Enjoyed the episode? Want to listen later? Subscribe by searching 80,000 Hours wherever you get your podcasts, or click one of the buttons below:

Say you come along and are like, “Well, I did this estimate and I found that there’s this amazing global priority that no one else is working on and it’s like 100 times better than what everyone else is doing.” So then the question is, should you just trust that, or should you figure that you’ve probably made a mistake somewhere? And because your calculation has said there’s this thing that’s amazing compared to what everyone else is doing, most likely you’ve made an error in the direction of it being better than it actually is.

Ben Todd

Today’s episode is the latest conversation between Arden Koehler, and our CEO, Ben Todd.

Ben’s been thinking a lot about effective altruism recently, including what it really is, how it’s framed, and how people misunderstand it.

We recently released an article on misconceptions about effective altruism – based on Will MacAskill’s recent paper The Definition of Effective Altruism – and this episode can act as a companion piece.

Arden and Ben cover a bunch of topics related to effective altruism:

  • How it isn’t just about donating money to fight poverty
  • Whether it includes a moral obligation to give
  • The rigorous argument for its importance
  • Objections to that argument
  • How to talk about effective altruism for people who aren’t already familiar with it

Given that we’re in the same office, it’s relatively easy to record conversations between two 80k team members — so if you enjoy these types of bonus episodes, let us know at [email protected], and we might make them a more regular feature.

Get this episode by subscribing to our podcast on the world’s most pressing problems and how to solve them: type 80,000 Hours into your podcasting app. Or read the transcript below.

Producer: Keiran Harris.
Audio mastering: Ben Cordell.
Transcriptions: Zakee Ulhaq.

Key points

Misconceptions about effective altruism

An interesting recent example was Sam Harris when interviewed Toby Ord on his podcast, and when he defined effective altruism, he defined it as, “Taking the actions that most help the people who are most in need alive today.” And so his definition explicitly ruled out helping future generations and then for that reason, maybe ruling out working on existential risks, even though Toby Ord is on his podcast, who’s one of the creators of effective altruism and has also just written a book about existential risks.

I think Sam Harris is quite a representative example and it’s not so much that people within the community misunderstand it, but even very adjacent people often misunderstand it, and of which Sam Harris is an example, but there’s many others. And yeah, as to where they get the idea, well I’m not sure to some degree, but a couple of thoughts. One is that effective altruism is a very abstract idea. So the way I see it is, I see effective altruism as trying to seek the very best ways to have a positive impact or to contribute to the common good is how I like to frame it. And a lot of people it’s like… Yeah, that’s quite hard to grasp in abstract what exactly that’s saying.

So instead people end up focusing on particular things that people discuss in effective altruism. Particular ways of doing good, I mean. And, in particular, the ones that are most grabbing and memorable. And so much of effective altruism, in the past, it really focused a lot on donating to global health charities and how it’s surprisingly easy to save a life and those kinds of things. And if I had to sum up the misunderstandings with effective altruism in one line, it’s that people think that effective altruism is just about the claim that we should donate money to evidence-backed interventions that help the world’s poorest people.

Ben’s definition of effective altruism

So, there’s maybe two different questions. One is, what are the best ways to explain it in practice? Which is kind of more of a marketing question. And then the other question is just, what actually is it? If we just want to be really clear about the claims. For what actually it is, just a very high level definition I would use is effective altruism is about seeking the best ways to contribute to the common good. And so then you can roughly divide that into two projects.

First is the intellectual project, which is trying to figure out what are the best ways of contributing to the common good. And so that’s more like a field of research either in academia or in nonprofits where people are trying to have this debate.

And then the second part is the practical project which is like, given our current answers about what are the best ways of helping others who want to contribute to the common good, let’s see if we can actually put those into practice and actually tackle some of the world’s most pressing problems. And that’s more than the social movement around effective altruism of people who sometimes work together who are trying to tackle these big issues.

