--:-- Dwarkesh Patel: Okay, today I have the pleasure of interviewing William MacAskill. Will is one of the founders of the Effective Altruism movement, and most recently, the author of the upcoming book, *What We Owe The Future*. Will, thanks for coming on the podcast.
--:-- Will MacAskill: Thanks so much for having me on.
--:-- Dwarkesh Patel: My first question is: what is the high-level explanation for the success of the Effective Altruism movement? Is it itself an example of the contingencies you talk about in the book?
--:-- Will MacAskill: Yeah, I think it is contingent. Maybe not on the order of, "this would never have happened," but at least on the order of decades. Evidence that Effective Altruism is somewhat contingent is that similar ideas have been promoted many times during history and not taken on.
--:-- We can go back to ancient China. The Mohists defended an impartial view of morality and took very strategic actions to help all people, in particular providing defensive assistance to cities under siege. Then, of course, there were the early utilitarians. Effective Altruism is broader than utilitarianism, but has some similarities. And then even Peter Singer in the 70s, he had been promoting the idea that we should be giving most of our income to help the very poor and hadn't had a lot of traction until even like early 2010, after GiveWell launched, after Giving What We Can launched.
--:-- What explains the rise of it? I think it was a good idea waiting to happen at some point. I think the internet helped to gather together a lot of like-minded people that wasn't possible otherwise. And I think there were some particularly lucky events, like Alex meeting Holden and me meeting Toby, that helped catalyze it at the particular time it did.
--:-- Dwarkesh Patel: If it's true, as you say in the book, that moral values are very contingent, then shouldn't that make us suspect that modern Western values probably aren't that good? They're probably mediocre or worse, because ex ante you would expect to end up with the median of all the values we could have had at this point. And obviously we'd be biased in favor of whatever values we were brought up in.
--:-- Will MacAskill: Absolutely. I think taking history seriously and appreciating the contingency of values, appreciating that if the Nazis had won the World War, we would all be thinking, "Wow, I'm so glad that moral progress happened the way it did, and we don't have Jewish people around anymore. What huge moral progress we had then." That's a terrifying thought. And I think it should make us take seriously the fact that we're very far away from the moral truth right now.
--:-- So I think one of the lessons I draw in the book is that we should not think we're at the end of moral progress and we should not think, "Oh, we should lock in the kind of Western values we have now." Instead, we should think we want to ensure that we spend a lot of time trying to figure out what's actually morally right so that the future is guided by the right values rather than merely whichever happened to win out.
--:-- Dwarkesh Patel: That makes a lot of sense, but I'm asking a slightly separate question. Not only are there possible values that could be better than ours, but should we expect our values—I mean, we have this sense that we've made moral progress, so things are better than they were before or better than most possible other worlds in 2100 or 2200. Should we not even expect that to be the case? Should our prior just be that, yeah, these are kind of meh values?
--:-- Will MacAskill: I think our prior should be that our values are as good as what one would have expected on average. And then you can make an assessment like, "Are the values of the world today going particularly well?" There are some arguments you could make for saying no. Perhaps if the Industrial Revolution had happened in India rather than in Western Europe, then perhaps we wouldn't have wide-scale factory farming, which I think is a moral atrocity.
--:-- Having said that, my all-things-considered view is actually to think that we're doing better than average. If civilization were just a redraw, then things would look worse in terms of our moral beliefs and attitudes. The abolition of slavery, the feminist movement, liberalism itself, democracy—these are all things that we relatively easily could have lost and are huge gains.
--:-- Dwarkesh Patel: If that's true, does that make the prospect of a long reflection kind of dangerous? Because if moral progress is sort of a random walk and we've ended up with a lucky lottery, then you're kind of maybe reversing, maybe you're risking regression to the mean if you just have a thousand years of progress.
--:-- Will MacAskill: I think that moral progress isn't a random walk in general. There are many forces that act on culture and on what people believe. And one of them is just what's right morally speaking, like what the best arguments support. That is a force, but I think it's a somewhat weak force unfortunately.
--:-- The idea of long reflection is getting society into a state where, before we take any drastic actions that might lock in a particular set of values, we allow this force of reason and empathy and debate and good-hearted moral inquiry to guide which values we end up with.
--:-- Dwarkesh Patel: In the book, you make this interesting analogy where humans at this point in history are like teenagers. But another common impression that people have of teenagers is that they disregard wisdom and tradition and the opinions of adults too early and too often. So do you think it makes sense to extend the analogy this way and suggest that maybe we should be Burkean, small-L longtermist, and reject these inside-view esoteric threats?
--:-- Will MacAskill: I think the Burkean arguments for taking history seriously and what kind of moral views have stood the test of time—I think those are important arguments to engage with. My view kind of goes the opposite actually. We are cultural creatures, and in our nature we are very inclined to agree with what other people think, agree with tradition, even if we don't understand the underlying mechanisms.
