Effective Altruism and its Critics

Introduction to Problem

0:00My friend Curtis has a problem, a very good problem to have. He has a very wealthy father. For a number of years, as part of his estate planning, I guess, my father has been giving me and my sister a pretty large amount of money every year. And this is wonderful to have when you're a young man, when you're broke in college,

0:33in graduate school. As I got older and had my own income, I started to sort of realize how unearned this money was and how unneeded it was. So I'm helping Curtis with his problem.

Charity Options

0:47He's going to give his money to charity. But what charity? Or what charities? That's where I come in. Because it turns out, in the last 10 years or so, philosophers have built a movement around charitable giving. It's called effective altruism. And there are many critics of that movement. What better way to get into those issues than to try to convince Curtis? So Curtis, you have agreed to come on HiFi Nation for this episode and listen to arguments

1:18about the proper way to give to charity. Well, I've never really had a systematic approach to giving to charity. So what I'm going to do today is hear those arguments. And at the end, I'm going to make a donation to a cause related to the argument that I find the most persuasive.

1:38From Slate, this is HiFi Nation, philosophy in story form. Recording from Princeton University, here's Barry Lamb.

Effective Altruism Explained

1:47Effective altruism is one of the first global movements that has its roots in secular philosophical ethics. It's grown to such an enormous size that it's a major player in philanthropy today. You're going to hear about it and its critics. A lot of people think that effective altruism makes some serious moral mistakes. And there's been a big scandal recently.

2:17A leading donor in the movement has been charged with fraud and theft. Everyone agrees that there's a wrong way to give to charity. You shouldn't give your money to scams or charities that spend a lot of money on trying to help others, but they don't actually help anyone. But this is where agreements end.

2:41Because now there are ways of measuring the effectiveness of charities. And judged on these two dimensions for the past decade, effective altruism organizations emerged victorious, acquiring billions of dollars in pledges toward causes it deemed highly effective at doing good. Today, that movement is at a crossroads. Because of changes in focus, because of some scandals, and because many of its founders want

3:13us to give more and more money to fighting a catastrophic digital future.

3:21Today, we'll see how persuasive each side can be to an ordinary person with a desire to do good with his money.

Personal Experience with Charity

3:33All right, Curtis, let's get started by you telling us how you've approached charitable giving in the past with your dad's money. One year, the arrival of this check that he would send came at about the same time that I found out about an organization that a friend of mine had started locally here, which was providing personal care products to students at the local high school who didn't have those products. Things like shampoo, deodorant, feminine care products.

4:06It's called Teen Closet. Yeah, this was the first time I really thought that my money could have a meaningful impact on someone. It made me think about how I took it for granted to have shampoo and deodorant, and what it would be like to not have those things, and what a loss of dignity that would be. And I just thought, wow, if I could give some of this money that just fell into my lap, some kid will have these things I took for granted, and I just felt very empowered around that, and it excited me, and I wrote them

4:39a check. The other thing Curtis did is give money at the end of the year to charities that sent him, you know, persuasive mailers or brochures, a mix of the local library, some international organizations, a lot of things he gave to in the past. It's actually pretty close to how I do it.

Effective Altruism 1.0

4:56So the very first thing I want to present to you, I am going to call Effective Altruism 1.0. This is something that came about about a decade ago, and it was inspired by some philosophers and some economists. Effective altruism has grown into a movement. It's gone, you know, slightly different directions in recent years, but I'm going to talk about those original arguments. Philosopher Richard Yetter-Chapel of the University of Miami will be the one giving voice to these

5:29views. When you have some pot of money that you're donating for the purpose of trying to do good with this resource, you should try to do as much good as you can with it. In that sense, it's got this maximizing element. For instance, if there is a stranger that's in front of you, and you have some money, and there's actually a couple of things you can buy for them, they're very, very hungry, right? The most good you can do for them might be to buy them food. There might be something else you can do with that money that would be okay.

5:59Like, you could pay your friend to entertain them, for instance, right? And like, that's a little bit of good. Like, they'll be entertained and hungry rather than just hungry. But the idea is supposed to be, okay, you should do the most good that you can with this bit of money. So that's the first piece. The second piece is that you ought to be neutral, impartial when you're helping strangers. You should think of strangers as all on a par with each other.

6:31There isn't one kind of stranger that you should help more than another kind of stranger. I feel like we should really treat all strangers equally. And, you know, we should care about their well-being and want them to be better off rather than worse off. It doesn't seem to me that there's any principled case for caring more about, like, fellow countrymen than foreigners. They're all equally strangers to you. The fact that you're in the same place doesn't give you any principled basis for caring more about them. So now if there are two strangers in front of you, the fact that one's an American and one isn't doesn't matter, that one is your ethnic group and the other isn't doesn't matter.

7:04And the fact that one might be like 10 feet farther from you than the other doesn't matter. What matters is what's the most good you can do in this circumstance. So if one of them is hungry and the other one is just bored, so you could pay to feed the hungry person or you could pay to entertain the bored one. Well, the most good you can do is to alleviate hunger. What do you think, Curtis? I'm just thinking about the example of feeding versus entertaining. Like, if I feed somebody, they're hungry again the next day.