Objections to effective altruism

I think one is epistemic humility. This is the idea that, what should we believe about a difficult question like how much of a pressing priority is AI safety? One answer to that is to just try and figure it out for yourself. Another answer to that is just go with what other sensible people think and take an average across them. And some evidence that that kind of approach works is like in forecasting. So people have tried to measure, what’s the best way to predict geopolitical events, which are super hard to predict. And it turns out that getting like 100 people who are really good at forecasting, seeing what they think will happen, and then averaging that together is a really good way of doing it. And it’s usually better than any individual person.

And you can think of it as like, if someone else has a radically different view of what the biggest priorities are from you, and you’re in a similar position epistemically, why would you think you’re right and they’re wrong? Unless you have some really clear explanation of like, they’re making this mistake and you haven’t made some other mistakes that they’ve figured out and you haven’t figured out. So there’s a kind of symmetry argument for going with what other people think.

And yeah, it seems like if you put a lot of weight on that kind of argument, then you should be focusing on the priorities that a sensible coalition of other people would think are the biggest priorities. And that almost brings you back to the most common sense priorities by definition.

How to talk about effective altruism

I think most of our existing introductions tend to emphasize the community bit more than the intellectual bit, but actually both are really important and the nice thing about emphasizing the intellectual bit is it gets kind of more to the heart of effective altruism.

And so an example of the kind of community approach is, “Okay, effective altruism is a group of people who give 10% of their money to charities such as those supported by GiveWell”. And that’s, again, a very practical, ‘here’s what we do’ kind of explanation. Whereas the more intellectual project explanation would be, effective altruism is about the question of which ways to benefit others are most effective? And effective altruists like to do research into these questions and they try and tackle questions like, which global problems are most pressing? And what even is the common good? And what implications might different definitions have? And which interventions are most effective? And should we help people now? Or try and invest and help people later?

And so I like introducing effective altruism as a question and then giving examples of the types of questions that we think are really interesting and important. And then you can see if someone’s interested in one of those questions, then you can go more into, “Okay, well what do people say”? Some people say, “No, we should focus on the most evidence-based thing, because at least we can measure it.” Other people are like, “No, we should take a hits based approach so that we get the tail, the thing that’s most effective.” And that’s a debate within the community. And I think that type of approach is really good to appealing to someone who’s really interested in intellectual debate.

Transcript

Rob’s intro [00:00:00]

Robert Wiblin: Hi listeners, this is the 80,000 Hours Podcast, where each week we have an unusually in-depth conversation about one of the world’s most pressing problems and how you can use your career to solve it. I’m Rob Wiblin, Director of Research at 80,000 Hours.

Today I’m delighted to bring you another conversation between Arden Koehler, and our CEO, Ben Todd.

Ben’s been thinking a lot about effective altruism recently, including what it really is, and how it’s framed.

Effective altruism is widely misunderstood, sometimes even by its supporters, and this episode is our latest attempt to try to clarify the views of this community.

We also recently released an article on misconceptions about effective altruism, based on some work by Will MacAskill, which you can find in the show notes.

This is another ‘80k team chat’, which means it’s all very off-the-cut compared to our regular episodes, and even though he’s our CEO the things Ben says here aren’t ‘official 80,000 Hours’ positions or anything like that.

Arden and Ben first tackle common misconceptions about effective altruism, like how it isn’t just about donating money to fight poverty, and how it doesn’t necessarily include a moral obligation to give.

They then move on to clarify how Ben defines the central claim of effective altruism, and how to a potentially robust argument for effective altruism based on 3 premises.

They then discuss how you might object to the core ideas of EA by denying one of these premises. Though don’t cover the many (potentially good) objections to EA in practice, that will have to wait for another.

Finally, they cover a few ideas on how to communicate EA given all of the above.

Given that we’re in the same office, it’s relatively easy to record conversations between two 80k team members — so if you enjoy these types of bonus episodes, let us know at [email protected], and we might make them a more regular feature.

Alright, without further ado, I bring you Arden and Ben.

