{"url": "https://www.youtube.com/watch?v=sfnScBUwwMU", "page_url": "https://www.youtube.com/watch?v=sfnScBUwwMU", "title": "Sam Altman and William G. Gale discuss Taxation Solutions for Advanced AI", "show_or_channel": "Centre for the Governance of AI (GovAI)", "date": null, "duration_s": 3152, "type": "talk", "speakers": ["Sam Altman", "William G. Gale"], "confidence": "medium", "notes": "Centre for the Governance of AI: taxation solutions for advanced AI.", "people": ["sam-altman"], "id": "0000_centre-for-the-governance-sam-altman-and-william-g-gale-discuss-t", "audio_duration_s": 3151.7373125, "worker": "jc-w08", "transcribe_s": 137.7, "model": "large-v3", "segments": [{"start": 0.72, "end": 29.92, "text": "Hello, I'm Anton Korinek. I'm the economics lead of the Center for the Governance of AI, which is organizing this event. And I'm also a Rubenstein Fellow at Brookings and a professor of economics and business administration at the University of Virginia. Welcome to our first seminar of 2022. This event is part of a series put on by the Center for the Governance of AI dedicated to understanding the long-term risks and opportunities posed by AI."}, {"start": 29.92, "end": 53.98, "text": "Future seminars in the series will feature Paul Charest on the long-term security implications of AI and Holden Karnofsky on the possibility that we are living in the most important century. It's an honor and a true pleasure to welcome our two distinguished guests for today's event, Sam Altman of OpenAI and Bill Gale of Brookings."}, {"start": 53.98, "end": 83.38, "text": "Our topic for today is how to ensure shared prosperity in an age of transformative advances in AI. This is a question of very general interest. I personally view it as the most important economic question of our century. But what made us particularly excited to invite Sam for this is a blog post that he published last year entitled Moore's Law for Everything, and to which we have also linked from the event page. In his post,"}, {"start": 83.38, "end": 110.74, "text": "Sam describes the economic and social challenges that we will face if advanced AI pushes the price of many kinds of labor towards zero. The goal of today's webinar is to have a conversation between experts on technology and on public policy represented here by Sam and by Bill because the two fields are often too far apart and we believe this is a really important conversation."}, {"start": 110.9, "end": 140.4, "text": "Sometimes, technologists and public policy experts even speak two different languages. For example, we will be talking about HEI in today's webinar, and that has two very different meanings in the two fields. In public policy, specifically in US tax policy, HEI is adjusted gross income and is used to calculate federal income taxes. In technology circles, HEI, of course, means artificial general intelligence."}, {"start": 140.4, "end": 170.83, "text": "To be honest, I have certain mixed feelings about both forms of AGI. I want to start our event today with Sam to hear about the transformative potential of AI from him. But let me first introduce Sam a little more. Sam is the CEO of OpenAI, which he co-founded in 2015 and which is one of the leading AI companies focused on the development of AGI that benefits all of humanity. Sam was also"}, {"start": 170.83, "end": 199.06, "text": "A former president of Y Combinator. Sam, OpenAI has made its name by defining the cutting edge of large language models. To give our conversation a technical grounding, can you tell us about your vision for how we will get from where we are now to something that people will recognize as more like human level AI or HEI? And are you perhaps willing to speculate on a timeline?"}, {"start": 199.63, "end": 226.61, "text": "First of all, thanks for having me. I am excited to talk about the economic and policy implications of all of this. We spend most of our time really thinking hard about technology and the sort of very long-term future. But I think that the short and medium term challenges to society are going to be immense. And it's nice to be able to have a forum with smart people to talk about that with. And we're looking for as many good ideas here as we can find. I think that..."}, {"start": 226.8, "end": 256.08, "text": "There are people who think that if you just continue to scale up large language models, you will get AGI, not the tax version. We don't think that is the most likely path to get there. But certainly, as you create models that can do more work with different modalities, have more intelligence that they have developed, that they've learned, and the ability to operate over long time horizons."}, {"start": 256.08, "end": 285.87, "text": "Accomplish complex goals, pick their own data they need to train on, do the things that a human would do, you know, read more books about a specific area of interest, try some experiments, call a smart friend, whatever. I think that that's going to be something closer to what feels like an AGI. I don't think it will be an all at once moment. I don't think we're going to have this like, you know, one day of takeoff or one week of takeoff. But I do expect it to be an accelerating process. And one of the things that we believe."}, {"start": 285.9, "end": 314.77, "text": "At OpenAI, about the way to carefully steward AGI into existence is that we need a continuous deployment and a roughly constant rate of change is better for the world. So rather than have people all of a sudden wake up and be like, whoa, I had no idea this was coming in. Now there's an AGI. We would like it to just feel like this continuous arc where society, institutions, policy, economics, people all have time to adapt. And importantly, that we can learn along the way."}, {"start": 314.77, "end": 346.74, "text": "What the problems are, how to align these systems, well in advance of having something that would be recognized as an AGI, really have learned how to align these systems. And I don't think our current techniques are going to scale without new ideas there either. But I think that's what the path looks like. There will be a lot of new research. There'll be a lot more scale. There will be a lot of complex systems integration and engineering. And there will be this ongoing feedback loop with society and this sort of"}, {"start": 346.74, "end": 376.4, "text": "Societal inputs, the infrastructure that trains these models together, all of that will be at some point recognizable as an AGI. In terms of timeline, I don't know. I do think this will be the most important century, I'll say that. Thank you, Sam. Now let's turn a little bit more towards the economic implications. What do you view as the implications of technological advances towards AGI for our economy?"