--:-- I think that works well in a low-change environment. The environment we evolved towards—things didn't change very much. We were hunter-gatherers in small bands for hundreds of thousands of years, millions of years if you include other Homo species. Whereas now we're in this period of enormous change where the economy is doubling every 20 years, new technologies arrive every single year. That is unprecedented. And I think it actually means we should, much more than would make sense historically, just be trying to figure things out more from first principles.
--:-- Dwarkesh Patel: But at current margins, do you think that's still the case? If a lot of EA and longtermist thought is first-principles thought, do you think more history would be better than the marginal first-principles thinker?
--:-- Will MacAskill: I think two things. If it's about an understanding of history, then yeah, I actually would love EA to have a better historical understanding, mainly just as a on-the-margin thing. I think the most important subjects, if you want to do good in the world in an EA way, are philosophy and economics. But we've got that in abundance, whereas there's very little in the EA community in terms of historical knowledge. And I certainly felt like I learned a huge amount over the last few years understanding that better.
--:-- Should there be even more first-principles thinking? Yeah, probably. I think the kind of first-principles thinking paid off pretty well in the course of the coronavirus pandemic. From January 2020, my Facebook wall was completely saturated with people freaking out, or at least taking it very, very seriously in a way that the existing institutions weren't. And they weren't because they were just in this mode of, "Oh, business as usual, don't panic." They weren't properly updating to a new environment and new evidence.
--:-- Dwarkesh Patel: In the book, you point out several examples of societies that went through hardship—I mean, hardship is putting it mildly—but Hiroshima after the bomb, Vietnam after the bombings, and Europe after the Black Death. And they seem to have rebounded relatively quickly. Does this make you think that perhaps the role of contingency in history, especially economic history, is not that large? And it implies a sort of Solow model of growth where, yeah, even if bad things happen, you can just rebound and it really didn't matter?
--:-- Will MacAskill: Yeah, in economic terms. I think that's the big difference between economic or technological progress and moral progress. In the long run, at least, I think economic or technological progress is very non-contingent. It's actually fascinating—some historical contingencies you can see in technology. The Egyptians had an early version of the steam engine. Semaphore was only developed very late, yet it could have been invented thousands of years in the past, similarly with Kay's flying shuttle.
--:-- But in the long run, the instrumental benefits of tech progress and the incentives towards tech progress and economic growth are just so strong that I think we get there in the end in a very wide array of circumstances. In particular, just imagine there are a thousand different societies and none are growing, but one is. Then in the long run, that one becomes the whole economy.
--:-- Dwarkesh Patel: That particular example you gave of the Egyptians having some ancient form of a steam engine—maybe that points towards there being more contingency, because maybe the steam engine comes up in many societies, but it only gets turned into an Industrial Revolution in one.
--:-- Will MacAskill: In that particular case, there's a big debate about whether the quality of metalwork was actually possible to build a proper steam engine at that time. I was mentioning those examples to say that historically, you actually do get some amazing examples of contingency prior to the Industrial Revolution. I think it's still contingency only on the order of centuries to thousands of years. And then in the post-Industrial Revolution world, I think there's much less contingency. There are some, but I think it's much harder to see technologies that wouldn't have happened within decades if they hadn't been developed when they were.
--:-- Dwarkesh Patel: So I guess the general model here of yours is that there are these general-purpose changes in the state of technology, and those are very contingent and it would be very important to try to engineer one of those, but other than that, it's going to get done by some guy starting a startup anyways?
--:-- Will MacAskill: No, I think more generally. Even the case of the steam engine or semaphore that I was pointing to, which historically seem kind of maybe contingent—I think in the long run, they get developed. If the Industrial Revolution hadn't happened in Britain in the 18th century, would it have happened at some point? Or would similar technologies have been developed that were vital in the Industrial Revolution? And I think yes, because there are very strong incentives for doing so.
--:-- If you've got a whole bunch of cultures, and they're all in a random walk, and one hits upon, "Hey, we're going to do industry, we're a culture that's into making textiles and doing that in an automated way," as was true of England in the 18th century, then that economy just takes over the world. And that's why there's this structural reason why economic growth is much, much less contingent than moral progress.
--:-- Dwarkesh Patel: Usually when people think of somebody like Norman Borlaug and the Green Revolution, it's like, "Oh, if you could have done something like that, you'd be the greatest person in the 20th century." Obviously, he's still a very good man. But would that not be your view? You think the Green Revolution would have happened anyways?
--:-- Will MacAskill: Yeah. Norman Borlaug is sometimes credited with saving a billion lives. I think he was huge. I think he was an enormously important and good force for the world. I think it's not the case that had Norman Borlaug not existed, a billion people would have died. Rather, similar developments would have happened shortly afterwards. Perhaps he saved tens of millions of lives, and that's a lot of lives for one person to save. But it's not as many as just simply saying, "Oh, this tech was developed, this tech was used, a billion people who would have otherwise been at risk of starvation used this technology." And in fact, even at the time, there were similar kinds of agricultural developments not long afterwards.
--:-- Dwarkesh Patel: So then counterfactually, what group of people, what kind of profession or career choice, tends to lead to the highest counterfactual impact? Is it moral philosophers?