7:36But if I let's you could imagine entertaining them in a way where it like, you know, the memory of that experience like helps them be hopeful or believe in themselves or I don't know, survive the future hunger. So those are the kinds of tricky calculations that you'll need to make if you buy into this idea that in giving you should be maximizing. So while there are going to be tricky cases, I think that the effective altruists tend to think that if you're talking about global charities, right, they think that there are

8:10some relatively untricky assumptions that you can make that people who are suffering from, you know, certain kinds of diseases, that's just much weightier than other people not having music in their life. So there is the element of the prevention of suffering, and then there's the element of promoting happiness. That's basically classical utilitarianism.

8:36So the next part of the view is that there actually are right and wrong ways to prioritize these things. Relieving suffering of strangers is really better than promoting some musical joy of other people. So a lot of people giving money to opera and theater groups for children or things like that, according to the effective altruists, they're making a moral mistake. Suffering people just are more important than kids who get joy out of art.

9:08If you end up donating to like opera or some other thing that's like good to some extent, but maybe arguably not comparable with saving people's lives, it reveals that your motivation wasn't genuinely altruistic in the right kind of way. There's one qualification though. So your own kid's joy, for instance, buying them a toy car while some other stranger is suffering. That can be okay because on Richard Yetter-Chapel's view, you're allowed to be partial to some people in your life.

9:38So I think the most reasonable forms of partiality are just ones that are based on personal connections to your loved ones. Friends and family seems reasonable to care more about the people that you have those close relationships to than about total strangers. But, and here's the key part of his view, the actions that you, Curtis, do on behalf of friends, family, and loved ones, those aren't really altruistic acts. Those are acts that are at best acts of self-interest or mutual self-interest.

10:09They're acts of caring, but they're not really doing good for strangers, which is this other thing that you really ought to do. Now, do you think you agree with that idea? I mean, I agree that it's not as altruistic to do good for a group you are partial to than to a group you are not partial to. So the example of the giving that was so exciting to me, giving to the local organization, giving personal care products to, you know, local students, is a group I'm partial to in that

10:45it's my local community. And I felt really good about doing that. He would challenge me. He would say, well, you're not really being altruistic. Yes, I think that's right. To be fair to you, I think there's another interpretation, which is just that you were being altruistic. These are strangers. You just weren't acting optimally about your impartiality. So you should think of the kids without shampoo in your local community. But you should also think about what kids halfway around the world are also lacking.

11:20But another way of thinking about it is actually you were acting on your own behalf because these are people you care about because they're local to you. It affects you in your local community, whether they have personal care hygiene products. On that view, then it's like you should feel good, but you should feel good in the way that you should feel good for doing something for your wife, for your child, right, for your neighbor, for your colleague. And that's OK. That's not charitable giving. It's not altruism. To put it in the terms that I described, I was excited to provide for someone something that

11:53I took for granted. Someone halfway around the world can't even take for granted being alive in the way that I take for granted staying alive, let's say. And so I should be supporting other people in staying alive. Yes. And this is a central feature of effective altruism. The idea that in any act of giving, it's not coarse and it's not morally wrong. Instead, it's morally obligatory to think of a trade-off.

12:24If with the same amount of money, the trade-off was somebody getting shampoo and deodorant locally or somebody staying alive somewhere else, what is the most good? It's clearly somebody staying alive. And doesn't that mean in a wealthy place like the United States, you will just never give money to anything in the United States? I think it does mean that.

12:53On the effective altruist scheme.

Critique of Effective Altruism

12:55This is Savannah Perlman. She's a critic of effective altruism. It's always wrong, or at least non-ideal, to manage local poverty because poverty within the U.S. is always going to be more expensive than poverty abroad. So, on some level, you have to be willing to say to, not only to the kid that doesn't have deodorant, but to the library that I value so much and the program that gets wrongly

13:26incarcerated people out of prison. You'd have to be willing to kind of say to all of them, you in your current situation have more than these people who are not able to stay alive. So, until all of that is taken care of, you get nothing. That's right. You know, one way of thinking about this is that it's because it's so cheap to save someone's life in certain parts of the world and so expensive to save their lives in this country that this happens to be true.

13:57The second thing to say is that it's not completely total in that way. Like I've mentioned before, there's a way of construing the local giving as, well, it's because it affects me. But in that case, the right way to think about it is not as altruism, but it's like self-interest. It's indirect self-interest. And the impact of that on me is just, if I want to do those things, I have to stop congratulating myself on it. Yes. Just own it as a self-interested investment. That's correct.

14:27I mean, I got to say, the idea of congratulating yourself and feeling good about stuff, the effective altruists that I meet, they are not happy, generally, people. Like, they seem to be overly burdened by suffering all the time. Honestly, when they give, it's not like they feel happy about things. They're just like, oh, God, I've alleviated some suffering, but there's just so much more suffering. So, there's a couple more things I want to say about this, right? So, altruism is indifferent to strangers.