}, {"start": 378.61, "end": 403.2, "text": "It's always hard to predict the details here, but at the highest level, I expect as the, let's call it the price of intelligence, how much you have to pay to get a task that requires a lot of thinking or intellectual labor done. That gates a lot of other things throughout the system. And in fact,"}, {"start": 403.2, "end": 430.27, "text": "There's some levels where just no one person or no group of people that can coordinate well is smart enough to do something. So there's a whole lot of things that don't happen. And as these AI systems get way more advanced, I expect the price of cognitive labor to trend way down. And that has a ton of implications that are positive for people. It also has enormous implications on wages for cognitive labor."}, {"start": 430.27, "end": 460.21, "text": "You titled your blog post on this topic, Moore's Law for Everything. Could you perhaps expand a little bit on what does Moore's Law for Everything mean to you? Well, as you can probably tell from the title, I think this whole idea that, let's call it the cost of intelligence can fall in half every two years or 18 months, whichever version of Moore's Law you want. I think that's a good thing. I think that..."}, {"start": 460.21, "end": 487.3, "text": "It's sort of well-known that compound growth is this incredibly powerful force in the universe that almost all of us underestimate, even those of us who think we understand how important it is. That's when the curve goes like this. It's equally powerful in other cases when the curve goes like this. And this idea that we can just have twice as much of the things that we value every two years and that there will be..."}, {"start": 487.3, "end": 515.66, "text": "Not just quantitative jumps, but qualitative ones, things that just weren't possible before. I think that's great. You know, if we look at the last several decades in the US, think about what wonderful things have been accomplished by the original Moore's Law. And think how happy we are to have that. I mean, I was just thinking today, like what the pandemic would have been like if we all didn't have like these massively powerful computers and phones that, you know, so much of the world came to really depend on. And that's just, that's one little example."}, {"start": 515.66, "end": 544.53, "text": "And then contrast that with industries that have had, let's say the opposite of that, just like runaway cost disease and how we feel about those. So I think we can, we must embrace this idea that AI can deliver a technological miracle, a revolution on the order of the biggest technological revolutions we've ever had. And that is how society gets much better. I think the challenges society has faced."}, {"start": 544.53, "end": 569.97, "text": "At this moment feel quite huge. I think people are understandably quite unhappy about a lot of things. But I think lack of growth underscores a lot of those. And if we can restore that, one of the best ways to restore that is AI cutting the cost of cognitive labor. And we can all just have much more and a great rate of progress. I think a lot of the challenges in the world can get much better."}, {"start": 570.54, "end": 599.54, "text": "This is really fascinating. So cutting the cost of everything by 50% every two years or doubling the size of the economy every two years, no matter which we put it, is really a radical change from the growth rates that we face today. Society doesn't really do a good job with radical ideas anymore. We don't think about them. We no longer seem to believe they're very possible, but sometimes technology is just too powerful and we get it anyway."}, {"start": 599.86, "end": 630.0, "text": "Now, some people assert that once we have systems that do reach something like the levels of AGI, there will be no jobs left whatsoever because AI systems will be able to do everything better. They will be better academics, better policy experts, even better CEOs. What do you think about this view? Do you think there will be any jobs left and what kinds of jobs would they be? I think there will be new kinds of jobs. I mean, I think there will be a big class of things where people want"}, {"start": 630.0, "end": 660.3, "text": "The connection, the humanity of another human. And I think that will happen. But I think we're even seeing some of the things that people do when they have way more money than they need to spend. And they still want to buy status. And I sort of believe that the human desire for status is sort of unlimited. And I think NFTs are a fascinating case study. And we can see much more of things in that direction. But, you know."}, {"start": 660.3, "end": 690.51, "text": "It's hard to sit here and predict the jobs on the other side of the revolution, but I think we can say some things about human nature that could help us guess what they might be. It's always been a bad, it's always been sort of like wrong to say that after this new technological revolution, there will be no jobs. The jobs have looked very different on the other side. In this case, I expect the jobs to look the most different of any of the technological revolutions we've seen so far, because this idea of cognitive capabilities is..."}, {"start": 690.51, "end": 719.81, "text": "Such a big part of what makes us human. Sort of the most remarkable of all of our capabilities. And if that gets done by technology, then it is different. But, you know, I think we'll find new jobs. They'll feel really important to the people of the future. And they'll seem like quite silly and frivolous to us in some cases. But there's a big universe out there. And we or our descendants are going to hopefully go off and explore that. And there's going to be a lot of new things in that process. That's really interesting."}, {"start": 719.81, "end": 747.38, "text": "But let me move to the realm of public policy now. One of the fundamental principles of economics is that technology determines how much we can produce, but our institutions determine how it is distributed. You wrote that a stable economic system requires growth and inclusivity. I guess the growth will emerge naturally if your technological predictions materialize."