What’s wrong with effective altruism? With Martin Sandbu

This is an audio transcript of the The Economics Show with Soumaya Keynes podcast episode: ‘What’s wrong with effective altruism? With Martin Sandbu’

Soumaya Keynes
The effective altruism movement has been on a wild ride. Now, when I first engaged with it in the early 2010s, it seemed to be about how to give money to charity more effectively, how to do the most good. And that meant more money for things like anti-malaria bed nets or deworming pills, things like that. But since then, effective altruism has evolved. It became about crazy and sometimes uncomfortable thought experiments. It became about the risk of AI wiping us out. It started shaping academic research agendas. It became the darling of Silicon Valley tech bros. Everything got a bit intense, and arguably, it went a bit wrong. Today, we’re going to talk about effective altruism. Does it make sense? 

[MUSIC PLAYING]

This is The Economics Show. I’m Soumaya Keynes in London, joined in the studio by Martin Sandbu, my colleague here at the Financial Times. Hi, Martin.

Martin Sandbu
Hi.

Soumaya Keynes
Are you an effective altruist?

Martin Sandbu
Absolutely not. I could hardly be further away from it, I think. 

Soumaya Keynes
Right. Well, lots of ambiguity there then. OK, Martin, I wanted to talk to you because last December, we both wrote about effective altruism. My piece, as I’m sure you will remember, was “An effective altruist’s guide to Christmas”. And the point was that lots of EAs — that’s what they call themselves — take themselves very seriously, and maybe they shouldn’t. And reading the comments of that piece, yes. Yes, they do. I did not persuade any of them to take themselves less seriously. But you wrote a much longer, more thoughtful essay on what was going on with this movement. And I wanted to ask why, now, why did you pick that topic then?

Martin Sandbu
So there are a couple of personal reasons for this. So, before I joined the FT about 15 years ago, I was an academic for . . . I did a PhD, you know, I did economics and philosophy for a long time. I did my undergraduate work at Oxford. I came out of the sort of environment that the effective altruism founders and the movement really came out of. I had never seen it coming, and I was very surprised by it. So that was one thing that had puzzled me for as long as I’d heard about effective altruism.

The other reason was that in the last year, effective altruism had popped up in very FT stories that you wouldn’t really have thought it had a place in. So in particular, the collapse of FTX, Sam Bankman-Fried’s crypto exchange. It turned out Sam Bankman-Fried was a self-avowed effective altruist. He said the motivation for making so much money was precisely to give money to effective altruism. That was one example. The other example was OpenAI and the ousting and then return of Sam Altman as the CEO of this biggest AI company, where some of the board members were core members of the effective altruism movements. They were, in the end, the ones who lost this boardroom battle and left. But it just struck me as extraordinary that this quite new, quite nerdy movement popped up in two of the biggest and most influential, consequential business stories of the decade.

Soumaya Keynes
Right. Yes, there has been a lot of focus on effective altruism or EA as they like to call themselves. OK, we should step back though and sort of define what this movement is, how it started. I think the one-sentence summary is doing the most good that you can, but there are obviously different flavours of that. Martin, could you talk a bit about how it started, the idea there?

Martin Sandbu
Yeah. So it started in two ways. Right? There’s kind of an intellectual history to talk about, and there’s just a kind of organisational history. So let’s just do the last bit first. Around 2009, a couple of young Oxford philosophers, Will MacAskill and Toby Ord, both came to this idea of doing the best, doing the most good you can. And they started setting up a kind of organisation around this. Actually, several organisations. There was something called Giving What We Can. There was something called the Centre for Effective Altruism.