14:59It's indifferent to species. Yes, absolutely. So, yeah, species is not an intrinsically relevant difference. Just if you're talking about relieving a certain amount of suffering that either a chicken would experience or a human, there's no intrinsic reason to prefer to relieve the human suffering than the non-human animal's suffering, as long as it's of equal intensity. So, like, if you have a scale of suffering of animals that dwarfs, say, the suffering of one human being, then it very well could be that the right thing to do to maximize the

15:29altruistic acts, your altruistic acts, is to give to causes concerning animal welfare. Aren't there organizations that are trying to eradicate malaria by basically exterminating the mosquito that spreads it? That's right. You know, making it extinct? Yes, absolutely. So, the effect of altruism, people would not support donating to that? No, no. Here's their advice to you. It's called GiveWell.org. GiveWell.org is the organization that actually measures these things.

16:00Like, what charitable organizations in the world are the ones that are most effective in reducing the suffering and promoting the well-being? For every buck that you give, you get the most bang for your buck. Mosquitoes are on the top of the list. They're all malaria prevention interventions. It's not the eradication of mosquitoes. It's bed nets. It's a single pill that'll prevent malaria. It is vitamin A supplements to prevent blindness, things like that.

16:34Here's philosopher Richard Yedder-Chappell. When you're comparing different charities, some of them save people's lives, others improve people's lives. How do you make those trade-offs against each other? And so, there's this idea of a quality-adjusted life year, where if you gave someone five extra years of life in full health, that would be five quality-adjusted life years. If you basically made their quality of life go from 50% up to 100% for 10 years, that improvement in their life would also count as five quality-adjusted life years. And so, it's trying to get a common scale for making these different kinds of comparisons.

17:05The recommendation on this way of thinking is going to be any one of these top four charities. It'll be medicine to prevent malaria, malaria bed nets, $5 provides one net, and they have to be replaced every so often. Supplements to prevent vitamin A deficiency, this is the most cost-effective. It's $3,500 per life saved. And then there's cash incentives for routine childhood vaccinations. It turns out that there's a lot of barriers to people in these developing countries.

17:36The travel to a medical clinic, $160 will give a child all of their routine vaccinations. The idea would be to get the most quality-adjusted life years for the amount of money you're donating. Those are the arguments for effective altruism 1.0. Tell me what you think. You know, I have the feeling that the kind of suffering they're trying to eliminate is so overwhelming, so vast in its scale. What's their argument about how we should feel about the overwhelming scale of this suffering

18:10relative to the size of the contribution a normal individual can make to it? Well, every problem can be understood at a large scale. So, you gave to Teen Closet. But there is a scale of a problem. There are millions and millions of teenagers who don't have care products, right? Yeah, got it, got it, yeah. If you think about it that way, then you're like, oh, I'm barely making a dent. But another way of thinking about it, I made a huge dent. There's one kid who has, you know, care product.

18:40And then the way they think about it, too, is it may be a huge problem. Yeah, yeah. But when you give to a highly effective charity, I can actually tell you how many years you've added to a single person's life for the amount of money that you've given. Something about the monetary calculations, there's something like grim about it to me. When they say you can save a life for this amount of money, there's a part of me that says, well, what kind of a life will that be?

19:12You know, to live in poverty, say, or, you know, war-torn region or evading malaria feels like just the first challenge in life. But I think if I have that reaction, I want to be very conscious that, in a sense, I'm assuming that, like, some lives don't deserve to exist or, you know, are not worth saving. And I recognize that as an artifact of a racist attitude toward Africa and, you know, the global South in general.

19:49Which you want to fight against, right? Yeah, yeah. So I appreciate the way the effect of altruism challenges that. You know, what it's saying is that there is more to life and none of those things exist without life. And the grimness is, in fact, the thing you're trying to address.

20:07I do think the, like, economic calculations and everything, it feels like they take not just the joy out of the giving, but the, like... Right. The humanity out of the giving. They take the humanity out of it. I don't exactly know what I mean.

20:21It's also making me realize that I might be okay with not being an altruist. With giving up altruism as a motivation and seeing it as a self-investment.

20:34Hi-Fi Nation will return after these messages.

Alternative Approach to Charity

20:37Okay, let's meet our next advisor. Hi, my name is Savannah Perlman. I'm an associate instructor in the Department of Philosophy at Indiana University Bloomington. So Savannah Perlman is a critic of effective altruism. I think that it's a great thing to do to donate one's money to effective charities. I'm not recommending that people avoid donating to highly effective charities. However, I think there's a possibility that we can have multiple ways to discharge this broader duty of helping our fellow humans,

21:15or sometimes referred to as our duty of benevolence. In fact, the kind of thing that she likes giving to is the kind of thing that inspired you to give to Teen Closet. Things that involve the display of care, and it's much easier to display care to people that you can, like, be face-to-face with, or people you can talk to. For instance, where I live in Bloomington, Indiana, we have a large homeless population. When I do something like donate $100 to buy a heater for my local homeless community to have in their encampment,

21:54it's not that I'm doing the morally wrong thing to do. She likes mutual aid, which is establishing a local network of people who can help each other out when one has a particular need. Last year, I saw a woman within my mutual aid group post on our Facebook platform saying, my teenage daughter had passed away from cancer in 2019, and I haven't gotten my hair done since then. Is anybody willing to come over and cut and color my hair?