--:-- Will MacAskill: Not quite moral philosophers, although perhaps sometimes. Sticking on science and technology, if you look at Einstein, the theory of special relativity would have been developed very shortly afterwards. However, the theory of general relativity was plausibly decades in advance. So you do sometimes get these surprising leaps. But I think we're still only talking about decades rather than millennia.
--:-- So who really does make a very long-term difference? I think moral philosophers could be one. I think Marx and Engels made an enormous, very long-run difference. So did religious leaders. I think Muhammad, Jesus, Confucius made an enormous and contingent long-run difference. And moral activists as well—abolitionist campaigners, the Quakers, and other groups too.
--:-- Dwarkesh Patel: If you think that the changeover in the landscape of ideas is very quick today, would you still think that somebody like Marx will be considered very influential in the long future? I mean, communism lasted less than a century, right? Maybe the consequences are huge, but it's all in expectation.
--:-- Will MacAskill: As things in fact turned out, probably Marx will not be very influential over the long-term future. But that could have gone another way. It's not such a wildly different history where rather than liberalism emerging dominant in the 20th century, it was communism. And if it had, it's totally on the table for me that that persists for an extremely long time. The better technology gets, the better a ruling ideology is able to cement its ideology and persist for a very long time.
--:-- You can get a set of knock-on effects where communism wins the war of ideas in the 20th century, in the limit forms a world government based on those ideas, and then via anti-aging technology, genetic enhancement technology, cloning, or artificial intelligence, it's able to build a society that literally persists forever in accordance with that ideology.
--:-- Dwarkesh Patel: The death of dictators is especially interesting when you're thinking about contingency. When Mao dies or when Stalin dies, there are huge changes in the regime, which makes you think the actual individual there was very important, and who they happened to be was contingent and persistent, or at least important in some interesting ways.
--:-- Will MacAskill: For sure. If you've got a dictatorship, then you've got a single person ruling the whole of society, and that means it's heavily contingent on what the views, values, beliefs, and personality of that person are.
--:-- Dwarkesh Patel: Going back to your designation in the book, you're very concerned about fertility, because it seems your model about how scientific and technological progress happens is number of people times average research productivity. And if research productivity is declining and also the number of people isn't growing that fast, then that's concerning.
--:-- Will MacAskill: Number of people times fraction of the population devoted to R&D. Yeah, thanks for the clarification.
--:-- Dwarkesh Patel: It seems that there have been a lot of intense concentrations of talent and progress in history, like Venice, Athens, Bell Labs, or even something like FTX—20 developers making this multi-billion dollar company. Do these examples suggest that organization and congregation of researchers matters more than the actual total amount?
--:-- Will MacAskill: I actually think the model works pretty well. Throughout history, you're starting from a very low baseline, a very low technological level compared to today. Most people aren't even trying to innovate. Or if they're trying to innovate, it might be in things that we wouldn't now call science or technology. It might be theology.
--:-- One argument for why Baghdad lost its scientific golden age is because the political landscape changed such that what was incentivized was theological investigation rather than scientific investigation in the 10th and 11th century AD. Similarly, one argument for why Britain had a scientific and industrial revolution rather than Germany was because all of the intellectual talent in Germany was focused on making amazing music, and that doesn't compound in the way that making textiles cheaper does.
--:-- If you look at Sparta versus Athens, what was the difference? I think it's just that they had different cultures such that in Athens, intellectual inquiry was rewarded. And because they're starting from a much lower base, even just hundreds of people or thousands of people trying to do this thing that looks vaguely like science or intellectual inquiry has enormous impacts.
--:-- Dwarkesh Patel: But if you take an example like Bell Labs in the late 20th century, the low-hanging fruit is mostly gone, but then you have this one small organization that does six Nobel Prizes. Is this a kind of coincidence and lucky break?
--:-- Will MacAskill: I wouldn't say that at all. I should acknowledge that the model where what you're working with is just size of the population times fraction of the population put towards R&D—that's a toy model. It's the simplest model you could have. Bell Labs is punching above its weight. You can obviously create amazing things from a certain environment where not only are you getting the most productive people, but you're putting them in an environment where they're 10 times more productive than they would otherwise be.
--:-- However, when you're looking at the grand sweep of history, those effects are comparatively small compared to the broader culture of a society or the sheer size of a population.
--:-- Dwarkesh Patel: I want to talk about your paper on longtermist institutional reform. One of the things you advocate in this paper is that we should have one of the houses be dedicated towards longtermist priorities. Can you name some specific performance metrics you would use to judge or incentivize the group of people who make up this body?
--:-- Will MacAskill: Sure. I'll caveat with longtermist institutions—I'm actually pretty pessimistic about them in the sense that I have this paper exploring it, but there's this fundamental issue that if you're trying to represent or even give consideration to future people, you just have to face the fact that they're not around and they can't lobby for themselves. So you're going to have co-option by people in the present.