}, {"start": 747.38, "end": 779.17, "text": "But what policies do you advocate to make the growth inclusive? Make everybody an owner. I am not a believer in sort of paternalistic institutions deciding what people need. I think it ends up being wasteful and bureaucratic and they're probably mostly wrong about how to allocate it. I'm also not a believer that you can have a long-term successful capitalist society in which most people don't own part of the upside."}, {"start": 779.17, "end": 805.34, "text": "Now, look, I am not an economist, even less of a public policy expert. So I think the part you should take seriously about the Morse law essay is the technological predictions. I think those are probably like better than average. And the specific economic and policy predictions are like probably quite bad. I meant it as like a starting point of a conversation and sort of like my best guess at predicting where things will go."}, {"start": 805.34, "end": 835.17, "text": "Well out of my depth. And I mostly wanted to contribute on the technological piece and some ideas about where I think the technology can shape or will shape the policy and economic landscape. But the principle that I feel confident about is we need a society where everyone feels like an owner and the forces of technology are naturally going to push against that. I think that's already happened. I think in the US, I don't remember the last number that I heard, but it's like something like half of the country."}, {"start": 835.17, "end": 864.19, "text": "Owns no equities. And that I think is really bad or land or whatever. And so a version of the policy that I would like is rather than having increasingly sclerotic institutions that I think have a harder time given the rate of change and complexity in society keeping up sort of to say, okay, now we're gonna have this new program and now we're gonna have that one, now we're gonna have that one. Literally just find a way to say,"}, {"start": 864.19, "end": 889.5, "text": "Here's how we're going to redistribute some amount of ownership in the things that matter. And so everyone can participate in the updraft in society. That's very thought provoking to redistribute ownership as opposed to just redistributing the output itself. Now, before we turn over the discussion to Bill, let me ask you one more question."}, {"start": 889.5, "end": 918.86, "text": "What do you think about the political feasibility of proposals like redistributing ownership? Or maybe let me make it more concrete. What could we do now to make a solution like what you are describing politically feasible? I feel so deeply out of my depth there that I hesitate to even hazard a guess. But it seems to me like the Overton window is expanding and could expand a lot."}, {"start": 918.86, "end": 951.42, "text": "I think we're ready for a real change and things are not working that well for a lot of people. I certainly don't remember people being this unhappy in the US in my lifetime, but maybe it's just getting old and bitter or something. That's a very honest thing to say. I'm afraid none of us is a real expert on all these changes because as you say, they are so radical that they are..."}, {"start": 951.42, "end": 979.23, "text": "Kind of hard to conceive of and kind of hard to imagine what they will lead to. So thank you for this fascinating initial conversation, Sam. And let me now turn it over to Bill. Bill is the Miller Chair in Federal Economic Policy and a senior fellow at Brookings. He's an expert on tax policy and fiscal policy and a co-director of the Tax Policy Center of Brookings and the Urban Institute."}, {"start": 979.23, "end": 1005.02, "text": "Phil has also been my colleague at Brookings for the past half year, and I've had the pleasure to discuss some of these themes with him. Phil, Sam has predicted that Moore's law will hold for everything if HEI is developed. And just to be clear, I mean artificial general intelligence now. Now, economists have long emphasized that there is this second force that runs counter to Moore's law."}, {"start": 1005.02, "end": 1031.54, "text": "And that has slowed down overall productivity increases, even though we have had all these fabulous technological advances in so many areas since the onset of the industrial revolution. And the second force is Baumol's cost disease. Can you explain a little bit more about this second force and what would it take to neutralize it so that Moore's law can truly apply to everything?"}, {"start": 1031.95, "end": 1058.66, "text": "All right, well, thank you. It's a pleasure to be here and to read the very stimulating proposal that Sam put forward. I am not an AI expert, but I've read a few pieces about it in the last week, and I think there's huge potential for tax issues here. So I'm very excited about being part of this discussion. Let me answer the question in three parts. The first part is generally economists think technology is..."}, {"start": 1058.66, "end": 1085.38, "text": "Technological change is a good thing. It makes labor more productive. There are adjustment costs. And if we did things right, we would compensate people. We normally don't. But in the long run, you have to believe that technological change has been not just a good, but a fantastic thing. We've had 250 years of technological change. At the same time, the economy has steadily increased. People have steadily been employed."}, {"start": 1085.38, "end": 1115.06, "text": "So then AI comes along, and the question is, what's different about AI relative to other technology? And the answer, I think, is the proposed or expected speed of the adjustment, the scale and the scope. For example, if some technology, driverless cars, took place instantly, and all people that were involved lost their jobs, all the Uber drivers and Lyft drivers, et cetera, lost their job overnight."}, {"start": 1115.06, "end": 1142.96, "text": "That would be very disruptive in an economic sense. Whereas if that happens gradually over the course of many years, then those people cycle out of those jobs. They look for new jobs. They're employed in new sectors. So the speed with which AI is being discussed or the speed of the effects that AI might have is actually a concern. From a societal perspective, I think it's possible that technological change could go."