? Well, a, a, an analogy I go back to is retirement, right? When we look at retirees, we don't say, "Oh, they no longer have this great meaningful work." They seem pretty happy. They live in a world of abundance, go on cruises, go see their friends, go to have dancing. Summer camp was an analogy I like. They go make pottery, and maybe my version of that would be making pottery and singing songs and getting together with my growth friends and having the AI teach us the latest growth model. I think that'll be, that'll be my version of summer camp. Um- Okay, let me say one other thing though, because contrary to what you might have taken from the talk so far, I'm not, uh, a total optimist on this. I would even say I'm very nervous about our future. Why? Because of catastrophic risks. And I- I'll say something about it and, and I think it's really important to discuss this honestly and openly and without the sort of, you know, pejorative criticism that sometimes gets associated with it. I think this is something we absolutely should take seriously. So, so there are two versions that people talk about. There's the bad actor version and the alien intelligence version. The bad actor version, I think everyone can, can understand very easily and, and believes is likely to be a problem. So what if there's some bad actor, some hacker in North Korea or wherever, that decides they want to do harm, and they have access to a jailbroken version of GPT-8 or Opus 7? All these models are jailbroken the day they come out. And they ask the model, you know, this oracle, by GPT-8, by Opus 7, these are gonna be able to do anything the smartest humans can do. If it's possible to design a virus that's more lethal than Ebola and takes three months to display symptoms, the AI will figure it out, right? Uh, we got through nuclear weapons so far because they were so rare. Only a handful of people had red buttons that they could push and do serious damage. If eight billion people have access to the red button, can we make sure no one pushes it? So this is a problem that I, I take very seriously. The other version is more speculative, but there's a quote from this computer science professor from Berkeley that I found helpful. So alien intelligence. Suppose we found out this evening that there's a spaceship on its way from Pluto toward the Earth, right? Some alien spacecraft heading toward Earth. How would we feel? We'd be pretty excited at first, and then we'd say, "But, you know, wait a minute. When advanced societies or species encounter less advanced societies and species in our history, it hasn't gone well for the less advanced." [laughing] Right? And Stuart Russell's quote, "How do we retain power over entities more powerful than us forever?" I... You know, that's worth thinking about. Let me say o- one other thing. I'm gonna skip the how much should we spend to avoid risk. Let me say one other thing, because it's related to weak links and it's, it's the thing I'm, I've been worried about just in the last month as I, as I developed the, the research paper. Chain is only as strong as its weakest links. I told you one consequence of that is the benefits come slowly, right? Because you have to automate all the weak links. You, you need to strengthen all of them to get the full benefit, to get growth to explode. Okay, but remember the space shuttle Challenger, the, the O-ring problem, or a chain. If you break one link in the chain, all the value gets lost, right? And so a weak link model is very slow to improve, but it's very fragile on the downside. So already, think about Mythos, the, the model that Anthropic, um, didn't release publicly, but said, "Look, this model's discovering bugs in twenty-five-year-old software that's been battle tested by humans for twenty-five years that we didn't find." Right? It's, it's discovered thousands of bugs that people didn't find. Um, in six months, in a year, we'll have an open source version of Mythos that anyone can use, right? How sure are we that a bad actor doesn't take it and hack the electric grid, hack the financial system, hack the banking system? Or it can communicate, it contacts a, a bio lab somewhere in, on, in, in, across the Earth and says, "Help me design a virus. I'm a, I'm a scientist at Stanford trying to do something." Or, you know, you know... And like, if you, if you hack the electric grid, you hack the banking system, we zero out everyone's bank balance in your favorite financial institution, that would be a huge problem. Not an existential problem, but I think that's a problem we have a good chance of facing in the next three years, right? The AI can already do that. Mythos can already come close to doing that. Wait a year or two or three at the rate these models are getting better, this is a problem that's around the corner, and I think it's something, uh, we should be concerned with. Um, okay. Let me, uh, let me just end with this, this last thought. So a question I, I find helpful, how much did the