--:-- However, you could have this assembly of people who have some sort of real regulatory power. How would you constitute that? My best guess is you just have a random selection from the population. How would you ensure that the incentives are aligned? There are things you can try. In 30 years' time, their performance will be assessed by a panel of people who look back and say, "Were the policies being recommended here good or not?" Perhaps the people who are part of this assembly have their pensions paid on the basis of that assessment.
--:-- Then the people in that 30 years' time—their assessment of both their policies and their assessment of the previous 30-years-prior futures assembly—get assessed by another assembly 30 years after that, and so on. There's some math and economic analysis such that under certain conditions, this checks out. You have this backwards chaining where people in a thousand years' time are evaluating the people in 970 years' time, who are evaluating the people in 910 years' time. Can you get that to work? Maybe in theory. I'm a little bit more skeptical in practice, but I would love some country to try it and see what happens.
--:-- The other thing I should say is that there is some evidence that you can just get people to take the interest of future generations more seriously by telling them, "This is your role." There was one study that got people to put on ceremonial robes and act as trustees of the future. And they really did make different policy recommendations than when they were just acting on the basis of their own beliefs or self-interest.
--:-- Dwarkesh Patel: If you are on that board that is judging these people 30 years before you, is there something you'd be like, "Okay, this is the metric I care about most"? Expected future GDP growth or something that you think would be the most informative about how good those decisions were?
--:-- Will MacAskill: There are things you could do. It could be GDP of the country. It could be you agree on a set of metrics like homelessness rate, perhaps some expert measure of technological progress. I think you would absolutely want there to be expert assessment of risk of catastrophe as well. We don't have this at the moment, but you could imagine having a panel of superforecasters who are predicting what is the chance of a war between great powers occurring in the next 10 years, and that gets aggregated into a war index. I think that would be a lot more important than the stock market index, and we don't have it. But you could imagine that being fed in as well, because you wouldn't want something which is just incentivizing economic growth at the expense of tail risks.
--:-- Dwarkesh Patel: Would that be your objection to a scheme like Robin Hanson's about just maximizing expected future GDP using prediction markets and making decisions that way?
--:-- Will MacAskill: I think maximizing future GDP is more an idea I associate with Tyler Cowen. But yeah.
--:-- Dwarkesh Patel: Robin Hanson's idea of futarchy—vote on values, bet on beliefs—people can just vote on what collection of goods they want to have, where GDP might be one of them, but also unemployment rate or whatever. Beyond that, it's just pure prediction markets. Again, it's something I'd love to see tried. I think it's an idea in a vein of speculative political philosophy or reasoning about how a society could be extraordinarily different, structured very differently. That is incredibly neglected. Do I think it will work in practice? Probably not. Most of these ideas wouldn't work in practice.
--:-- You do have issues when it comes to prediction markets where they can be gamed or they're just simply not liquid enough. It's been notable since he developed those ideas and really worked on prediction markets—there hasn't been a lot of success at prediction markets, whereas at least there has been a fair amount more success on kind of forecasting. Perhaps you can solve those things. You have laws about what things can be voted on or predicted in the grand prediction market. You maybe have government subsidies to ensure there's enough liquidity. Overall, I think it's pretty promising. You could try it out on a city level or something, see how it goes.
--:-- Dwarkesh Patel: Let's take a scenario where the government starts taking the impact on the long-term seriously and institutes some reforms to integrate that perspective. You can look at the environmental movement for an example of this, where there are environmental review boards that try to assess the environmental impact of new projects and they can repeal any proposals based on this. The impact here, at least in some states and in some cases, has been that groups that have no strong plausible interest in the environment are able to game these mechanisms in order to, in some cases, prevent projects that would actually help the environment.
--:-- Especially when you're talking about longtermism, where it would take a long time to assess what the actual impact of something is, but policymakers are tasked with evaluating the long-term impacts. Are you worried that it would be a system that would be really easy to game by malicious actors? And what do you think went wrong with the way that environmentalism was codified into law?
--:-- Will MacAskill: It's absolutely a worry. It's potentially a devastating worry. You create something trying to represent future people, they're not actually around to lobby themselves, so it can just be co-opted. My understanding of environmental impact statements has been similar, for similar reasons. It's not like the environment can represent itself. It can't say what its interests are.
--:-- What is the right answer there? It's super tough. There are these speculative proposals about having a representative body that assesses these things and is judged by people in 30 years' time. That's kind of the best we've got at the moment. But I think at the moment, we need a lot more thought to see if any of these proposals would actually be robustly good for the long term, rather than just things that are more narrowly focused.
--:-- Regulation to have liability insurance for dangerous biolabs is not in any way about trying to represent the interests of future generations, but it's very good for the long term. So at the moment, I primarily think that if longtermists are trying to change the government, let's focus on a fairly narrow set of institutional changes that are very good for the long term, even though they're just not in the game of representing the future. That's not to say I'm opposed to all such things, but there are major implementation problems with any of them.
--:-- Dwarkesh Patel: If we don't know how we would do it correctly, do you at least have an idea of what went wrong? How could environmentalism have been codified better? Why was that in some cases not a success?