Misconceptions about effective altruism [00:02:12]

Arden Koehler: Hi listeners, I’m Arden. I’m a researcher at 80,000 Hours.

Ben Todd: Hi, and I’m Ben, CEO of 80,000 Hours.

Arden Koehler: And today we are going to talk about effective altruism, how it’s framed, what it is and how people misunderstand it. So, I know you’ve been thinking a bunch about effective altruism recently. 80,000 Hours is an organization in the effective altruism community. So why has this been on your mind?

Ben Todd: Yeah. So effective altruism has been around since about 2012 and yeah, I think something that just keeps bugging me is it seems like we haven’t done a great job of explaining what effective altruism is to the wider world.

Arden Koehler: Well now’s your chance, Ben. What would you say effective altruism is and how have we been going wrong?

Ben Todd: Yeah. So I’m excited to talk about this. I mean, an interesting recent example was Sam Harris interviewed Toby Ord on his podcast and when he defined effective altruism, he defined it as, “Taking the actions that most help the people who are most in need alive today.” And so his definition explicitly ruled out helping future generations and then for that reason, maybe ruling out working on existential risks, even though Toby Ord is on his podcast, who’s one of the creators of effective altruism and has also just written a book about existential risks.

Arden Koehler: So where did he get this idea, do you think?

Ben Todd: Yeah. So, I think… Maybe just also say, I think Sam Harris is quite a representative example and it’s not so much that people within the community misunderstand it, but even very adjacent people often misunderstand it, and of which Sam Harris is an example, but there’s many others. And yeah, as to where they get the idea, well I’m not sure to some degree, but a couple of thoughts. One is that effective altruism is a very abstract idea. So the way I see it is, I see effective altruism as trying to seek the very best ways to have a positive impact or to contribute to the common good is how I like to frame it. And a lot of people it’s like… Yeah, that’s quite hard to grasp in abstract what exactly that’s saying.

Ben Todd: So instead people end up focusing on particular things that people discuss in effective altruism. Yeah. Particular ways of doing good, I mean. And, in particular, the ones that are most grabbing and memorable. And so much of effective altruism, in the past, it really focused a lot on donating to global health charities and how that can save a life for… This is how it’s surprisingly easy to save a life and those kinds of things. And if I had to sum up the misunderstandings with effective altruism in one line, it’s just that people think that effective altruism is just about the claim that we should donate money to evidence-backed interventions that help the world’s poorest people.

Arden Koehler: Yeah. I wonder if this is partly because so many social movements are about something more specific, it’s kind of weird to have a social movement that’s about something as abstract as “Try to figure out and then do whatever it is that is actually doing the most good in the world”.

Ben Todd: Yeah. I think that’s one reason. Yeah, just before we start going more into where this came from, I want to say where this is most pronounced is if you look at external coverage of effective altruism, so basically people who aren’t in the community writing about the community. So there’s all these critiques of effective altruism, like on the Boston Review of Books and some academic papers about it and they pretty much all focus on why it’s not a good idea to focus on evidence-backed global health charities as a way of doing good. And so they’re all replying to that specific way of doing good than the broad concept. The actual underlying concept. If you want more details on this, Will MacAskill wrote this paper “The Definition of Effective Altruism”, which has a section on common misconceptions, which we just released an extract of onto the blog, and on there it lists lots of other coverage of effective altruism and how it’s actually responding to the misconception rather than the core idea.

Effective altruism isn’t just about donating money to fight poverty [00:05:56]

Arden Koehler: I guess it seems like there’s at least two differences between this definition of effective altruism that Sam Harris offered and that people seem to work with and the one that you gave, and maybe more, but it seems like two that jump out are one, it’s not just about donating money. At 80,000 Hours we talk a lot about doing good with your career, or really any other resource that you have at your disposal seems like it can be included.

Ben Todd: Yeah it could potentially be advocacy or political campaigns as well as your job and as well as your money.