Speaker 41:02 internet change the world between nineteen ninety and two thousand and twenty? A lot, we think. Despite the two percent growth, right? Again, you don't know the counterfactual. How much is AI gonna change the world between, say, twenty fifteen and twenty forty-five? How many internets is it worth? I think multiple internets, many internets. I think it's, it's more transformative than anything we've seen probably so far. But it's probably gonna take a longer time than we thought. But just because it takes thirty years instead of five years doesn't mean the effects won't ultimately be huge and transformative. And then I add this last bullet point. I think the downside risk can come sooner. That's what the weak link view of the world delivers. And so, uh, we should be using the intervening years, uh, that we have to prepare for these risks, both the inequality risks, the labor market risks, the political economy risks, and the catastrophic type risks that might be there. So sorry to end on a down note. I should say class of two thousand and six, hooray. [laughs] [cheering] [applauding] Okay. So I, I think we have fifteen minutes or so for questions. There are people with microphones, so how about wait until someone comes to you with a microphone so everyone can hear. Um, happy to talk about a- any of the things on these slides or macroeconomics or how Stanford's changed or... I guess up here.

Speaker 42:30 Um, thank you for that. Uh, a, a question around, uh, the slow growth rate. It seems counterintuitively slow. Um, and I'm, I'm sure you've heard this objection a million times, but I'm curious how you, how you respond is, um, what about, like, the infinite free chess lessons that I get that aren't monetized? Um, and there's value there, but I don't think that it's reflected in GDP, and that's just gonna multiply going forward. So even if the GDP growth rate looks slow, there's so much value. Um, how are we capturing that?

Speaker 43:08 Yeah, this is a, this is a question you can ask historically as well. GDP is mismeasured for sure, but it's always been mismeasured, right? You invent, uh, antibiotics. How are those captured? How are the gains in life expectancy over the 20th century? Massive gains. Um, Bill Nordhaus, the w- Nobel Prize winner from a couple years ago, um, asked this question. He said, "Suppose you could only have one of two things. You could have the GDP growth of the 20th century, or you could have the gains in life expectancy, but not both." Life expectancy went from 50 years to, you know, 77 years, right? And he surveyed a bunch of people, and half chose one, half chose the other. He says, "The gains in life expectancy, which are not nearly adequately captured in GDP, are as important as the gains in GDP." So, so I totally agree with you, things are mismeasured. An interesting question is whether things will be increasingly mismeasured or not, and I, I'm open to the possibility that the answer's yes. Um, so the, the simulations I'm showing you are kinda holding constant the measurement, and that would be one reason why things could be a little faster. I, I think that's totally a fair point. I appreciate it. Yeah, I think there was a question up here. Oh, over here. Sorry.

Speaker 44:18 Well, it's on my face. Uh, so one question I have, and I don't know, it could be not the topic of your talk, is about distribution and the short-term effects, short being next 10, 20 years, is that if AI comes and automates certain tasks and completely eliminates the need for those tasks, a big chunk of the economy dies almost immediately. And until we reach that 50 to 100-year, uh, trajectory of whatever growth we have, this is gonna be a shock in and by itself, which could be a disrupting factor to AI growth or any other growth. Um, do you have any scenarios simulating that?