22:28In any other context, we would count people involved in mutual aid as doing really good things, charitable things. But in the effective altruism context, you would say that they were being slightly immoral because they're actually letting other people die so that they can cut and color the hair of a human being. That's the basis of her criticism of effective altruism. She calls it the flattening of values. She thinks that there are some things that are valuable, irrespective of how much good they actually can do in the world.

23:06Imagine that you were to take care or solidarity or community building and try to assign a point, some sort of scientific point system based only on the good consequences of participating in those acts. You're losing some of the major value of what it is to be a human experiencing care or taking part in solidarity.

23:33A good example is the Warsaw Ghetto Uprising. The Jews within the ghetto knew that they were not going to actually be able to defeat the Nazis, but they came together in solidarity in that situation, knowing that despite the fact they would not be able to create the consequence of escaping the ghetto, there was something worthwhile about coming together and being in solidarity and taking a side against their situation. And that's precisely the sort of important moral content that I worry that effective altruism will ignore.

24:11Perlman's a pluralist about charitable values. She thinks that maximizing the effectiveness of your giving is a good value, but she also thinks exercising care for local people through mutual aid is a good charitable value. And she thinks lots of other ways of expressing goodness are good charitable values. And there's no single way to compare or calculate which one is the right way.

24:46This is so antithetical to effective altruist thinking. Like Richard Yetter Chappell said that people like Perlman are explicitly saying that coloring and cutting someone's hair is just as important as saving someone's life. And he thinks that that's repugnant. But in the same way, people like Savannah Perlman think that doing that kind of comparison is repugnant. That's the most fundamental disagreement.

25:19If the kind of reductio ad absurdum version of effective altruism is like the closure of all local arts institutions, the folding of all local nonprofit organizations, and the redirection of those resources to malaria prevention, is the extreme version of what she's suggesting the one where people in developing nations are just left to die of malaria because, like, they don't have access to the mutual aid networks that are so core to the kind of altruism she's describing?

25:53I think that if you wanted to draw out the extreme possibility, I think that's got to be yes. It's a possibility of her view, but it depends on there being no one that values the lives of malaria victims in the developing world. And I'll just say, I mean, through the course of the pandemic, like many people, I developed a wider network of acquaintances, I guess I would say, online through, you know, apps like Discord.

26:25And I think I've started doing more charity of the type Perlman is describing, where someone who basically is a stranger to me, if they post, hey, I'm having trouble making rent this month, I've started throwing them 50 bucks. That's what she's describing, right? That's exactly what she's describing. She's not prescribing it. It's the values that she likes to enact. What's important for her is not to prescribe, because I think that would be partaking in this whole values flattening that she worries about. What resonates with me in her critique of effective altruism is the rejection of a prescriptive form of altruism.

27:04It's unappealing to be told there are right and wrong ways to direct your money, even though on some level, I actually believe there are right and wrong ways. So maybe what I mean is it's unappealing to be told the right way is a pretty narrow way. Okay, so here is Savannah Perlman's recommendation to you.

Effective Altruism 2.0

27:28Curtis, you should donate to whatever cause you care about, whatever communities you feel connected to and want to help. My view is that it's perfectly okay to donate to any charitable cause that's going to help others, especially if you're more likely to give more than just your money down the line. And that's because at the end of the day, it's not about the money. It's not about the numbers. It's about connection and care.

27:56All right, let's move on to the next one. So I'm going to call this Effective Altruism 2.0. I'm Shaquille Hirschem. I'm the head of communications at the Center for Effective Altruism. Effective Altruism, this very same movement, has in the last few years taken a turn to asking individuals to redirect their money from malaria bed nets, malaria medicine, and global public health to other causes, causes that try to mitigate the risk of total human extinction.

28:29A big one is pandemics. In the very near future, especially as synthetic biology develops, you might be able to have engineered pathogens that are much, much more virulent, much, much more deadly, and could pose the threat of extinction. Nuclear war is another big one. There are very few resources that go into mitigating it or trying to tackle the risks of nuclear war. And then the third big bucket is risks from advanced artificial intelligence. If an AI became smarter than us, which seems possible, and pursued the wrong goals,

29:03we could lose control over the future of humanity. So whereas EA 1.0 wanted people to focus their money mostly on global public health, which is human suffering now, the shift is to have people focus mostly on human suffering in the future. It's called long-termism. Here's Richard Yetter-Chapel. It certainly seems commonsensical that we should care at least some extent about future generations. You know, we wish past generations had been more concerned about climate change

29:36and other problems that are now bedeviling us that, you know, could have been alleviated somewhat by earlier attention. And so I think in the same way, we just want to be a bit more forward-looking ourselves. But then philosophers want to take this common sense idea and then try to extend it in a really principled way and say, actually, there's no limit in time to when people matter. You could imagine just like spatially, no matter how far away in space you are from me, you still morally matter. Someone who exists in the far future a million years from now in principle still matters just as much.