Arden Koehler: Right. And then the other difference which you were pointing to more before is… Or when you were talking about existential risk, is that it seems like effective altruists understand themselves as wanting to work on a lot of different kinds of issue and not just global health. Are there–

Ben Todd: Yeah, exactly. And when we look at where money is actually being spent in the community, now it’s maybe something like 30% on global health, but actually therefore the majority is on other issues, which is a lot to do with existential risks. And then there’s also global priorities research, community building and then reducing factory farming. And also, actually, a lot of other issues like criminal justice reform and scientific research.

Arden Koehler: Yeah. And actually, just pausing on global priorities research for a second, that’s… So, research into what in fact is the best way to do good and I guess that’s another thing that’s sidelined in Sam Harris’ definition and sorry to Sam, I feel like now we’re using him as a shorthand, but–

Ben Todd: Yeah no, and I mean, his definition was actually still pretty accurate, it was just different about who the beneficiaries should be.

Arden Koehler: Ah. Okay. Yeah.

Ben Todd: Whereas often people confuse not only whether it’s about global poverty, which is the thing we’ve just been talking about, but also whether it should be just donating or helping people in general. And then, the third aspect, where the misunderstandings are, is whether it’s just about evidence-backed interventions, especially those tested with randomized control trials, like malaria nets or whether there could be some broader notion of what to focus on. And actually again, on that, now if we look at where most resources are going, I think more resources are going on what Open Philanthropy calls a ‘hits-based’ approach to doing good rather than an evidence-backed one.

Arden Koehler: So what does that mean? What’s that distinction?

Ben Todd: Yeah. So in the evidence-backed one, basically what you do is you try and test lots of things with randomized control trials, and then do the things that seem best based on those tests. Whereas a hits-based one is a bit more like investing in startups. The idea is most of the value, most of the impact will come from a small number of things that are really unusually good. And your goal as a person aiming to have an impact is to identify those amazing opportunities through whatever means are best and focus on those. And yeah, it would often end up focusing on high risk/high reward things which maybe don’t have very much evidence behind them, but where you have some qualitative arguments or a back of the envelope calculation, or you’ve looked at something that’s very important, neglected and tractable, that seems good based on that rather than things that are very measurable.

Arden Koehler: So, just to clarify with high risk/high reward, I think sometimes when people hear that phrase they imagine something that could cause a lot of harm or a lot of benefit, but I actually feel like many people in the effective altruism community are very anti-taking risks of that kind, things that could cause a lot of harm.

Ben Todd: Yeah, no, sorry. Yeah. I wasn’t trying to point to that distinction. I was just meaning things where maybe it could be amazing, but it’s maybe only a 10% chance of it being amazing.

Arden Koehler: Where the other 90% is like it’s neutral or maybe a little bit good or something, but not incredible.

Ben Todd: Yes.

Arden Koehler: So I guess if we’re turning briefly to the question of where people could have gotten this idea, one thing that strikes me is that it seems like for each of these elements, it is very related to something that is very important to effective altruism. So the evidence-backed thing, well, it seems like people are moving toward a hits-based approach sometimes, but the emphasis on evidence-backed interventions does come from the place of being like, “No, actually do the thing that makes the actual difference as opposed to something that seems like a good idea from the armchair, but you haven’t really thought very much about it” or something like that. Does that seem right?

Ben Todd: Yes. Though I would say the things that are most interesting and important about effective altruism, at least from my perspective, are these insights like longtermism, maybe existential risk is an important thing, like movement building, global priorities research, the now versus later debate–

Arden Koehler: What’s that?

Ben Todd: Just how should you trade… Like, if I can invest resources now and have more in the future, when is that better than just trying to help right away, trying to have an impact right away? I think these kinds of big picture questions are some of the things where effective altruism has really added the most compared to what’s already out there and have really big impacts on what you should focus on and all those kinds of things. It’s more about, I would say using philosophical reasoning rather than using evidence. And then, on the other side, evidence-backed charity or evidence-backed development is already a thing. That’s the randomistas. And so that’s not actually a novel contribution of effective altruism. And so by saying “using evidence” it’s getting people to think of the stuff that’s maybe not the most interesting and also not the most novel compared to what’s already out there.