Speaker 44:57 I, I don't. I think this is an incredibly important question. I had those two slides on the labor market effects, and I kinda mentioned, yeah, I'm not the right person for that. I, I am-- The follow-up paper we're working on is exactly on this question, so, you know, come, come back in a year or two and hopefully I'll have a better answer. I, I will say as I think about it, um, I, I think the Waymo example or the radiologist example I find very helpful because, again, Geoff Hinton, world's expert on deep learning, you know, deep neural networks, is telling you we're not gonna need radiologists, and yet we have more and they're better paid. Or, you know, Waymo when, when it was... With, with self-driving cars, 2012, if you go back and read The Wall Street Journal, they're saying, "Self-driving cars are here. In five years, no one will need to drive ever again," and we're a long, long way from that. Why? Because weak links and the lesson of history that all these changes take a lot longer than you think. And so, uh, I'll t- uh, uh, the thing that's gonna be automated first, I believe, is software engineering. Ask yourself, do you think we'll have as many software engineers 10 years from now as we do today? And it's easy to say no, but let me give you the scenario where we say yes. AI's gonna transform the entire world. Integrating AI into every business in the world is a long, drawn-out process that requires lots of software engineers. You might say someday we'll have AI that can do it, but ask yourself, you're the CEO of your company or the CIO, are you just letting Anthropic press the button, or do you want some people in there that you can talk to and do it slowly and make sure it works right with your precious data? I, I think we're gonna have software engineers around, and that's probably the first thing to be automated. So, so I think o- one, one of the things I took away from that, the scenarios being relatively slow, is that we might have more time than we think to avoid some of these scenarios that we're very concerned about. But that's just a very preliminary thought, and I, I think this is a really important question.

Speaker 47:02 I'd just like to, uh, double-click on the radiologist versus Uber driver. If you look at the number of people that are getting scanned, um, and have more active, uh, preventionary healthcare, it's exploded over the last 50 years. And so possibly a reason for the r- the rise of radiologists and also incomes for radiologists is because there's more variability, and, um, due to the fact that there are more medical technicians that increase the variability by placing patients into a scan in various different ways, and there are more things that are being seen. Um, code is very different, uh, and I would argue that if there is, you know, uh, automation in terms of placement and also machine learning in terms of imaging, that, you know, uh, y- the weak link could, uh, become much stronger. Um, so this all kind of feels like Von Neumann's elephant, uh, to a certain degree and, you know, at what point are we saying, like, you know, with five points, uh, five variables we can, like, you know, make the elephant's tails wag?

Speaker 48:36 Yeah, no, the, the, these are all excellent points. I don't disagree with anything you said. I, I, I do think that, um, uh- Yeah. Th- things taking longer than we might have expected based on Waymo. Um, h- here's another way to think about it. So someday I think we'll have robots teaching our kindergartners, right? Why do I say that? You might say, "No, I'd never want a robot teaching my kindergartner." Well, remember, the average kindergartner teacher is not nearly as good as the world's best in history kindergarten teacher. Once we invent the robot that can be trained to be the world's best kindergarten teacher in history or even better, we can then replicate them a million times and give every kindergarten classroom that robot, okay? That's gonna happen, but is it gonna happen in 10 years? No way, right? It's Waymo, you know, triple clicked, right? Are we gonna let the robot take care of little, you know, whoever, mine, Audrey, little Audrey? Uh, no, because, you know, we wanna make sure it's 100 and thousand percent safe, right? So, so I do think these things are gonna take, uh-

Speaker 49:42 Just one follow-up.

Speaker 49:42 Sure.

Speaker 49:43 Also, if you look at, uh, uh, you know, the growth of, uh, the Taiwanese economy or the South Korean economy, uh, IT output for those economies are, uh, substantial. So we've outsourced manufacturing of semiconductor chips to those economies, and we've replaced it with services, m- namely healthcare and financial services over the years. Um, those are the areas that are easily at threat from AI. So how does your model, like, kind of account for the fact that you're replacing manufacturing with services and weak links?

Speaker 50:26 Yeah. I think historically manufacturing is kind of the easy thing to automate, and kindergarten teachers and the nurse holding the hand of my father when he has Alzheimer's at night, you know, tho- those things, again, we'll get a robot someday to do it, but, you know, maybe not in the next 20 years. I think the services m- they, to me they seem like the, the weak link. I think over here. Sorry, I need to look for the people in the red shirts. That's my, my rule.

Speaker 50:48 Okay, so people have talked about the K-shaped economy. You know, we've all heard, of course, of Meta doling out $100 million to top software engineers, uh, or AI engineers. It feels very Gilded Age right now, at least in the United States. So my question is, in capitalism, what stops the hyper concentration of the capital share of income, i.e. basically an oligopoly, big tech on steroids capturing more and more of the economy?