}, {"start": 1142.96, "end": 1170.91, "text": "Although generally we want to increase the rate of technological change. We'd be careful what we wish for. Baumol's cost disease applies to things like, just to give an extreme example, the technology of giving a haircut probably hasn't changed in the last 300 years. It still takes the same amount of time to use the same thing. It's hard to see the productivity of that doubling every couple of years or every five years."}, {"start": 1170.91, "end": 1199.76, "text": "Of course, that's a silly little example. But industries like healthcare and education, where there's a lot of labor input and a lot of human contact, you might think that Moore's Law wouldn't come into effect as fast as it would in the computer industry, for example. Now, I'm not reading Sam's paper literally that productivity of everything is going to double in two years. Productivity of..."}, {"start": 1199.76, "end": 1228.62, "text": "50% of the economy doubled every four years. I think it would claim a victory. That would be a massive change. So there are these forces pulling back on the speed of technological change. And to me, as an economist, the question is just how different is AI from all these other technological changes we've had since the Industrial Revolution. Thank you, Bill. Now let's turn to the public policy responses."}, {"start": 1228.62, "end": 1257.33, "text": "If AGI is developed, so that means if the price of intelligence, as Sam is saying, and that means the price of many, maybe most forms of labor, really converges towards zero, what is our arsenal of policy responses to this type of AGI? And let's look at that perhaps both in the realm of fiscal spending and in the realm of taxation. Great. Let me say, if labor income goes to zero,"}, {"start": 1257.33, "end": 1287.33, "text": "I don't know. That's just a totally different world, and we will need to rethink a lot of stuff. But if we're simply moving in that direction faster, there's a couple of things to be said on the tax and the spending side. I want to highlight something that Sam wrote about the income tax, which that it would be a bad instrument to load up in this case. And the point he's made in his paper is that the income tax has been moving more toward labor."}, {"start": 1287.33, "end": 1316.94, "text": "away from capital. And on top of that, you've got the payroll tax. So loading up the income tax now basically means more tax on labor. And it's a subtle point. It's not even well understood in the tax world. But I think it's exactly right that the income tax is a wrong instrument to load new revenues on, at least on the labor portion of the income tax. On the spending side, there's a variety of things that people suggest, some to accelerate the change, like giving people a new education."}, {"start": 1317.33, "end": 1345.41, "text": "And some to cushion the change, like a universal basic income, as Sam wrote in his paper. I actually am much more sympathetic to UBI than most economists are, and not as a replacement for existing subsidies, but as a supplement. And that's even without worrying about the downside of the AI revolution. The other things you could do,"}, {"start": 1345.41, "end": 1374.06, "text": "It sounds good in theory, but hard to implement is to have a job guarantee. With a job guarantee, a federal job guarantee, the whole thing depends on what wage you're going to guarantee people. If you guarantee a job at $7.25 an hour, that's a very, very different proposal than guaranteeing a job at $15 per hour. And lastly, the classic economic solution to this is wage subsidies. Think of the earned income tax."}, {"start": 1374.06, "end": 1402.72, "text": "Credit and scale it up massively. And so that, for example, somebody making $10 an hour would get a supplement from the government that's two times that, right? And that would be a way, it's essentially, this is a little loose, but it's essentially a UBI for people that are working, right? And so our welfare system tends to focus on benefits for the working poor and does not provide good benefits for the non-working poor. So some combination of UBI and wage subsidy."}, {"start": 1402.72, "end": 1430.51, "text": "Could cushion a lot of people and give incentives to work a lot. Let me focus a little bit more on the tax side now. Can you tell us a bit more about what the menu of tax instruments is that we would be left with if we don't want to load taxes on labor and how they would compare? Yeah, sure. The obvious candidate is the wealth tax. What Sam has proposed is a variant of a wealth tax."}, {"start": 1430.51, "end": 1458.77, "text": "I mean, literally, that's where the money is. And wealth taxes, as we've talked about a little bit, sort of have administrability issues. And sometimes a good substitute for a wealth tax is a consumption tax. And they come in different forms. But presumably, if people are generating wealth, it's because they want to consume the money. If they just want to save and create a dynasty, a consumption tax doesn't get at that."}, {"start": 1458.77, "end": 1488.85, "text": "But you can design these things, especially in combination with the universal basic income that on net hit high income households very hard and actually subsidize low income households. A paper that I wrote a couple of years ago showed that with a value added tax and a UBI, you can get results that are more progressive than the income tax itself. Now, let's turn to a few more specific proposals. So, for example,"}, {"start": 1488.85, "end": 1515.7, "text": "Bill Gates has advocated a robot tax. Sam has proposed the American Equity Fund and a substantial land tax in his blog post, Moore's Law for Everything. What is your assessment of these proposals and especially proposals that are in the realm of capital taxation? And what would you propose if AGI is developed and you are tasked with reforming our tax and spending policies?"}, {"start": 1516.14, "end": 1544.42, "text": "Yeah, both the Gates proposal and the Altman proposal are motivated by good thoughts. The difference is, I think the Gates proposal is a really bad idea. I don't understand what a robot is. If it's any labor-saving technology, right, then your washing machine is a robot, your PC is a robot, your operating system, Microsoft Windows is a robot. So I don't think we want to tax..."