Arden Koehler: So I guess just to push back a little bit, I mean, we’re talking about what is effective altruism? In some sense it’s like, who are we to make up the definitions or whatever? But, I guess, rephrasing it in terms of what really feels like the most distinctive and valuable thing to this community, I might be tempted to say that it’s really distinctively valuable that people in the effective altruism community say there’s a relentless focus on whether this is in fact the best use of your resources and that push to constantly be asking yourself, “Okay, but is this investment actually better than, for instance, saving a life, in expectation, by donating to the Against Malaria Foundation which is pretty good at saving lives”, is maybe that’s the core contribution and even though there’s this evidence-backed…

Arden Koehler: Or I guess, the relationship to evidence-backed is that maybe if you’re really focused on that, then you’re really obsessed with figuring out whether the thing you’re doing is in fact going to have the effect you think it will, in which case you search really hard for evidence. And if you’re not obsessed with that, then you maybe don’t search so hard for evidence and that’s the way it’s connected to…

Ben Todd: Yeah. So I might frame it as people interested in effective altruism want to seek out the best evidence that is available, but then that will often be a philosophical argument or lots of experts think this is a good idea or something that would not, in common sense, normal language, that wouldn’t be thought of as… Like normally when people talk about, “What’s the evidence for this?” They’re referring to some kind of more objective thing and so–

Arden Koehler: Or more empirical.

Ben Todd: Yeah, more empirical. Yeah, exactly. So it’s like, there is obviously a sense in which yes, of course we’re trying to use whatever evidence is available and have an evidence-based decision making, but we’re thinking of evidence in this broader way than what is most commonly used. And so really foregrounding being evidence-backed is like giving people a bit of a red herring.

Arden Koehler: Okay. I wonder if a better term could be something like “best justified” or something, where it’s like, it could be by empirical evidence, but it could be by a philosophical argument, but you’re really seeking that justification.

Ben Todd: Yeah. I mean, the standard phrase is using evidence and careful reasoning and I think … Yeah, careful reasoning’s also a pretty reasonable phrase to sum things up with. But yeah, the problem is that the evidence-backed thing has been so much more foregrounded in the past that people just only focus on that element.

Arden Koehler: And then, on the idea of oftentimes people equate effective altruism with trying to do these global health interventions, I guess, the kernel of truth in that is that it is one of the big insights of effective altruism, and maybe it sounds a little bit basic, but is that your money can go way farther in some areas than others. And one of the first ways that people made that concrete, was thinking, oftentimes you can do a lot more good in a poorer country or for people who are less privileged than for doing something in the US or the UK.

Ben Todd: Yeah. And that’s a really powerful insight; you can actually do way more good than you might think. And yeah, I mean… And also I mentioned, maybe something like a third or 40% of the money that’s being spent in an effective altruism style is being spent on global health. Among people who are changing their careers in an effective altruism way I would say it’s probably more like 10%. So it’s a bit lower than the money, which is because GiveWell is such a big factor, but it is still a major thing that people are working on. So yeah, there is a huge kernel of truth in this. Yeah. I guess… Well, I wanted to talk a bit more about how the misconception… How that idea got going.

Arden Koehler: Yeah, okay. Yeah. Please do.

Ben Todd: So one big thing to bear in mind is yeah, most people who’ve engaged with effective altruism haven’t read one of the primary sources like effectivealtruism.org, or even Doing Good Better, but rather they’ve read some media coverage about effective altruism because that media coverage has millions of views, so it’s many more views than have viewed the original sources. And all that media coverage is based on basically the media campaigns that were done between 2011 and roughly 2016 around Giving What We Can, GiveWell, Doing Good Better and also, Peter Singer giving his TED Talk and writing his book on effective altruism. And all four of those sources really focus a lot on the drowning pond, giving–

Arden Koehler: The drowning child.