Speaker 51:19 Yeah, no, th- this is totally a valid question to worry about. I'll, I'll, I don't have the answers, but I'll tell you some things that I've been thinking about along these lines. Um, so first, I think what we're seeing with AI-- So contrary to what I said, historically we thought manufacturing would be the first thing to be automated. Low-skilled workers could be replaced, and us creative types, how could you ever replace creativity? No, it turns out ChatGPT does creativity like this, right? I'm gonna be replaced long before the electricians and the plumbers, right? Think about what that does for inequality. That's actually good. The electrician's wage is going up like crazy, and my wage is, I think it's probably not going down actually, but, but, you know, it's, it- [laughs]

Speaker 51:59 [laughs]

Speaker 52:00 Fing- fingers crossed. [laughs] Fingers crossed. But, but it could be good for inequality in the short term. Second thing I'll say, the high-skilled cognitive labor that's getting replaced by AI, the doctors, the lawyers, the economists, the professors, um, we all own shares of the S&P 500. If you own shares in the stock market, you're getting some of this capital income. And so actually I think the people who own shares are fine. It's the people who don't own, own shares you have to worry about. But then I would say the government owns a lot of shares in GDP. Why? Through its tax system, right? So the government taxes and transfers, and that allows them to help out the less fortunate. We do that already, maybe not enough, but, but that at least puts a, a floor that hopefully can be improved. But this is a, this is a political economy question, and if you wanted to say, "Chad, you know nothing about political economy. Look at the last 20 years," that's a fair point. So I, I think it's a, it's a complicated problem for sure. Okay. Uh, over here please.

Speaker 53:06 Uh, so I really liked your framework of weak links, uh, being the source of scarce- scarcity, and therefore potential for high returns. So given that framework, what advice would you give for younger generation as well as us-

Speaker 53:20 [laughs]

Speaker 53:21 -to earn the high returns? [laughs]

Speaker 53:24 I do think that, um... I think management, the things we teach at a business school, are actually gonna be some of the things that are still valued in the world 10, 15, 20 years from now. Why? I think we don't wanna give power to the AI and let it make all the decisions unchecked by humans. Managers are the people who are gonna consult with the AIs, and then they make the final call. They're gonna be the thing that's, you know, in some sense scarce and incredibly valuable. So I actually think, you know, I'm gonna be automated in two years. You guys are safe for another 15 years, I think. After 15 years all, all bets are off. Own shares of the S&P 500, though, and I think you'll, you'll be okay.

Speaker 54:07 [laughs]

Speaker 54:07 So that's, um, anyway, that's kinda how I think about it. Yes.

Speaker 54:12 Uh, thanks again for the talk. I thought it was great. So if speed depends on your willingness to go at the weak links, uh, strong, won't it be possible for different societies to go at different speeds, and won't there be huge differences if someone is willing to move a lot faster?

Speaker 54:31 Yeah, no, th- th- this is an excellent question that I, you know, even less than the labor market inequality question I have thought not nearly enough, and I would say there are lots of economists thinking about the labor market inequality question. I just happen not to be-- That's not my focus yet. Almost no one's thinking about, what does this look like in a global economy? Right? So I think the economies that have the software engineers and that are creating the ideas behind neural networks and OpenAI and, and Anthropic, you know, and we own shares of those companies, I think one way or the other, in a world of abundance, we're fine. What about in developing countries that, uh, you know, are y- much less well off, don't have claims on the S&P 500, uh, have enormous inequality, and the less- least fortunate of those countries are in terrible, terrible shape? I don't know how that works. You know, the, the US redistributing from rich to poor, we already do some of that, maybe not enough of that. But the US doesn't redistribute that much to the poor people around the world, except through the ideas we invent. So the ideas we invent can help out, but, but I, this is-- I, I think the, the global implications of AI for the global economy, uh, fascinating, not studied nearly enough. I, I saved an article just yesterday that I'm gonna read on the plane tomorrow about this, but it's, um, it's the only one I could find. Uh, yes.