The basic idea was, you know, we are morally obliged to relieve suffering in the world and to do that as well as we can. And it’s immoral not to give at least some of our money to the things that could most help relieve suffering. Because it turns out that for very . . . a very little, very small monetary cost to ourselves, you know, give up your morning coffee or whatever, you can pay for something like malarial bed nets, antimalarial bed nets, that will save lives. So Giving What We Can was an organisation where people make pledges to set aside, I think, 10 per cent of their income and spend it in very effective ways. The Centre for Effective Altruism was set up to study, do the research to find out, well, where do you get the most suffering relieved by per pound or dollar given? And a whole network of organisations like that came about. Very quickly, this sort of very well meaning but quite a little bit nerdy kind of set-up got a lot of followers, lots of smart young people in the best universities in the world, the Oxford, Cambridge, the Ivies, and so on, were taken with this, and soon enough, a lot of money started coming into it as well.

Soumaya Keynes
Yeah. I mean, so there’s another strand that you didn’t mention, which is the 80,000 Hours …

Martin Sandbu
That’s another one. Yeah.

Soumaya Keynes
 . . . bit of it, right? And so that is saying, look, you spend a lot of your career doing stuff. So you should make sure that that stuff that you’re doing is the best it can possibly be. And in some cases, that might mean joining a hedge fund, making a bazillion dollars, and giving all of that to charity. That’s OK. But, essentially, when making these financial decisions, you should use the effective altruism framework to make all of them. Right? And so, the framework was sort of utilitarianism with some tweaks. So, you know, when making decisions, you need to quantify the benefits of those decisions, make sure that that benefit is maximised relative to the cost. And it’s funny. If you go on internet forums, EA online forums, this framework can be applied to everything and they do. So it’s, you know, what food do you eat, right, when considering animal welfare costs? Or it’s, you know, should the movement buy a castle to house their conferences? You can kind of use this to reason your way to a lot of different things, and they apply this framework very, very broadly. But, Martin, you must have studied this, obviously, as an academic in the world of philosophy and economics. Could you give your definition of utilitarianism?

Martin Sandbu
It’s a 200-year-old political philosophy. The original formulation was: Do what gives the greatest happiness to the greatest number of people. You know, that captures it pretty well. It’s seeing good as a quantity and a quantity to be maximised, and that’s the basics of ethics. That’s what we ought to do as moral beings. So it’s a very powerful theory because it’s so applicable to everything. And back in the day, it was a very radical theory. It was obviously an anti-religious theory. It, you know, swept away all kinds of dogma or religious commandments. It just said, look at what makes people happy and look at what makes most people happy. And it could often be a tool for social reform and progressive reform historically, you know.

Let’s also remember that in mainstream academic philosophy, utilitarianism is basically a sideshow. If that is because most moral philosophers have long since been thinking that it gets too extreme, it produces implausible conclusions, and it’s basically not the right way of thinking about ethics in the first place. Utilitarianism is also actually at the root of modern economics. It started to see people as utility functions, essentially, as people who enjoy something called utility or happiness.

Soumaya Keynes
Right. And, of course, all of this is very familiar to me during my undergraduate economics days when, you know, the first thing you’re presented with in microeconomics is a utility function. Right? Economics is about maximising that utility function, at least in its basic form. So, you know, I can see why this framework could be very appealing to someone who wanted simple solutions to complex problems. It kind of reduces them to this fairly straightforward analytical framework.

Martin Sandbu
I think that the affinity between economics and utilitarian philosophy that you point out, Soumaya, is really important. Historically, they developed together, or rather economics came out of utilitarian political philosophy. They were kind of the same thing up until the late-19th century. And so the methods came out of precisely seeing humans as these containers of utility. And to this day, in the part of economics where one tries to assess outcomes of policies, cost-benefit analysis, which is the technique used in finance ministries around the world, it’s basically utilitarianism. You put a money value on people’s utilities, and then you add things up and see which policy is better. And in the other direction too, economics is extremely attractive to utilitarian philosophers because it gives all those techniques by which you calculate total utility or the greatest happiness for the greatest number. So this affinity between utilitarian, political, and moral philosophy, and economics, that’s been there all the way up.

Soumaya Keynes
Why do you think this became so popular? Right? I mean, there was this explosion in Silicon Valley tech bros saying, yes, this is what I want to use as a framework. Why?