--:-- Will MacAskill: Honestly, I just don't have a good understanding of that. I don't know if it's intrinsic to the matter, or if you could have had some system that wouldn't have been co-opted in the long term.
--:-- Dwarkesh Patel: Are corporations the most longtermist institutions we have today? Their incentive, theoretically, is to maximize future cash flow, which at least explicitly and theoretically should give them an incentive to try to do the most good they can for their own company. That implies that if there's an existential risk, then the company can't be around.
--:-- Will MacAskill: I don't think so. I think different sorts of institutions have different rates of decay associated with them. A corporation, even a corporation that is in the top 200 biggest companies, has a half-life of only about 10 years. They're surprisingly short-lived. Whereas if you look at universities, Oxford and Cambridge are kind of 800 years old. I think the University of Bologna is even older. These are very long-lived institutions.
--:-- Corpus Christi Church at Oxford was making a decision about whether it should have some new tradition that would recur only every 400 years. That's the sort of decision it makes because it's such a long-lived institution. Similarly, religions can be even longer-lived again. I think that natural half-life really affects what sort of decisions a company versus a university versus a religious institution would make.
--:-- Dwarkesh Patel: Does that suggest there's something fragile and dangerous about trying to make your institution last for a long time, if companies try to do that and they're not able to?
--:-- Will MacAskill: Companies are composed of people. Is it in the interest of a company to last for a long time? Well, is it in the interest of the people who constitute the company—the CEO, the board, and the shareholders—for that company to last a long time?
--:-- Will MacAskill: And it's like, no, they don't particularly care. At least most of them don't. Some do, but most don't. With other institutions, I think it goes both ways. This is the issue of lock-in that I talk about at length in *What We Owe the Future*. You get these moments of plasticity, the formation of a new institution—whether that's the Christian church or the constitution of the United States—and that locks in a certain set of norms. That can be really good if the set of norms and laws are good. I think the US Constitution, looking back, seems kind of miraculous. It was the first democratic constitution, as I understand it, created over a period of four months, and it really seems to have stood the test of time. Alternatively, it could be extremely dangerous. There were obviously horrible things in it. The legal right to slavery was proposed as a constitutional amendment. If that had gone in, it would have been a horrible piece of lock-in. So it's hard to answer in the abstract because it really depends on what the thing is that's persisting for a long time.
--:-- Dwarkesh Patel: Do you say in the book that you expect our current era to be a moment of plasticity? Why do you think that is?
--:-- Will MacAskill: Yes, I think the specific time is a moment of plasticity for two reasons. One is that the world is completely unified in a way that's very historically unusual. You can communicate with anyone basically instantaneously, and there's a great diversity of moral views. So we can have arguments, people can debate what's morally correct. It's plausible to me that one of many different sets of moral views might win out or become the most popular ultimately. Secondly, we're at a period where things really can change, but it's a moment of plasticity because it also plausibly could come to an end. In the coming decades or centuries, the moral change that we're used to could end. If there was a single global culture or world government—a global communist state, a global Nazi state, or some other world government that preferred ideological conformity—then combined with technology, it becomes unclear why that would end over the long term. The key technology here is artificial intelligence. The point in time, which may be sooner than we think, where the rulers of the world are digital rather than biological could persist. Once you've got that plus global hegemony of a single ideology, there's not much reason for that set of values to change over time. You've got immortal leaders and no competition. The other sources of value change over time can be accounted for too.
--:-- Dwarkesh Patel: But isn't the fact that we are in a time of interconnectedness that won't last if we settle space a reason for thinking that lock-in is not especially likely? If your overlords are many millions of light years away, how well can they control you?
--:-- Will MacAskill: The way I see it, the control will happen before the point of space settlement. If one day we took to space and there are many different settlements in different solar systems pursuing different visions of the good, then you're probably going to maintain diversity for a very long time. Given the physics of the matter, once the solar system has been settled, it's very hard for other civilizations to come along and conquer you, at least if we're at a period of technological maturity where there aren't new groundbreaking technologies to be discovered. But I'm worried that the control will happen earlier. I'm worried the control might happen this century, within our lifetimes. I don't say that's very likely, but I think it's seriously on the table—10% or something.
--:-- Dwarkesh Patel: Going back to the long-term of the long-term as a movement, there are many instructive foundations that were set up about a century ago, like the Rockefeller Foundation and Carnegie foundations. They don't seem to be especially creative or impactful today. What do you think went wrong? Was it value drift, or just some decay of competence and leadership and insight?
--:-- Will MacAskill: I don't have super strong views about those particular examples, but two natural thoughts come to mind. One is that for organizations that want to persist a long time and keep having influence, historically they've tended to specify their goals in far too narrow terms. A fun example is Benjamin Franklin. He invested a thousand pounds for each of the cities of Philadelphia and Boston to pay out after 100 years and then 200 years for different fractions of the amount invested. But he specified it very specifically—it was to help blacksmith apprentices and so on. It's like, oh man, this doesn't make much sense once you're in the year 2000. Whereas he could have said something much more general, like "for the prosperity of people in Philadelphia" or "for the prosperity of people in Boston," and then it would have had at least plausibly more impact. The second is just a regression to the mean argument. You have some new foundation doing an extraordinary amount of good, as I think the Rockefeller Foundation did. Over time, if it's exceptional in some dimension, it's probably going to get closer to average on that dimension. As a matter of changing the people involved, if there are some people who are exceptionally competent and far-sighted, the next people just statistically are probably going to be less so.