}, {"start": 1544.42, "end": 1568.91, "text": "Labor saving technology in particular. And so I think not only because of definitional reasons, but because we subsidize investment. Most of our policies go towards subsidized investment. Turning around and then taxing it is kind of counterproductive. It would create complicated incentives. So I don't think it's a good idea. I love the spirit behind Sam's wealth tax proposal. I love the idea."}, {"start": 1568.91, "end": 1597.47, "text": "of making everybody an owner. The issue with wealth taxes is the ability to administer them. So for example, if you taxed public corporations and private businesses and land, you'd get people moving their money out into bank accounts and gold and art and yachts and stuff like that. So then this is basically the whole wealth tax debate in a nutshell, where people say you should tax all wealth and people say, well, you can't."}, {"start": 1597.47, "end": 1627.1, "text": "Tax all wealth because how are you going to come up with the value of things that are not exchanged on markets? How are you going to do that every year for every person? So then the answer, the somewhat throw your hands up in the air answer is, well, we're just not going to tax it. So that's a wealth tax where you just erode some of the base. Sam is coming out from the other way saying there are certain components of wealth that we can tax."}, {"start": 1627.1, "end": 1656.1, "text": "Market value, essentially, and land, which are two good targets. So my nerdy, wonky tax concerns are kind of in the weeds about the administrability of the tax and the amount of avoidance and tax shifting it would cause. But I like very much the general idea of saying here are these changes. They're going to displace some people. They're going to greatly benefit other people. Let's use, as you said, Anton, the institutions that we have."}, {"start": 1656.1, "end": 1685.25, "text": "To offset some of that changes and share the wealth essentially so that everyone can be better off from AI rather than AI causing the immiseration of a substantial share of the population. I see Sam has just raised his hand on that. That reminded me, a point that I cut from the essay that I had meant to make more clearly is I think a wealth tax on people actually would be great, but it's too difficult to administer."}, {"start": 1685.25, "end": 1715.23, "text": "And a very nice thing about a wealth tax on companies instead of people is the share price is the share price. Everyone's aligned. Everyone wants to go up. It's easy to take some. And I think that would be super powerful and great. So yeah, I think it didn't come through. But part of my hope with the proposal is that I think these are the two classes of things that are going to matter more than anything else. Sure, people can buy yachts and art, but it's not actually clear to me why stashing away a billion dollars and not spending it"}, {"start": 1715.23, "end": 1742.19, "text": "That's not compounding. It's not creating this runaway effect. So the goal, really the design goal was administrability and focusing on where the big wealth generation will go in the future, which I think is these two areas and trade off the kind of like perfect fairness of taxing art and boats too, given that it would be less important."}, {"start": 1743.28, "end": 1772.34, "text": "So it looks like we have alignment on this notion that some sort of wealth taxes are very desirable and that there are difficulties for certain classes of wealth. Now, Bill, would you like to add anything more on this? No, let's have a discussion. I think there are a lot of interesting issues raised and I'd be happy to respond or clarify anything that I said earlier. Okay, great. So let's continue with the Q&A part then."}, {"start": 1772.34, "end": 1806.83, "text": "Please raise your virtual hand and get ready to unmute yourself if you have questions that you would like to pose to Sam or Bill. Or feel free to add your question into the regular chat if you prefer that I read it out. So I see we have a question from Robert. Hi, thanks. So yeah, my question was on the administration side as an outsider. So this is for both of you. Yeah, in the piece, Sam writes,"}, {"start": 1806.83, "end": 1836.69, "text": "There will obviously be an incentive for companies to escape the American equity fund tax by offshoring themselves. But a simple test involving a percentage of revenue derived from America could address this concern. And that bit, just as someone who doesn't know about tax stuff, that bit confused me. It wasn't clear to me how that would be simple or how that would work. So I was just looking for more detail, I guess, from either of you on how that could work. So for example, what percentage of Alphabet's revenue is derived from America? How do you calculate that?"}, {"start": 1837.33, "end": 1867.92, "text": "Thanks. I think they have to report that. So I think these companies say how much of their revenue is from different geographies. The hope would be that eventually the whole world realizes this is a good idea and we sort of agree on a global number and a free tax at the same rate and there's no reason to move around. But that's probably a pipe dream. It won't happen. There will be at least one jurisdiction who says, come here and no tax. But I think that companies report."}, {"start": 1868.34, "end": 1898.34, "text": "Yeah, I wrote a 280-page book on tax and fiscal policy and presented it to a bunch of tax economists. And every question I got was on the administrability of the estate tax reforms I proposed. They drilled down way into the weeds, into the details. I don't really want to do that here. The big picture here is the idea that the tax is essentially on market value. It could be paid in shares. It could be paid in taxes."}, {"start": 1898.34, "end": 1925.92, "text": "But the international aspects of it, I want to emphasize, are solvable in the sense that the US and the current corporate tax, we tax foreign corporations that do business in US on their US income. So there would be a way to do that. Someone just put in a comment about formulary apportionment, which, again, has the feature."