Martin Sandbu
It’s a really good question, and I’m not totally confident of my answers. I tried to give some answers in that long piece in December. But the starting point is that I still find it really surprising. When I was doing philosophy and economics, utilitarianism was sort of, we basically all concluded with some exceptions that this wasn’t a very good theory. It did happen. So why did it happen?

I can think of a couple of reasons. One is that it’s a sort of thinking that appeals to a certain personality type: people who kind of want a very clear answer, who are morally committed, but they want to know, so what am I supposed to do? Well, utilitarianism allows you, it gives you an algorithm for choosing between your actions. You know, I could do this. I could do this. I could do this. OK. Maybe I shouldn’t be a social worker or a volunteer in Africa. If I want to do the most good, I actually should make a lot of money and then give money to those organisations. Right? That sort of thinking appeals to a kind of personality type. And I spoke to quite a few academics for this piece, and they tell me that it’s often the kind of techie types of students who are most attracted to this, kind of economics-ey type. So that’s one reason.

Another reason is that it happened at a very specific time in our recent history. Right? You just had the global financial crisis. Couple years before that, you’d had 9/11 and the Iraq war. There was a sort of moment of angst, I think, and the climate disaster was starting to impress itself on our consciousness. And I think people who were, let’s say, 17, 18 in 2009, you can see the attraction of a comparatively clear moral framework that gives you a direction to take in a world that looks like it’s going to pieces. So that’s another thing.

But the darkest possibility, I think, and several harsh critics in philosophy departments that I spoke to, critics of EA, they said, well, it’s a very convenient theory for the tech bros, for Silicon Valley, and so on, because it has actually become a sort of moral indulgence. It says it’s really OK to be super rich, to take advantage and be privileged of a very unequal economic structure so long as you use it for good. So in a sense, a lot of people see it, the critics see it as a very anti-progressive movement in the sense that it does not question at all the general social structure, just says what we should do within that social structure. In a sense, perhaps justifies or dispel some moral qualms people might otherwise have.

Soumaya Keynes
Yeah. I mean, that all sounds highly plausible. So this idea has captured the imaginations of these Silicon Valley tech bros who want this neat framework for thinking about the world. Can we talk now about how the ideas developed? Because from when I first engaged with this and EA was about bed nets and deworming pills, it seems to have strayed quite a bit from those earlier ideas. Right? Now, if you go on the 80,000 Hours top courses website that kind of lists things that you should do with your life, the top ones are working on risks from AI, catastrophic pandemics, nuclear war, a great power war, and climate change. You know, some of those are pretty mainstream. Then you’ve got a whole kind of community building set of courses. Right? So, it’s a good thing for your 80,000 hours to be spent building effective altruism. Sounds a bit convenient, but OK. And then you’ve got some slightly unusual things. So space governance, improving individual reasoning and cognition, like a clever pill. OK. And then at the bottom of the list, you have the things that I thought that this movement was all about at the beginning, which is, you know, factory farming or, you know, easily preventable and treatable illnesses. Not what I would have expected. How did we get there?

Martin Sandbu
It’s a really good question, isn’t it? And, you know, I scratch my head a bit too, and I was asking people about this. When I spoke to Toby Ord, one of the founders, he pointed out that — and others pointed this out, you know, we talked about pandemics early on — they did. You know, let’s give them credit for that. That was great foresight. Of course, it’s what the scientists were talking about too, and they took the science seriously. Unlike the politicians.

But then you have the seemingly zany stuff. It all goes under the rubric of existential risk. So it can be, you know, how do we avoid AI from breaking out and taking over the world? But it’s even seemingly zaniest stuff, craziest stuff, like if there’s a small chance that humanity can colonise the universe, there’ll be so many lives that can be made happy that kind of what happens now is so small in comparison that we should put in all our efforts to make sure that this happens, to conquer space …

Soumaya Keynes
Yeah. I mean, this is one frustration that I came across when reporting for my piece, which is that sometimes you can get an infinitesimally small probability of something bad happening. But as long as you multiply it by a large enough number of lives saved, it becomes magically worthwhile looking into. So, naught point naught naught naught naught naught 2 per cent probability of aliens invading, when put alongside the number of lives in the future that would be saved by stopping that alien invasion, suddenly alien invasion becomes a major thing that EAs should be working on. And, of course, criticism of that would be that some of these numbers are made up or sort of so uncertain that it’s hard to have a meaningful conversation about them.