30:07And if future people matter, then whether they get to live matters. I think the key idea here comes from the philosopher Derek Parfitt. He pointed out that there's a big difference between 99% of people dying versus 100% of people dying. Because 99% of people dying, there's still a chance the species can continue and there will be future people who are morally relevant people. Whereas if everyone dies, then we've deprived humanity of its future.

30:41So extinction on this philosophy is uniquely bad and much, much worse than almost extinction. So effective altruism wants to focus now on threats to the long-term survival of human beings. And you might think that that means that climate change is a high priority. And it is a priority, but not a high one. And that's for two reasons. The first reason is just that there's already a lot of money going to climate change and climate change risk mitigation.

31:19And the other reason is because climate change might very well make the earth uninhabitable for like 90% of the population, but 10% of the population would still be able to inhabit it. But it also means that EA 1.0 global health causes are not as weighty because they involve fewer people than all of the future people that you're trying to save. I mean, my initial thought is that that sounds like a satire of effective altruism 1.0 to me,

31:51or a kind of taking of its axioms to an extreme case. The gut reaction I'm having, which is that future lives, and not just like the next generation, but 1,000 generations from now, should be weighed the same as present lives. I just revolt against that idea.

32:18You know, I mean, I really believe the person in the global south is more valuable than the person 1,000 years from now. Shaquille Hashim actually doesn't disagree with that. Which is why I asked him, you know, 3,500 bucks will save a life in the global south now. But he wants us to redirect it towards these existential risk causes. So how could he justify that when the beneficiaries might not even be born yet? I think it comes down to this fundamental thing of, yes, I do think a given amount of dollars can save a given amount of lives in public health.

32:56But I think we can probably save even more in existential risk reduction. I think my three and a half thousand dollars in AI reduction could quite plausibly save more than one life. And I think that's a bet worth taking. You know, the nice thing about effective altruism 1.0 is a normal person like me can understand what they're going to do. We're going to give somebody a net. We're going to give somebody vitamin A. We're going to give somebody a vaccine. We're going to do things that I believe in a deep way are going to help that person.

33:26Whereas I would guess the things they would name as preventing nuclear war or a pandemic are probably going to feel to me like they have at best a chance of preventing those things. That's right. That's exactly right. You've just lowered the risk by 0.01% by your giving of $5,000 to this nuclear monitoring organization, something like that. But I might feel good about doing that, I guess. I would need to think about it. Shaquille, what are the nonprofits in nuclear war, pandemic, and AI risk mitigation actually doing with the money that you're asking people to give to them?

34:08So a lot comes down to lobbying, meeting with governments, trying to explain the risks to them, trying to help them draw up policies. A lot is technological development. So in pandemics in particular, developing new vaccine platforms to try and create a system where when a new biological threat appears, we can have a vaccine within a few weeks. What about an AI risk? The things that we need are more technical research to figure out how to make these systems safe.

34:40We don't understand how these systems work, and we don't know how to get them to do what we want them to do. It's possible that there are mathematical and computer science techniques that will solve that problem. And so we need more research being done in that space. A lot of that research is happening at nonprofit organizations or in academia. The other is regulation. A lot of the money going to risk mitigation is going to PhDs in expensive countries so that they can have jobs doing their PhD research.

35:15And if you compare that to something like global health, I don't know. Well, the salary of a PhD doing AI risk mitigation is like $100,000 a year, and the amount of good you can do for a present person for that amount in the global health space just is enormous. And these were the kinds of calculations that effective altruists really cared about. So there better be like a real return on that. Is that the right way to think about this? I think so, yeah, and I think that's a really fair point to raise.

35:47So one thing I say to people is I think AI is at where climate change was in the 70s. Experts realize it's a problem, but the rest of the world hasn't cottoned on yet, and there's not much money going towards it. We hope that that one PhD researcher could end up having an impact that saves all of humanity. I guess the number of potential future lives is always going to be vastly larger than the present.

36:18That's exactly the argument, Curtis. Yeah, you can see it. You can see it. I mean, so it's making me realize I think on some level, I don't really think it's a bad thing if humans go extinct.

36:28You know, to me, I'm much more worried about the world where there's only 10% inhabitable. That sounds like a recipe for a ton of suffering to me. Whereas 100% annihilation of humanity, which will be an awful experience to undergo with tons of suffering. But I guess I see that as, isn't it just as plausible to say, okay, that's the end of human suffering? Okay, so here are Shaquille's recommendations for you.

Recommendations for Giving

36:59Curtis, I think you should give 60% to tackling existential risks, 30% to tackling global health, 5% to tackling animal welfare, and 5% to tackling climate change. In true effective altruist economic reasoning fashion, you should temper your giving with uncertainty. That is, these things matter, but you don't know how big of a difference they'll make, or you don't know how effective it'll be.