--:-- Dwarkesh Patel: Going back to that dead hand problem, if you specify your mission too narrowly, it doesn't make sense in the future. Is there a trade-off where if you're too broad, you give future actors—maybe they're malicious, or maybe they're just not as smart or creative as you—the ability to take the movement in ways you would not approve of? So if it's just "do good for Philadelphia," it turns into something Ben Franklin would not have thought is good for Philadelphia.
--:-- Will MacAskill: It depends crucially on what your values and views are. If Benjamin Franklin—I don't think this was true—but if he was like, "No, I just really care about blacksmith's apprentices and nothing else matters," then he was correct to specify it in that narrow way. But as a matter of fact, certainly my own values, but I think more generally, they tend to be quite a bit more broad than that. Secondly, in general, I expect people in the future to be smarter and more capable. That's certainly the trend over time. If we're sharing similar broad goals and they're implementing it in a different way, then probably they're right and I'm wrong.
--:-- Dwarkesh Patel: Let's talk about how good we should expect the future to be. Have you come across Robin Hanson's argument that in the future we'll all just be subsistence-level ems because there'll be a lot of competition, and you'll try to minimize compute per digital person, which will just be a miserable, barely worth living experience for every entity?
--:-- Will MacAskill: I'm familiar with the argument, but we should distinguish the idea that ems are at subsistence level from the idea that they would have bad lives. Subsistence means that given their income per capita and population growth, if they were any poorer, deaths would outweigh additional births. That actually doesn't tell you about their well-being. You could be very poor as an emulated being, but you're just in bliss all of the time. That is perfectly consistent with Malthusian theory. That future still could be very good. Those ems, while at subsistence, still could have lives thousands of times better than ours.
--:-- Dwarkesh Patel: Speaking of being poor and happy, there was a very interesting section in the chapter where you mentioned a study you had commissioned trying to find out if people in the developing world had a life worth living. It turns out that 19% of Indians would not want to relive their life every moment, but I think it was 31% of Americans who said they would not. Why are Indians seemingly much happier at less than a tenth of the GDP per capita?
--:-- Will MacAskill: I think the numbers are lower than that from memory. It depends exactly on the question asked, but I think it was something more like 9% of Indians wouldn't want to live their lives again if they had the option, and 13% of Americans or something. But you are right that on this metric of how many people are happy to have lived, the Indians we surveyed were more optimistic about their lives, happier with their lives than people in the US were. Honestly, I just don't want to generalize too far from that because we were sampling comparatively poor Americans and comparatively well-off Indians, so perhaps it's just a sample effect. There are also weird interactions with Hinduism and the belief in reincarnation that could mess up the generalizability of this as well. So I basically don't want to draw any strong conclusion from that. But it is pretty striking as a piece of information, given that normally when you look at people's well-being, richer countries are considerably happier than poorer countries on average, at least.
--:-- Dwarkesh Patel: But you do generalize in the sense that you use it as evidence that most lives are worth living today, right?
--:-- Will MacAskill: Exactly. I put together various bits of evidence where very approximately 10% of people in the United States and 10% of people in India seem to think that their lives are net negative—they wouldn't want to be reborn and live the same life if they could. If you look at other studies as well, there's another study that asks people in the US or other rich countries about how much of their conscious life they would want to skip if they could. Skipping means you blink and come to the end of whatever activity you're engaging with. Perhaps I hate this podcast so much that I would rather be unconscious than be talking to you, in which case I would have the option of skipping—obviously not the case—but if you look at that and also ask people about the trade-offs they would be willing to make as a measure of intensity of how much they're enjoying or not enjoying a certain experience, you get the conclusion that from memory, a little over 10% of people regarded their life that day as worse than if they'd been unconscious the entire day.
--:-- Dwarkesh Patel: Jumping topics a little bit. On the 80,000 Hours podcast, you said that you expect scientists who are explicitly trying to maximize their impact might have an adverse impact because they might be ignoring foundational research that wouldn't be obvious in this way of thinking but might be more important. Do you think this could be a general problem with longtermism—that if you're really trying to find the most important things that are important long-term, you might be missing things that wouldn't be obvious thinking this way?