Martin Sandbu
I think that’s right. I mean, it’s a fair criticism, but I think that’s it’s a fair retort to say, well, you know, we need to think about it. So, you know, what’s the best we can have? We need to think of some numbers, and let’s make them as sensible as we can.

Soumaya Keynes
Yeah. And, like, I suppose it is important to be precise when you’re talking about things. Right? That, I will grant, is an advantage of quantifying as much as you can.

Martin Sandbu
But here’s the problem. Suppose that the numbers are exactly right, there’s still something really disturbing about the conclusions. And here we come back to the ideas at the foundation of this because I think these sorts of conclusions, you asked how do we get to this. I think that was kind of built in to begin with because that’s how the whole structure of basically utilitarianism works. You just put it very beautifully. If the outcome is significant enough, then no matter how small the probability gets, it can still outweigh helping real people here and now. If there are enough future lives with enough of a probability, even if it’s very small, people here today that you could help right now matter less. That’s in the structure of the theory. You don’t kind of get away from that unless you kind of put in place some sort of check on how far you’re willing to take the theory.

So, you know, you go back and I was reading this stuff 10 years before EA came out to the people I think of as sort of godfathers, if you like, of EA. So there were utilitarians like the Australian philosopher Peter Singer, whose articles the founders of EA really cite as making a huge impact on them and helping them develop this. But other people like Derek Parfit and John Broome, both Oxford philosophers, a lot of their thought experiments really are about, well, what if there are all these possible future lives that we could do something for now. What if there are an identical you know, you can have twice as many lives in two identical universes. These sorts of thought experiments that are important, interesting, fantastic scholarly work, but it does rather lead you to these sorts of views that you then get the mega tech bros taking up for real or thinking about for real, and you get a real reallocation of resources. You know, let’s not exaggerate it. There’s still a lot of EA money that goes to things like antimalarial bed nets and other causes. But it does seem like the whole area of existential risk is taking up quite a lot more of the resources, but also the kind of headspace of the community than it did before.

Soumaya Keynes
Yeah. I mean, I want to bring in a term that we’ve both alluded to, which is longtermism. And so this is the idea that you should value future lives with equal weight as you do current lives. Right? And because there will be more future lives than there are lives today, that can take you to strange places. I think another sort of strange place that you can be taken to by this framework is when you start to kind of add up lives in your cost-benefit analysis, there sort of comes to this question of, well, should some lives be weighted more than others? Young versus old, babies versus grown adults . . . 

Martin Sandbu
Geniuses versus ordinary people . . . 

Soumaya Keynes
Yep. It gets pretty scary, pretty dark, pretty quickly. You know, there are pretty horrendous conversations about ableism, ableists, and how that interacts with this. There’s a whole strand of EA research into valuing animals’ lives differently in how you trade off the lives of, say, shrimp against the lives of cows. Sure, if people wanna spend their life doing that, OK. But the requirement that you have to add everything up takes you down these very strange and sometimes dark paths.

And I think just building on this, another critique of utilitarianism, effective altruism, is that some things just can’t be quantified. Right? What if antimalarial bed nets are not the most effective thing we can do with our money, they’re just the most effective thing that we know about because of these other benefits to other interventions that are difficult to put a number to?