Ben Todd: Sorry, the…

Arden Koehler: This is the Peter Singer thought experiment. You can listen to our podcast with Singer, I mean, if you haven’t heard of it already.

Ben Todd: Yes. Well actually, it’s mainly Peter Singer who does the drowning child example, but then all the others, they’re talking a lot about evidence-backed global poverty charities. So there were these media campaigns and there’s loads of people who wrote about that and then when you stumble across effective altruism online, you mainly stumble across one of those media articles about those things. So that’s another big reason why that’s happened. Maybe also just another point is, Peter Singer is still much more famous than effective altruism. If you look at how much search traffic both terms get, I think Peter Singer is about 10 times the search traffic (actually closer to 4 times the search traffic – Ed.) and Peter Singer focuses much more on global health, factory farming and climate change and focuses less on hits-based giving, global catastrophic risks, some of the other things that we talk about on the podcast.

Effective altruism doesn’t include a moral obligation to give [00:16:08]

Arden Koehler: So another thing that I think talking about Peter Singer makes salient, is it seems like oftentimes effective altruism is also identified with or associated with in some way a certain demanding moral theory. So it’s not so much about what you should do to do good… Or well, okay, I think everyone knows it is about that, but oftentimes people think there’s a second part, which is that also you should be spending a larger amount of your resources to make a difference. Yeah. What do you think of that aspect of the public image of effective altruism?

Ben Todd: Yeah. I mean, I think that’s another misconception in a way. And so, Will wrote this philosophy paper called “The Definition of Effective Altruism”, and in that he argues that we shouldn’t define it as having a moral obligation component. And one of his arguments for that is just, he polled lots of people in the community about whether they think it should be defined that way and, on average, people agreed it shouldn’t. So if we wanted to define effective altruism in the way that the people who actually practice it think it should be defined, which seems like a pretty reasonable way of doing it, they would all say, “No, it’s not about the moral obligation to give, rather it’s about supposing you do want to give some time or money or some other resource to help others, how can you do that most effectively?” And it’s… Well, yeah, it’s encouraging you to really focus on finding the very best ways to use those resources.

Arden Koehler: You didn’t use normative language in the definition that you just gave or moral language, but I might’ve thought that effective altruism does have a sort of moral claim, even if it’s limited. Yeah. I don’t think I was around to take this poll, but I might’ve thought that it was something like, “Okay, with the resources that you are going to dedicate to helping others, you actually should dedicate them to the ways that you think are going to do the most good and not just”… So, there is a should in there, but it’s just limited to the resources you are in fact already–

Ben Todd: Planning to contribute to the common good.

Arden Koehler: Yeah.

Ben Todd: Yes. No, I’m more okay with that being in there. I think later we can get onto how I would actually want to define the claim of effective altruism. But yeah, the key thing is that a lot of people go around and they’re like, “Well, I can never be an effective altruist because I don’t want to give all my money to charity, and they’re people who are saying that we have to give all our money away,” but that’s not what we’re saying. We’re more saying, “Supposing you want to give some away, let’s figure out the best ways to do it.

Ben Todd: And again, the reason for that is just that’s how people think we should define it. I think another reason is that’s a much more interesting insight because there’s already been thousands of years of moral philosophy about the question of what are our obligations to others? And we’re not going to be able to go and settle that debate, but I think we do have something much more unusual to add about the question of… Well, just making this point, that some ways of helping others is way better than others.

Arden Koehler: So, okay. Just to push back a little on this, and I know that I’m going up against the crowd here and that maybe means that since we’re just talking about how to define a social movement that consists in its members, maybe that doesn’t make sense, but so here’s one way that the insights of effective altruism seem related to this question about moral obligations. So one of the things that the effective altruism movement has emphasized is that some ways of helping others don’t just do a little bit more good, but they actually do a lot more good, right?