37:31So rather than thinking about giving to one cause, you should divide your causes into a portfolio, like we would do if you were investing in your own retirement. Yeah, and I do find that appealing relative to Effective Altruism 1.0, where everything should be devoted to this one cause until that's solved. Then we can figure out what the next optimal cause is. The idea of the portfolio generally, and the uncertainty of what is going to do the most good, I think is appealing to me.

38:04Hi-Fi Nation will return after these messages.

Critique of Effective Altruism 2.0

38:10Okay, so let me introduce you to the last advisor. Yeah, so I've got a microphone that I've just got out of the cupboard. This is Seth Lazar. Seth is a moral philosopher who is a very strong critic of effective altruism, the strongest critic that I've found. And Seth is actually involved in AI risk research. So let me tell you a little bit about what he says specifically about AI risk, and then generalize it to what he thinks is mistaken about effective altruism. The kind of thing that we have to worry about, says Seth, is not the robot apocalypse.

38:45It's actually a lot more familiar. Effectively, it's like creating the most kind of spectacular computer worm or virus that has ever been developed that can kind of call upon the skills of all the hackers that have gone into its training data. As you can imagine systems like this down the line, being used alongside a sort of human amanuensis, someone who has the kind of spirit of an Oklahoma bomber, you know, an incel who wants to get famous for destroying things, who then uses this to create a virus that is able to kind of adapt and autonomously undermine our ability to use digital technologies kind of across the board.

39:22And Seth thinks that this kind of AI threat is not something that he wants you to give any money towards mitigating. I think this is not really an area where philanthropic funding is an appropriate contributor. There's a lot of interest in it from government. There's a lot of interest in it from private companies. And these aren't even the existential risks that effective altruists want us to be worried about. They're concerned about AIs that can have certain goals that will, like, enslave people or annihilate them.

39:52The way that Seth thinks about the AI robot apocalypse is that could happen, but it would involve technology that we don't have today. So there's nothing you can do about it now, even if you had billions and billions and put all your resources into it, because you don't know what that technology is going to be doing. It would be sort of like trying to figure out now how to mitigate a threat from aliens who might land on Earth in 100 years' time. And, you know, like, given that we don't know what their technology is going to be, what can you actually do? Not a lot. Seth thinks that this makes the effective altruist turn to AI risk mitigation actually hypocritical.

40:28If you're asking people to put money into mitigating a risk, but you actually have no idea how that threat will happen, you're using money ineffectively. One of the biggest critiques of effective altruism in its early days was that it was obsessed with metrics. I think that in a way the EA people, like, they overdid it, but they were more or less right. It's important to be able to understand and actually measure the significance of your donations. It is unfathomable to me, this shift to investing in AI risk mitigation research.

41:02Because, Seth thinks, how much good the money is doing isn't measurable in the same way that it was under EA 1.0. Like, how do you measure how many future lives you're saving with a particular bit of AI research? Like, are you reducing the risk of extinction by 1% or 4%? Now, Seth thinks that this makes a lot of EA 2.0 causes stack up really poorly to the EA 1.0 causes.

41:40And there's another thing. Recent scandals in the effective altruism community. There's already been an expose about two organizations with an effective altruism that collapsed. First, Alameda Research and the Future Fund. Both were run by Sam Bankman-Fried, CEO of FTX, the crypto trading company. SBF was an effective altruist who stated that his aim was to take almost all of the money he was earning through FTX and donate it to effective altruist causes.

42:132.0 causes. To prevent human extinction and create an optimized human future. And he did donate. He funded a lot of effective altruist causes and gave away a lot of money. As of the recording of this episode, SBF is sitting under indictment and house arrest for defrauding investors out of billions of dollars in his crypto exchange company, some of which he allegedly siphoned to Alameda and the Future Fund.

42:45There's no question that what SBF did was bad. The question is whether there's something in the philosophy of effective altruism 2.0 that made him think he was justified. Because if that's true, the problem is with the movement, not the bad billionaire. All right, Curtis. So Seth thinks there is a connection between bad behavior connected to the EA movement and the philosophy of effective altruism. So he's a critique of effective altruism 2.0, really.

43:19That's right. He thinks that the effective altruists who go from EA 1.0 to 2.0, he thinks there's a mistake in that move. He calls it moral inflationism. You know, you talk about the kind of the scale of loss that could be possible from a genuine existential threat. And you then say, well, how could anything else possibly matter as much as that? There's nothing that's on the same scale to anything else that you might care about. So let's not worry about the fact that we really can't do an awful lot about it.

43:50Because gosh, the very small chance that you might be able to do something about it is worth taking. In essence, that's the way in which the sort of public facing versions of long termism, that's the way in which, as far as I can tell, they seem to argue. And I think it's understandable that people go that route. But I think it's very deeply morally mistaken. So you're like nuclear war, that's going to annihilate everything. AI risk is going to annihilate everything. What it clouds you into thinking is that it's worth anything. It's worth everything that you can possibly give.