}, {"start": 1925.92, "end": 1954.29, "text": "Like this proposal that it doesn't let the perfect be the enemy of the good. Right. And I feel like a lot of times in tax policy, people shoot for the perfect policy. They never get it. And as a result, they end up with not even a good policy. Cool. Thanks. Thank you. Let's continue with Daniel and let's maybe collect several questions. So Daniel, followed by Marcus and Jenny."}, {"start": 1954.35, "end": 1987.46, "text": "Terrific. Thank you, Anton. Real pleasure to be with everyone this evening. Spoken a lot so far about the distribution problem, about how to best share out the prosperity created by new technologies. I'd be really interested in your thoughts on a different problem, which is what I call the contribution problem. It seems to me that today, social solidarity comes from a feeling that everybody is pulling their economic weight through the work that they do and the taxes that they pay."}, {"start": 1987.46, "end": 2013.74, "text": "And if people aren't in work, there's a sort of expectation that they ought to go out there and actively look for work if they're willing and able to do so. And one of my worries about things like a universal basic income or something like universal basic assets where everybody might have a stake in the sort of productive assets in the economy is that it undermines that sense of social solidarity, that it means that some people might not be paying into the collective pot through the work that they do."}, {"start": 2016.75, "end": 2045.68, "text": "I'm interested to know your reflections on that contribution problem. It seems to me that economists spend a lot of time thinking about distributive justice, about what a fair way to share our income in society is, but we don't spend enough time thinking about contributive justice, about how we provide everyone with an opportunity to contribute in society and to be seen to be contributing. And things like universal basic income and universal basic assets, those sorts of things don't really engage with the contribution problem. That would be my challenge. But yeah, very interested in your thoughts."}, {"start": 2045.9, "end": 2076.46, "text": "So I strongly agree with the problem framing. And I think that universal basic income is only half the solution. And universal basic meaning, or fulfillment, or whatever you want to call it, is participation equally important, or almost as important, at least. I think there are going to be a lot of ways to solve this. My hope is that the tools of AI are going to empower people so much that people will, like, there will be incredible new art and immersive experiences and games and sort of"}, {"start": 2076.46, "end": 2105.3, "text": "Useful things that get created for others and people will love that. It just might not look like work in the way that we think of it right now. I also think that to get the AGI we want, and how we make governance decisions about that, that is going to be something that we want mass participation in. Tons of people helping to train the AI, helping to think about expressing the values we want the AI to."}, {"start": 2105.49, "end": 2136.21, "text": "To incorporate. So my hope is that this idea that we are racing towards this thing that is going to, and this set of decisions that we have to make in its creation are going to be, you know, one of the most exciting projects that humanity ever gets to do and sort of significantly impacts the course of the universe from here. And it is this like, you know, most important century in that sense, that that is the mission for humanity that can unite us, that a lot of people can, that all of us can to some degree participate in."}, {"start": 2136.21, "end": 2166.72, "text": "And that people really want to figure out how to do. So I am hopeful that that is going to be a kind of grand challenge for humanity and a great uniting force. But I think there will be a lot of more basic things where people just create and make value for their communities. Right. So the questions that are raised about the value of work and the conditionality of work have long dominated American politics. We have a long history of"}, {"start": 2166.72, "end": 2194.7, "text": "Supporting the working poor as much as other countries helping the non-working poor less. Having said that, there's this great concern that UBIs would basically cause people to drop out and just relax all day and listen to music. The studies don't suggest that that would happen. There's maybe a small reduction in labor supply."}, {"start": 2194.7, "end": 2222.38, "text": "Not in any sense dominant. And my interpretation of this study is that the humane aspects of a UBI, making sure people have enough to eat, making sure they have shelter and so on, totally dominate what appear to be relatively modest disincentive effects on the labor supply of the recipients. But if you were really concerned, you could combine a UBI with a job guarantee."}, {"start": 2222.38, "end": 2252.05, "text": "That actually, that's a little Orwellian because if there's a job requirement, it's not universal. But you could do a job guarantee or you can do a wage subsidy, which again, is like a UBI for the working, for people that work. I think ultimately, this is not a black-white thing. There'd be some balance between those two. I feel like we do a bad job at the very bottom for the disadvantage, extremely disadvantaged. And that's why I favor it."}, {"start": 2252.05, "end": 2281.14, "text": "But coupling it with a wage subsidy or something like that, which would help pull people into the labor force is an option. Thank you, Marcus. Yeah, I had a question for Sam. You said something that didn't quite fit with what I expected you to believe. So the thing you said was something like, even in the world where we have AGI, people will still have jobs. And I'm curious about what you mean there. Do you mean that we'll have jobs in the sense that we'll still have things that we do?"}, {"start": 2281.14, "end": 2308.58, "text": "Even though we won't contribute economic value because there'll be an AI system that could just do whatever the human did. Or is it the case that humans will be able to provide things that AI systems can't provide, for example, because they're just humans and people want to engage with human service providers or whatever it might be? Yeah. Well, I meant more the former. That was what I sort of meant by I don't think they'd be recognizable as jobs of today and they'd seem frivolous. But I think also many things that we do today wouldn't have seemed like important jobs to people from hundreds of years ago."