--:-- Will MacAskill: I think it's a risk. Among the ways people could argue against my general set of views, one way I find compelling is this. I argue that in general we should be doing fairly specific and targeted things: trying to make AI safe, trying to govern the values of AI, trying to reduce worst-case pandemics that could kill us all, trying to prevent third world war, ensuring good values are promoted and avoiding value lock-in. But some people—Tyler Cowen and Patrick Collison take this line—argue that it's very hard to predict the future impact of your actions, and it's a mug's game to even try. Instead, you should look at what things have done loads of good consistently in the past and try to do the same things. They might argue that means technological progress or boosting economic growth. I dispute that, but it's not something I feel I can give a completely knockdown argument to because it's about when we'll find out who's right—maybe in a thousand years' time. But one piece of evidence is the success of forecasters in general. People in effective altruism were realizing that the coronavirus pandemic was going to be a really big deal from a very early stage, worrying about pandemics far in advance. I think there are some things that are actually quite predictable. Moore's law has held up for over 70 years. The idea that AI systems are going to get much larger, that leading models are going to get more and more powerful—that's on trend. Similarly, the idea that we will soon be able to develop viruses of unprecedented destructive power—that's not actually that controversial a claim. So even though for loads of things it's very hard to predict and there will be tons of surprises, there are some things—especially fairly longstanding technological trends—where we can make reasonable predictions, at least about the range of possibilities that are really on the table.
--:-- Dwarkesh Patel: But it kind of sounds like you're saying the things we know are important now are important now. If something did turn out a thousand years from now to be very important, it wouldn't be salient to us now.
--:-- Will MacAskill: With me versus Patrick Collison and Tyler Cowen, who is correct? In some sense, we will only get that information in a thousand years' time because we're talking about which strategy will have a bigger impact on the long term. We might get suggestive evidence earlier. If me and others engaging in longtermism are making specific measurable forecasts about what is going to happen with AI or advances in biotechnology, and we're able to take action such that we are relatively clearly reducing certain risks, I think that's pretty good evidence in favor of our strategy. Whereas if in contrast they're doing all sorts of stuff, not really trying to make firm predictions about what's going to happen, but then things pop out where we think, "Oh, that was really good from a long-term future perspective," after we measure this in 10 years' time, that would be good evidence for their view.
--:-- Dwarkesh Patel: What you're saying earlier about the contingency of technology implies that even given their worldview, maybe you should think that if what's had the most impact in the past is changing values, then economic growth might not be the most important thing, or trying to change the rate of economic growth.
--:-- Will MacAskill: I really do take seriously the argument of looking at how people have acted in the past, especially for people who were trying to make a long-lasting impact. Towards the end of the 19th century, John Stuart Mill and the other early utilitarians had this longtermist little wave where they started taking the interests of future generations very seriously. Their main concern was that Britain would run out of coal and future generations would be impoverished. It's pretty striking because they had a very bad understanding of how the economy works. They didn't predict that we would transition away from coal because of continued innovation. And they had enormously wrong views about how much coal and fossil fuels there were in the world. That particular action just didn't make any sense given what we know now. In fact, trying to keep coal in the ground, given Britain at the time—we're talking about much lower amounts of coal, so the climate change effect is negligible at that level—it actually probably would have been harmful. But we could look at other things John Stuart Mill could have done, such as promoting better values. He campaigned for women's suffrage. He was the first British MP, I think in fact even the first politician in the world, to promote women's suffrage. That seems to have stood the test of time. That's one historical data point, but potentially we can draw a more general lesson there.
--:-- Dwarkesh Patel: Do you think the ability of global policymakers to come to a consensus is on net a good or a bad thing? On the positive, maybe it helps prevent some dangerous tech from taking off. On the negative side, it prevented human challenge trials, maybe it causes some sort of lock-in in the future. On net, what do you think about that trend?
--:-- Will MacAskill: The question of global integration is absolutely double-sided. On the one hand, it can help us reduce global catastrophic risks. The fact that the world was able to come together and ban chlorofluorocarbons was one of the great events of the last 50 years, allowing the hole in the ozone layer to repair itself. But on the other hand, if it just means we all converge to one monoculture and lose out on diversity, that's potentially pretty bad. We could actually lose out on most possible value that way. The solution is to do the big good bits and not have the bad bits. In a liberal constitution, you can have a country bound in certain ways by its constitution and certain laws yet still enable a flourishing diversity of moral thought and different ways of life. Similarly, in the world, you could have very strong regulation and treaties that deal with certain global public goods—mitigation of climate change, prevention of the development of the next generation of weapons of mass destruction—without thereby having some very strong-armed global government that implements a particular vision of the world.
--:-- Dwarkesh Patel: Which way are we going at the moment?
--:-- Will MacAskill: It seems to me like we've been going in a pretty good and not too worrying direction. But I think that could change.
--:-- Dwarkesh Patel: It seems the historical trend is that when you have a federated political body, even if constitutionally the central power is constrained, over time they just tend to gain more power. You can look at the US, you can look at the European Union. That seems to be the trend.
--:-- Will MacAskill: I think that's potentially a worry, depending on the culture that's embodied there. It might not be if the culture itself is liberal and promoting of moral diversity and moral change and moral progress. But that needn't be the case.
--:-- Dwarkesh Patel: Your theory of moral change implies that after a small group starts advocating for a specific idea, it may take like a century or more before that idea reaches common purchase. To the extent that you think this is a very important century—I know you have disagreements about that with others—does that mean that maybe there just isn't enough time for longtermism to gain power that way by changing moral values?