Arden Koehler: And in fact, there are these amazing opportunities to help others that maybe people weren’t really aware of or people are more skeptical that you can actually make other people better off. And so it seems like in the situation where some ways of helping others do that a lot more, it seems kind of commonsensical that you might have a stronger obligation to then use your resources to help others, since they can do so much rather than how it would be in a world where you can’t really help others that much with your resources. So that’s one way in which it seems connected.

Ben Todd: Yeah. No, there’s definitely a close connection and many people, in fact, after learning about effective altruism have decided to focus more on helping others because yeah, if one of your goals becomes easier than you thought, then it makes sense to focus more on that. But yeah, I just don’t want to have that as part of the defining or maybe most salient ideas of effective altruism. But yeah, if someone wants to draw that implication, then it makes a lot of sense to me.

Arden Koehler: Yeah. I mean, speaking personally, I feel like I was inspired when I started learning about effective altruism to be like, “Okay, actually, there’s so many amazing things we can do to help others, this makes me feel like I should be devoting more of my time and my resources”. And yeah, it does seem like a lot of people have that experience, but okay, so it’s a further implication that people are free to draw.

Ben Todd: Yeah. And I mean, I also would say the debate about how much to focus on helping others is just a really important and fascinating debate and I would encourage anyone interested to spend some time thinking about it and what it means in their own life. But again, that’s this much more general thing that has been around for centuries as a debate, and I wouldn’t say is the core thing in effective altruism.

Ben’s definition of effective altruism [00:21:02]

Arden Koehler: Okay. So you promised to give us your definition a little bit earlier. I feel like we’ve gotten bits and pieces. But do you want to just lay out how you would… If somebody said, “What is effective altruism, Ben?” What would you say?

Ben Todd: Well, yeah. So, there’s maybe two different questions. One is, what are the best ways to explain it in practice? Which is kind of more of a marketing question. And then the other question is just, what actually is it? If we just want to be really clear about the claims. For what actually it is, just a very high level definition I would use is effective altruism is about seeking the best ways to contribute to the common good. And so then you can roughly divide that into two projects. First is the intellectual project, which is trying to figure out what are the best ways of contributing to the common good. And so that’s more like a field of research either in academia or in nonprofits where people are trying to have this debate.

Ben Todd: And then the second part is the practical project which is like, given our current answers about what are the best ways of helping others who want to contribute to the common good, let’s see if we can actually put those into practice and actually tackle some of the world’s most pressing problems. And that’s more than the social movement around effective altruism of people who sometimes working together are trying to tackle these big issues.

Arden Koehler: Would you agree with the following statement that the intellectual project is… In some sense, the reason that it’s part of effective altruism is because people think in fact sometimes doing a bunch of research is a way of contributing a lot to the common good because it helps people do better actions later?

Ben Todd: Yeah. So, yeah, in a sense, we can see them all as just part of one thing. But I think in practice, it’s useful to split them because being an academic research field is very different from being a movement of people actually trying to do projects in the world, so they have a different qualitative character.

Arden Koehler: Okay. Not to belabor the definition point too much–

Ben Todd: We did both study philosophy, so it’s kind of–

Arden Koehler: Yeah. So forgive us listeners. You keep using this phrase, the common good, what do you mean by that?

Ben Todd: Yeah. So to some extent, I actually see the question of what is the common good as part of the study of effective altruism and part of the intellectual project of effective altruism along with other things like which problems are most pressing and so on. Well, yeah, so in Will MacAskill’s paper where he introduces the definition of effective altruism, he suggests that we should think of the common good as about what helps people from an impartial perspective and is what he calls tentatively welfarist.

Arden Koehler: Okay. So explain both those things, impartial perspective and tentatively welfarist.