Martin Sandbu
I think that’s a fair criticism. I suppose if you are a utilitarian or an effective altruist, your answer will be, well, but how can you know that these other things are better? At least here we know that we do some good. And, you know, here, let me let me come in defence of a modest version of effective altruism. You know, it’s really just about saying, you know, if you’re going to be altruist, let’s at least be smart about it. You know, pick causes that make a difference if that’s what you want to do, but don’t see it as a holistic theory of life. Now if that’s how you present it, it’s almost harmless. Right? It’s innocuous. Sure, sounds like a good idea. But then it’s hard to use that sort of very modest version to tell people that they should make life choices differently from how they feel because it could be more effective and so on. Then it doesn’t have as much grip. So if you do the modest version, it’s sort of a bit tame. That’s not what I think attracted so many followers. It’s the strong version that says, here is something that you can actually use to make sense of a lot of difficult questions in the world and in your life. It’s seductive in its clarity.

Soumaya Keynes
OK, well, I wanna go on to that seduction, but quickly before I do, I think it is really important to just note that, obviously, effective altruism is a very broad church. As you say, we have the weak version, the strong version, there are lots of different flavours within. 

OK. Let’s now talk about the scandal, the crisis that enveloped this movement when it became known that Sam Bankman-Fried, the head of FTX, this cryptocurrency exchange, and a major donor to the effective altruism movement, was in fact a fraud. Could you characterise the response within the EA community? What were people thinking?

Martin Sandbu
I think a lot of people took it as a real punch to the gut. I mean, they were sickened. I think, you know, we’ve talked about a lot of the criticisms of effective altruism as a theory and in practice, but there are a lot of people with integrity who take these criticisms seriously and try to engage with them. And I think a lot of these people, and I’ve spoken to some, were devastated by the revelations about Sam Bankman-Fried.

Soumaya Keynes
But could EA have actually contributed to the fraud? Could the ideas have been relevant?

Martin Sandbu
Look, it seems quite clear that Sam Bankman-Fried himself was a bona fide effective altruist. That’s what he said. Michael Lewis, for his book Going Infinite, chronicled the rise and fall of Sam Bankman-Fried, and there’s no doubt from his reporting and his view, he spent hours with Bankman-Fried, that Bankman-Fried and his entourage really believed in this stuff. They were committed to effective altruism. That’s what they were about. And some of this came out in the court testimonies as well that these were ideas they genuinely believed in. They had landed on this way of making quick big money, and they spent a lot of time and intellectual effort, including with some of the leaders of the EA movement, to think about where should they channel all these fantastic gains. So I think there’s you can’t really doubt that this was a big motivation in what was going on.

Soumaya Keynes
But I guess the question is whether the idea that you want to maximise the benefits, that idea that maybe a little fraud is OK if it makes you bazillions of dollars for these good causes, could that have contributed to Sam Beckman-Fried’s decision-making process?

Martin Sandbu
I think it could, and there are several reasons to think that this way of thinking could in fact have justified in his mind the fraud he was committing. I mean, one reason is the way he talked about making these gambles essentially. So there’s a famous quote from one of his, one person in his entourage that he had said, look, if he had the possibility to flip a coin and if it comes down heads, the world is destroyed. But if it comes down tails, everyone’s twice as well off, he would take the gamble. I mean, that’s kind of depending a bit on how you think about uncertainty. That’s what a utilitarian would say. Right? If the probability of a good enough outcome is high enough, you should do it. So that’s one sign.

But another sign I think is going back to this philosophical discussion we had at the beginning, it is kind of built into the theory. It gets the moral clarity from reducing everything to a single metric — maximised total utility — be as effective as you can. And it could just be a fact of the world that you can save the most lives, relieve the most suffering by defrauding some people who may not even realise they’re defrauded if Sam Bankman-Fried’s scheme had worked out, which could have, there might have been good reason to think that he would get away with it, and he would have saved tens of thousands of lives. Where in the theory do you find anything to stop you from doing that if that’s what you believe the facts to be? So I think the ideas really matter here.

Soumaya Keynes
OK. Well, I think we should throw to a break now. But when we come back, my question to Martin is: Are we done with EA? Is that it?

[MUSIC PLAYING]

[TECH TONIC TRAILER PLAYING]

Soumaya Keynes
We are back from the break. So, Martin, you’ve had a minute to think about that. Are we done?