44:22So here's how Seth thinks that makes EA generally a concern. When you think about morality in this way, there's something that is so bad that anything you do now is worth the price of trying to lower the risk a little bit. It gives license to commit any kind of present day crime in support of this altruistic act. Right. That makes sense to me. I wouldn't have thought of it without him pointing it out.

44:52That's, I guess, why I'm wondering, does this critique exist without a high profile example like Sam Brankman Freed? Like, I'm not sure the crypto billionaire shows the, like, moral emptiness of the whole movement. Yeah, that's right. That's right. That's fair. That's a fair thing to say. I mean, I asked Richard Yedder Chappell, I asked him whether the kind of utilitarian calculation that is part of this EA long-termism is what made people justify to themselves that they

45:23could lie and cheat. So I disagree with that very forcefully. I mean, I can't speculate about anyone else's psychology, but the thing I can say is that I don't think that utilitarianism truly justifies fraud in the service of effective altruism. For someone who is willing to just calculate expected utilities in a very explicit way and say, like, oh, yeah, whenever it seems to me like it would do more good for me to break these laws, that that would be worth it. And I think we should all be able to see in advance that that's not going to lead to more

45:54of the greater good in the long run. That kind of reasoning we know historically tends to do more harm than good. If you're wanting to do more good, the way to do it is to respect various kinds of common sense norms unless you've got really, really strong, robust evidence that you're in a weird situation where it really would do more good in the long run. I think clearly here, SBF's actions have not done more good than harm. They've done vastly more harm. So Richard Yedder-Chapel thinks it's bad utilitarianism, not good utilitarianism, that led to the SBF scandal.

46:24You can't ever justify acting dishonestly or harmfully in the interest of greater good because you're probably just going to fail. It's not actually going to be for the greater good. But Seth says that's disingenuous. If ultimately you are arguing to people that we have in our hands the ability to affect the possible creation or not of 10 to the 48 future human beings, you know, bearers of utility, then the notion that you can just sort of say to people, well, but, you know,

46:56it's also really important that you act with integrity. It's just bananas. There's just this intellectual inconsistency. The fact of altruism is attracting young people around the world in droves. So you want to have this big, urgent message. And it turns out that saving people from malaria wasn't a big enough and urgent enough message. And so you have this even bigger, more spectacular one. This bait and switch is just not tenable. You can't on the one hand say that these kind of 10 to the 48 future people, their existence matters and it depends on what we do, but then not say it's going to have a massive impact

47:30on what you're actually permitted to do. So I think that's why you're inevitably going to get things like these scandals.

47:39A lot of Seth's critiques so far presume that there's some kind of good faith moral reasoning on behalf of effective altruists that just kind of goes haywire. They're making some subtle mistakes. But Seth has an even harsher critique, not just of effective altruism 2.0, but also of 1.0, which leads to his recommendation to you, Curtis. If you want to help people, and if you want to help actual people, then you need to be

48:09sensitive to how those people actually want to be helped, right? None of us generally like, you know, receiving handouts or being told what's good for us or, you know, receiving one thing when actually we think that we'd rather have another thing. So the starting paradigm of giving what we can and the underlying effect of altruism was this notion that, you know, we, the EA people, are going to decide for others what's good for them, and we're then going to kind of do that. Over time, obviously, people realise that folks don't just like being helped without having a say about how they're helped.

48:40And so they do things like, you know, surveys and, you know, they've promoted this give directly approach, which involves just giving people resources, and they try and make it more conscientious. But another thing that you can do is try to help people who don't yet exist, because people who don't yet exist, they can't talk back and they can't say, well, actually, we'd rather you did this. You can claim the moral authority of, you know, the 10 to the 48 number of people that could exist if people listen to you, but you don't actually have to be accountable to the

49:13people that you're trying to help. Unfortunately, I think that that is a pretty strong reason why long termism has gained appeal. And, you know, it's why it has gained appeal among a cohort of people who like to have power. It's white men from a certain kind of background. Like your background. Yeah, like people like me, basically, you know, and like, the reality is that we who are morally motivated, often come up with situations where we try and do something that we think is good. And other people say, you know, actually, maybe you shouldn't be the ones who are deciding

49:45what's good here. And there's two things that you can do in response to that. One is say, actually, you're right and seed power, essentially. Or alternatively, you can kind of, you know, shift and try and find a problem framing that allows you to continue to have power. And I think the long termists have done the second thing. Personally, what I try to do is the first thing. OK, so Curtis, don't just think about how bad it would be if humanity went extinct.

50:15You should really reflect not just on, you know, the quality adjusted life years that you can bring into the world. One of the things that makes your options more choice worthy is that you're actually empowering people rather than deciding for people what they should do. Look at the options that are available to you and think, like, is it right that it's me deciding for these people how their lives go, whether they get, you know, a particular kind of treatment or not? Like, is it right for me that it should be me deciding, you know, whether resources are

50:46spent on bed nets or on something else that they might actually value more? How can I use my resources to empower people to deal with the problems that they face? That also recognize that my being in this position of power is itself something questionable and that what I should do if I have this kind of power is to try to seed it, try to lift people up and enable them to be more autonomous in their lives.