}, {"start": 2308.58, "end": 2336.08, "text": "But I also think there will be still a huge value on human stuff. Why do we prefer handmade stuff that's worse than machine-made stuff today? There's something we like about that. Why do we value a classic car or a piece of art when there could be a perfect replica made? It's because it's real and someone made it. So I think there'll be some of that as well. In the very long-term future, when we have created self-aware"}, {"start": 2336.08, "end": 2361.68, "text": "Super intelligence that is off exploring the entire universe at some significant fraction of the speed of light. I think at that point, all bets are off. And it's really hard to make super confident predictions about what jobs look like at that point or about what we broadly construed we do then. But in worlds of even extremely powerful AI, but the world looks somewhat like it looks today, I think we're really going to find a lot of things to do."}, {"start": 2361.68, "end": 2387.38, "text": "That people value, but they won't be sort of like what we would maybe consider economically necessary from our limited vantage point of today. Thank you, Sam. Jenny? Yeah, so I sort of had a question on the global implications of AGI and redistribution, because I think the assumption that you guys are making or the focus that you guys have is on the US or on developed countries. But I was just wondering,"}, {"start": 2387.38, "end": 2405.79, "text": "If, for example, the US or another developed country develops AGI, what would be the economic consequences for other countries that are less technologically capable? And would there be redistribution from the US, from Western Europe, to Africa, Latin America? And also, would it be"}, {"start": 2405.79, "end": 2437.66, "text": "Is it impossible for them to catch up in that scenario? Because a lot of the developing countries today caught up through cheap labor, but with the Moore's law for intelligence, it seems that developed countries no longer need that anymore. So yeah, that's kind of like a broad question. Look, I think the eventual solution has to be global. There's full stop. There's no way around that. I think we can maybe prototype things in the US. I also feel a little strange."}, {"start": 2437.66, "end": 2470.29, "text": "Here's what all the other countries should do too. But yeah, I think we'll need a very global solution. Bill, would you like to add anything on that? Just that in the interim, as we're approaching a global solution, if the US or Europe is ahead of the game, that's going to result in resource transfers to the US and Europe, not away from the advanced countries. And this idea of accelerating"}, {"start": 2470.29, "end": 2501.86, "text": "Technological change means that kind of once you get ahead, once you get farther down the curve than everyone else, you're going to continue down that curve at increasingly increasing the gap over time. So Anton's question about that, I think, is particularly relevant in an international context. Thank you. Katia. Thank you. I'm Katia. I run the Insured Prosperity Initiative at the partnership on AI. Thank you all."}, {"start": 2501.86, "end": 2528.16, "text": "My question is kind of building on Jenny's. In today's world in which a global solution for redistributing ownership doesn't exist and might not exist for a long while, what can AI companies meaningfully do or should be expected to do to soften the economic world from AI, the potential one, and tackle"}, {"start": 2528.16, "end": 2561.1, "text": "The medium term challenge that Sam acknowledged from the very top. This is low confidence, but I think my first instinct is that countries that have low wages today actually benefit from this the most, benefit from AI. The problem, the thing we all like about this vision of AI is that things won't get much cheaper and the things that we don't like, the thing that we don't like about it as much as wages fall."}, {"start": 2561.1, "end": 2604.59, "text": "And I think AI should be naturally beneficial to the poor parts of the world. But I haven't thought about that in depth, and I could be totally wrong. I'm curious what Bill thinks. Will you pass me? OK. We have another question by Phil Trammell. And let me invite anybody else who has questions to raise their hand. I think you're muted. Phil? So thanks. All this discussion about AI and AGI"}, {"start": 2604.62, "end": 2634.8, "text": "Has sort of treated the technology in the abstract as if it's sort of the same wherever it comes from. But, you know, looking back, it seems like, you know, if someone besides Marconi had invented the radio a month sooner, nothing would have been different. But like if the Nazis had invented atomic weapons sooner, history would have been very different. And I'm just wondering, you know, which category you think AI falls into. And if it's something more like the atomic bomb category, why is that sort of less..."}, {"start": 2634.8, "end": 2664.08, "text": "Less featured, at least in economic discussions about the implications of AI. Most of the conversations I have throughout most of the day are about that. I certainly think it matters hugely how it's developed and who develops it first. I would say I think about that like an order of magnitude more than the economic questions. But I think the economic questions are also super important. And given that I believe it's going to be this continuous sort of..."}, {"start": 2664.08, "end": 2695.14, "text": "Deployment in the world, I think we also need to figure out the economic and policy issues that we think lead to a good solution. So it's not for lack of a belief in the extreme importance of who develops it and how. That's really why OpenAI exists. It's just this conversation is focused on the economic and policy lens. But even for the economic implications, it seems like it could matter who develops it first. But it seems like there's this segmentation where"}, {"start": 2695.14, "end": 2727.2, "text": "In strategic conversations, people care about who develops it first. When we're thinking about implications for the wage distribution, it's more thought of as an abstract technology. What I would say is in the short-term issues, it matters less who develops it first. In the long-term issues of the AGI, as we all think it might happen, then it matters much more. I think it's a time frame issue is maybe how I'd frame it."