Martin Sandbu
Look, I would guess that as a sort of social phenomenon, EA may have peaked. Who knows how it picks itself up from some of these scandals? And, you know, maybe there’ll be a new hot thing for the next cohort of smart young people to latch on to. But, of course, it’s done a lot of good. Let’s first give credit where credit is due. A lot of money has been channelled into great things. That’s one thing to just put on the table.

I was speaking for this piece, to John Broome, one of these senior philosophers I thought of as one of the godfathers of EA. He’s now retired. And he was pointing out that it’s just an amazing thing to see that there are so many altruists out there. That’s what takes people to effective altruism to begin with. It’s the altruism bit. Right? And then it’s maybe the effectiveness bit that creates quite a lot of these problems we’ve talked about. But it’s surely something to be celebrated that people feel altruistic and that some people have signed a pledge to give away a tenth of their income or maybe more sometimes permanently. I mean, those are noble, morally admirable things. But, of course, you can do that without effective altruism. What do you think?

Soumaya Keynes
Yeah. I think I’m a bit more positive about effective altruism than you are. I mean, my take has always been that if a movement relies on one man, then that’s a problem for the ideas.

And there certainly has been a lot of kind of embarrassment about the connection with Sam Bankman-Fried. But the evangelists I spoke to were, you know, defensive about the movement. Sam Bankman-Fried provided money, but it’s not like he was a kind of core philosopher. You know, perhaps at the margin, there’ll be some crypto bros who might have been led towards it before, who won’t be now. But the kind of, you know, nerdy, maybe studied a bit of econ in the past, searching for meaning and kind of how to make a difference in the world. That group of people, I sort of see as firm believers, I don’t think they’re gonna be put off by the scandals.

I think the really interesting question within the movement is how they respond to the reduction in cash that they have. Right? And is it possible that some of that cash was fuelling research into these more outlandish causes, and actually with a bit of retrenchment, it will move back towards those original kind of causes that attract lots of people. 

So maybe we’ll see a bit more work on, you know, reducing animal suffering and a bit less work on space exploration. Personally, I’m fine with that. Actually, after reporting my piece, I ended up donating a huge chunk of money to charity and thinking much more carefully about which charities I was gonna give to. So, you know, I’m happy with that choice and I feel like that kind of had a positive effect on me and hopefully some others.

Martin Sandbu
Soumaya, I’ll admit that I too, kind of working on this, felt the moral tug of the arguments and thought I should be giving more as well. And that’s obviously that’s right. It’s good. And maybe I would take it back to how it is. It’s a good thing that people really think hard about how to be good in the world, how to be ethical. Utilitarianism, I don’t think, is the right answer, but even just thinking about it, it does clarify thought. And I think it might have the effect of strengthening our moral resolve, you know, wherever we end up in terms of the actual direct content of what we should do. It may help us be a bit more confident in saying, yeah, we should sometimes do things because they’re the right thing to do. And that’s surely welcome.

Soumaya Keynes
Yeah. Absolutely. I guess the final thing is I really enjoyed that both of our pieces ended in a pretty similar place from, you know, taking different directions. My conclusion was: Guys, you really need to take a joke. And your conclusion, I think, was …

Martin Sandbu
It was: If you really want to do the most good you can do, maybe you shouldn’t be an effective altruist.

Soumaya Keynes
Well, no. It wasn’t quite that strong. It was, you know, we shouldn’t take effective altruism too seriously, which I think is pretty hard to argue with. Martin, thanks so much for joining me.

Martin Sandbu
Thanks for having me.

[MUSIC PLAYING]

Soumaya Keynes
That is all for this week. You have been listening to The Economics Show with Soumaya Keynes. If you enjoyed the show, then I would be eternally grateful if you could rate and reviews us wherever you listen. This episode was produced by Edith Rousselot, with original music from Breen Turner. Sound engineering by Joe Salcedo. It is edited by Bryant Urstadt. Our executive producer is Manuela Saragosa. Cheryl Brumley is the FT’s global head of audio. I’m Soumaya Keynes. Thanks for listening.