51:10So he likes micro lending, for example? That's better, right? But he actually likes give directly, right? So which isn't even lending? Oh, OK. Yeah, yeah. Just unrestricted grants. Yeah, I mean, I find autonomy very attractive as a goal for people, for humanity. But like Perlman, and to some extent, like effective altruism 2.0, it lacks the metrics, I guess, right? You have to kind of, you have to have a lot of trust. What you're doing will actually increase autonomy.

51:43Yes, absolutely. No metrics involved. And this give directly is actually a charity. Yep, that's right. You know, they're not taking anything off the top or they'll take very little, like 1% off the top. It's hard to say that giving cash is actually like intruding on people's autonomy. Cash is sort of the autonomous thing. But Seth doesn't recommend a particular thing. He's not going to like say, do, give directly. Just give to anything that increases autonomy. You've given me a lot to think about right now. Now I don't feel like I can decide, which I think is, in a sense, good, because it means

52:16I heard something persuasive on all counts. So yeah, could you call me later and I'll make my decision then? You got it.

Decision and Conclusion

52:36Curtis, have you made a decision? Yes, I've made a decision. It was harder than I thought it would be. I'm actually really feeling like I want to talk through a little bit about how I thought about it. Sure. Go right ahead. At the end, I felt that the one that was definitely most comfortable for me was Perlman. That's comfortable because it's basically what I was already doing. I'd say the one that was least attractive to me was Shaquille Hashim and the long-termism. And the one that was most exciting to me was Lazar.

53:08Like, Lazar was exciting to me because this idea of autonomy is exciting to me. And like, I personally value autonomy very, very highly. Now, what's your opinion on effective altruism 1.0? And I'm actually totally convinced by their argument that the charities they recommend are very good charities, probably the best charity in terms of value for money. And I was thinking, well, I'll probably start giving to those charities in my end of year

53:40giving. But I just really don't want to stop giving locally. And I think I understood that as part of the effective altruism argument is that you have to optimize it. That just feels wrong and bad to me. And I can't let go of this feeling of how good I feel when I give locally. I think about this organization I give money to. It's a group that provides opportunities to participate in theater for people incarcerated in prisons around here.

54:12And they have a dance troupe for people who have gotten out of prison and learned to dance in prison through their programs. And then they can tour around dancing. And when I think about what it was like to sit and watch these men who had spent their entire adult lives in prison dancing, I mean, I played a small role in that. But part of the reason they're able to do that is because I gave money. I'm like, it's only by a perverse definition of altruism that that's not altruism to support

54:45something like that. Is your conclusion that you're going to continue to give as you were? Or is there any change? Like suppose the pot of money is 100%. Can you break it down how you're deciding right now on the basis of what you've heard? With the money I pledged for this episode and for this experience of hearing these arguments, I'm going to give it to give directly.

55:16To support autonomy is something I think is very valuable and very interesting to me. I didn't know I could do that. So in that sense, Lazar opened up a new understanding to me. So the money I pledged for this, I'm going to give to give directly. What is your takeaway from this experience? I think my biggest takeaway is that for me, there's a component of charitable giving that the arguments didn't seem to take account of, which I would describe as sort of how it makes

55:49me feel to support these causes. So, you know, I think the effective altruists are saying, don't think about that. But the kind of exclusivity of their argument, in a way, like requires giving up so many things that I feel are good. It's true. Effective altruism is trying to take the emotion out of giving. Or at least what they think are unreasonable emotions.

56:22Caring, they say, done impartially, is the better kind of caring. I don't want today's episode to sound like a takedown of effective altruism, 1.0 or 2.0, just because its critics were able to get the last word. Effective altruism is a large movement, a global institution. Large institutions have internal disputes and disagreements. Long-termism is only one branch of effective altruism. The original branch still thrives today.

56:54And one charity that Effective Altruism 1.0 has coalesced around in recent years is giving directly. It turns out giving people autonomy through cash also happens to be highly effective in prolonging their lives and increasing their well-being. The movement has done a lot of good, not just in lives saved, but also in holding the rest of the philanthropic world to account for their effectiveness. The question is whether it has also done any bad, either accidentally or by way of its

57:31fundamental commitments. And the question is whether it can continue to convince the world that it is the future of philanthropy. And that depends on whether you're convinced. Hi-Fi Nation is produced, written, and edited by Barry Lamb. Story editor for this season is Eleanor Gordon-Smith. For Slate Podcasts, Alicia Montgomery is VP of Audio, Derek John is executive producer of Narrative Podcasts, and Ben Richmond is senior director of Operations.

58:01Additional production assistance for this episode provided by Alec Opperman. Follow Hi-Fi Nation on Facebook, Twitter, and Instagram at Hi-Fi Nation. That's H-I-P-H-I Nation. Complete transcripts, show notes, and reading suggestions for every episode is available at Hi-Fi Nation dot org.