}, {"start": 2727.2, "end": 2754.26, "text": "So we have one question that Zoe Kramer put in the chat. And I think this is also mostly directed at Sam. How does your proposal, Sam, compare and contrast against Landemore or Tang's radical proposal for distributed decision mechanisms? So she writes, both ownership reform and real democracy result in the same thing."}, {"start": 2754.26, "end": 2783.39, "text": "People who carry the risks of intervention actually have real control over those interventions. Is there a reason to prefer moving the world towards land taxation and distributed ownership over distributed decision-making and citizens controlling the public policies? I think we need to do both of them. I think that they're both really important. I think one of the most important questions that will come as we develop Advanced AI is"}, {"start": 2783.39, "end": 2813.18, "text": "How do we get the world's preferences in, you know, like let's say we can solve the technical alignment problem. How do we get, how do we decide whose values to align the AGI to? How do we get the world's preferences encoded in that and taken into account? Like, of course, I don't think you should like literally make every decision about how AGI gets used, decided by a in the heat of the moment, passions running hot vote of everybody on earth for each decision."}, {"start": 2813.18, "end": 2845.97, "text": "But when it comes to like, let's say framing the constitution of AI and distributing wide power within that to people, but maybe sort of saying like, you know, here's the guardrails we're going to follow. Of course, I think we should have something like that. Phil, any thoughts on this question of how much of the distributional challenge we want to solve through private ownership or public distribution and public decisions? Well, the problem is that the private system won't"}, {"start": 2845.97, "end": 2875.66, "text": "It will solve the distribution problem. It will create surplus, but it will distribute it according to the market system. And the public system, I think Sam mentioned earlier, and I'd share this, I'm just incredibly concerned about the ability of the public system and the kind of the social values, political values underlying"}, {"start": 2875.66, "end": 2901.36, "text": "I'm really concerned about the ability to do anything that's both substantial and right. And we need, I think Sam's piece makes his point right at the beginning, if these are big changes, we need big, we need policy to go big and to actually get it right. Thank you, Sam and Bill. Ben."}, {"start": 2902.0, "end": 2929.34, "text": "Yeah, thank you both. I have a question on the importance of the idea of making people owners as a way of redistributing wealth. So I'm curious, I guess for both of you, what you see is this difference between a system where wealth or capital is taxed and then redistributed in the form of income but not capital ownership versus a system where shares of companies are actually redistributed to people in the country as opposed to just"}, {"start": 2929.34, "end": 2960.91, "text": "What's the importance of actually distributing ownership versus just merely income and wealth? I think in the short term, it doesn't matter much. But my impression is that people would be more likely to save shares than they would be to save cash. And so there's a potential over a longer term, 10, 20 years to build assets"}, {"start": 2960.91, "end": 2987.07, "text": "In a substantial share of the population, as Sam mentioned, a very large percentage of population does not own equity, does not have net worth above zero. So proposals like sharing corporate shares or in a social policy context, baby bonds have been talked about. I think those things could have effects over long periods of time."}, {"start": 2987.07, "end": 3017.55, "text": "As you kind of develop new generations that get used to saving this. But in the short run, I'm not sure to make that much difference. If we give someone $3,000 or one share of Amazon and then Amazon goes up 25% the next year. In the first case, people are like, yeah, fuck Jeff Bezos. Why does he need to get richer? But if they have the share of Amazon, that's awesome. I just benefited from that. And I think that's a significant difference. And I think most people..."}, {"start": 3017.55, "end": 3044.69, "text": "We have not effectively taught, or it's difficult to teach, or something, just how powerful compound growth is. And if people can feel that, I think it'll lead to a much more future-oriented society. And I think that's better. I also believe that ownership mentality is just really different. And then the success of society is shared, not resented. People really want it. People really work for it to happen."}, {"start": 3045.36, "end": 3075.81, "text": "I think one of the magic things about Silicon Valley was the idea that we compensate people in stock and you get a very different amount of long-term buy-in from the team than you do with high cash salaries. Thank you, Sam. Now, in our last minute, I see we have one more question from Andrew Trask. You mentioned the notion that prices fall based on abundance created by the cost of intelligence falling, recommending that we tax corporations."}, {"start": 3075.81, "end": 3104.43, "text": "In the extreme case of automation, what is the purpose of the cooperation if it is no longer employing meaningful numbers of humans for primary survival tasks? That is to say, assuming healthy competitive dynamics, would that cooperation be charging for its products at all? Would they not all become free? I have a long and complicated answer to this question, but there's no way I can do it in one minute. Let's say that like..."}, {"start": 3105.52, "end": 3134.13, "text": "People are still willing to pay for status and exclusivity and a beautifully engineered thing that solves a real problem for them over and above the cost of materials, good materials and laboring to a product today. The corporation can be clearly charged for that. Well, thank you. That was a great way to squeeze it into one minute. So let me..."}, {"start": 3134.13, "end": 3151.74, "text": "Thank Sam and Bill for joining us in today's webinar for sharing your really thought-provoking ideas on the question. And let me also thank our audience for joining. And we hope to see you at a future Gov.ai webinar soon."}]}