--:-- Will MacAskill: There are lots of people I know and respect very well who think that artificial general intelligence will very likely lead to singularity-level technological progress, extremely rapid rates of technological progress, and that that will happen more likely than not within the next 10 or 20 years. If so, then you're right. Value changes are something that pays off slowly over time. In the world today, I talk about moral change taking centuries. That's definitely more to do historically. I think we can have much faster change. The growth of the effective altruism movement is something I know well, where it's growing at something like 30% per year. Compound returns mean that's not change that happens on the order of centuries. If you look at other moral movements, like the gay rights movement, very fast moral change by historical standards. So if you're thinking we've got 10 years till the end of history, then probably don't just very broadly try and promote better values. But I think we should have at least a very significant probability mass on the idea that we will not hit some historical end this century. In those worlds, promoting better values could pay off very well.
--:-- Dwarkesh Patel: Have you heard of the slime mold time old potato diet?
--:-- Will MacAskill: I have indeed heard of the slime mold time old potato diet, and I was tempted as a gimmick to try it. But they're onto something because, as I'm sure you know, the potato is close to a superfood. You could survive indefinitely on just buttery mashed potatoes if you occasionally supplement with something like lentils or oats.
--:-- Dwarkesh Patel: Interesting. A question about your career. Why are you still a professor? Does it still allow you to do the things you would otherwise have been doing, like converting more SPFs and making moral philosophy arguments for EA? I'm curious about that.
--:-- Will MacAskill: It's fairly open to me what I should do. My best guess is that I do spend significant amounts of time co-founding organizations or being on the board of those organizations I've helped to set up. More recently, working very closely with the Future Fund, Sam Bankman-Fried's new foundation, and helping them do as much good as possible. That being said, if there's a single best guess for what I ought to do longer term, and certainly that plays to my strengths better, it's developing ideas, trying to get the big picture roughly right, and then communicating them in a way that's understandable and gets more people to get off their seats and start to do a lot of good, especially for the long term. I think I've had a lot of impact that way. From that perspective, having an Oxford professorship is pretty helpful. I can work on these big picture problems because they're too big for companies or other institutions. Some of these are: Is the future good rather than bad? If there was a global civilizational collapse, would we recover? How likely is this long stagnation? There's almost no work done on any of these topics. Companies aren't interested in grand-scale research. Academia, I think, has developed a culture where you don't tackle such problems. Partly that's because they fall through the cracks of different disciplines, and partly because they just seem too big, too grand, or too speculative. Academia is much more in the mode of making incremental gains in our understanding. But it didn't always used to be that way. If you look back before the institutionalization of academic research, philosophers would have grand unifying theories of not just ethics and political philosophy but also metaphysics, logic, epistemology, and probably the natural sciences and economics. I'm not saying all academic inquiry should be like that, but should there be at least some people whose role is to really think about the big picture? I think yes.
--:-- Dwarkesh Patel: Will I be able to send my kids to MacAskill University? What's the status on that project?
--:-- Will MacAskill: I'm really pretty interested in the idea of creating a new university. It is a project that I've been in discussion about with another person who's very excited about making it happen. Will it go ahead? Time will tell. But I think you can do education far, far better than currently exists. I also think you can probably do research far, far better than currently exists. It's extremely hard to break in and create something that's very prestigious because the leading universities are almost all hundreds of years old. But maybe it's possible, and I think it could generate enormous amounts of value if we were able to pull it off.
--:-- Dwarkesh Patel: Excellent. So the book is *What We Owe the Future*, and I understand pre-orders help a lot. It was such an interesting read because how often does somebody write a book about even the questions they consider to be the most important, even if they're not the most important questions? Just that kind of big picture thinking, but also looking at very specific questions and issues that come up. It was super interesting. Thank you so much. Anywhere else they can find you or any other information they might need to know?
--:-- Will MacAskill: Sure. *What We Owe the Future* is out on August 16th in the US and September 1st in the United Kingdom. If you want to follow me on Twitter, I'm at @WillMacAskill. If you want to try and use your time and money to do good, Giving What We Can is an organization that encourages people to take a pledge to give a significant fraction of their income—10% or more—to the charities that do the most good. It has a list of recommended charities. 80,000 Hours, if you want to use your career to do good, is a place to go for advice on what careers really have the biggest impact. They provide one-on-one coaching too. If you're feeling inspired, if you think you actually really want to do good in the world, you care about future people and want to help make their lives go better, then as well as reading *What We Owe the Future*, Giving What We Can and 80,000 Hours are the sources you can go to and get involved.
--:-- Dwarkesh Patel: Awesome. Thanks so much for coming on the podcast. This was a lot of fun.
--:-- Will MacAskill: Thanks so much. I loved it.
--:-- Dwarkesh Patel: Cool. Thanks for watching. I hope you enjoyed that episode. If you did and you want to support the podcast, the most helpful thing you can do is share it on social media and with your friends. Other than that, please like and subscribe on YouTube and leave good reviews on podcast platforms. Cheers. I'll see you next time.