Speaker 55:53 First, it probably won't be replaced within two years. Uh, but just think about the, your, your, uh, example for radiologists. Um, a lot of my friends, me include, actually become busier with the AI, and they, they take over the easier job, and we do the harder job. It become easier or harder. So I'm just thinking fast forward, I don't know how many years, there will be a small group of people become busier and, and harder and do this job, and as you mentioned, they, they have higher pay, whatever. And there's a, a much larger group probably can just take their summer vacation all day long. How the society will think about the wealth distribution, the resource distribution at that time? Did you ever think about that?

Speaker 56:34 Yeah, yeah, yeah. Th- th- this is y- y- the, the, the only answer I have which is not satisfactory is at least it will be a world of abundance, right? In a world of abundance where, you know, the, the growth graph looks like that, you know, we have enough income for lots of billionaires and for the worst among us to be millionaires, right? That's what infinite income and finite time creates a lot of resources to redistribute. Warren Buffett and Bill Gates give away a, away a lot of resources. The tax system transfers a lot of resources. So there's every possibility for that to work out well. It doesn't mean it has to, but I, I do think, um, a world of abundance is a great world to hope for. We just need to solve this subtle, complicated redistribution question. But I, I, I'm more optimistic about that. Again, call me Pollyanna. Yep.

Speaker 57:22 And this will be our last question.

Speaker 57:26 Oh, wow. Uh, [laughs] Yes. Um, uh, my name is Roberto Santana, Class of 2011. I-

Speaker 57:33 [cheers]

Speaker 57:34 Yay. [claps] I work at Google DeepMind, and, um, I'm trying to reconcile the model you're showing with the natural experiments we're running, and they don't match up. Uh, so the, I think the, the, the challenge I'm, I'm having with the model that you provided is that, um, it assumes that the constraint, that weak link constraint, is a human, and that the AI cannot take or improve in those tasks. So the coordination, judgment, taste, I've heard it in many different ways. And so i- it appears to me that the, the model gradual, gradualism is, is based on, uh, an assumption that AI has to be gradual. And so that, that is the piece that I'm trying to, to reconcile.

Speaker 58:26 Yeah, yeah. No, th- this is totally fair. And in some sense, that's y- your reaction is entirely the point of the paper. Because I went into it, uh, you know, I, I, I am in Silicon Valley, so I hear and see and read all the things that say the world is changing tomorrow, and I'm definitely open to that. I write down the model, calibrate it to historical data. I haven't told you how I calibrated, how strong are the weak links. That's obviously a key question here. Um, I don't have time to go into it now. You could argue we've calibrated that to be too strong. Um, but again, the, the 50-year thing takes a long time. The example I like when I think about is this plausible or, you know, are my friends in Silicon Valley who draw this line, you know, happening in three years. Uh, you know, the Waymo example's really compelling to me because Waymo, people were saying, like I said, in 2012, this is gonna be a solved problem. Think about how easy it is to solve a self-driving car problem. You turn the wheel left, you turn the wheel right, you step on the gas, you step on the brake. That's all you have to decide, right? We have every sensor in the world, incredibly machine learning algorithms that can do math better than me. Simple problem, seemingly, and yet not a simple problem. Why? Because there are all sorts of bottlenecks, weak links. The physical world is complicated. The cognitive world, I'm very open to the cognitive stuff going much faster. But to get that graph going, think about computers. Computers, factor shares gone down. You gotta automate everything else in the physical world, and I think that's really kind of what is, is the intuition behind the slowdown. But is it gonna happen in 20 years or 100 years? I don't know. I'm pretty sure it's not gonna happen in five years, right? Based on this weak link argument. But, uh, yeah. Anyway, thank you so much. Hope you have a wonderful time at your reunion. [claps]