Ben Todd: Yeah. So the impartial bit is kind of like the common bit. So impartial, very roughly, is just like we’re going to treat everyone’s interest as equal. And then the welfarist bit is then a bit about what does it mean to benefit someone’s interests? And there it’s saying, “Well, it’s about how much welfare people have” or sometimes also called wellbeing, which is about how good or bad their lives are, which could be understood as just happiness versus suffering, in a very narrow way. Or you could understand it in a broader way, like how flourishing are they? How fulfilled are they? But yeah, it’s something about your life being good or bad.

Arden Koehler: So this sounds extremely commonsensical, and of course we should not be partial or not privilege some interests over others and we should aim to make people happy and well off. So what are the salient alternatives to wanting to contribute to the common good, defined this way?

Ben Todd: Well most people don’t only want to contribute to the common good with their lives because they also want to have a personally fulfilling life and they also want to benefit their friends and family. They care more about their friends and family than strangers. And like we were saying, I don’t really want to get into the debate about how much you should focus on your personal goals versus the common good. I’m just starting from the perspective that the common good is one thing that matters and yeah, as to why that is, it’s just, well, I mean, most people believe if something you do will harm a stranger or you could easily benefit a stranger with little cost to yourself, then that would be a good thing to do. And that shows that we care about people in general, at least to some extent.

Arden Koehler: Yeah. I think the thing I was asking after was more something in between your personal goals and the common good. So lots of people aim to do good with their careers or with their donations, but they don’t say, “I’m aiming explicitly at the common good”. So what’s the difference between what they’re doing and what effective altruists are aiming at?

Ben Todd: I mean, it would depend a lot on the person and it’s often a bit hard to tell. Like some people, when they come to doing good, they think they should really focus on helping in their local community, which could be done because they care about impartiality. So then maybe they are impartial, but they just think that in practice, they can help the most in their local community because they know it the best. So maybe they are being impartial, but it’s just the thing that, in practice, works. But instead, maybe they actually think in some moral sense, they have greater obligations to their community and it’s actually morally better for them to help their local community. And it’s often hard to tell, but yeah, I’m not sure if that is the most common difference.

Ben Todd: Maybe one with the welfarism thing, if you’re focused on welfare, that might be welfare of sentient beings in general, not just humans. But it would mean that non-sentient beings don’t intrinsically have moral value. So yeah, the natural world has lots of value because it enables all these sentient beings to have good lives, but it doesn’t have value in and of itself. Whereas some environmentalists would say, “No, even if there were no animals and insects and any humans and anything that might be sentient, it would still be very valuable to have a pristine natural world.”

Arden Koehler: Okay. Yeah. So tentatively, effective altruists think… Or at least, the project is not about advancing those other values?

Ben Todd: Yes. Though yeah, the tentatively bit is to show that it’s not entirely certain, but this is our best guess of what to focus on. And yeah, the reasons for that is loads of people agree that, all else equal, if people have suffering or good lives, there’s an important difference between those two things. And we think, in practice empirically, there are big differences in how much wealth our different actions produce, so it’s a really important thing to focus on, even if it’s maybe not the final word on what matters.

Arden Koehler: Yeah. So speaking as somebody who believes that things matter besides welfare, but is also a member of the effective altruism community, it’s like, “Well, look, I think that it matters for people to have true beliefs or something”… Or I’m not sure about this, but that’s my sense, that would be my best guess that there are these other things that matter, but people not suffering and being happy seems extremely important and something that you’re probably not going to make a mistake by devoting a lot of resources to that.

Ben Todd: Yeah. Anyone who’s listening who wants more rigorous treatments of these … These are also just standard terms in moral philosophy, and so you can go and read about them on the Stanford Encyclopedia of Philosophy.

Arden Koehler: Yeah. And I think we have some links on our key ideas page where we summarize this a bit and link off to other readings. So, all right, is there anything else to say on just what is EA or what is effective altruism and how does it differ from the public perception?

Ben Todd: Well yeah, just quickly on the definition, my definition didn’t have “Using evidence and reason” actually as part of the fundamental definition. I’m just saying we should seek the best ways of helping others through whatever means are best to find those things. And obviously, I’m pretty keen on using evidence and reason, but I wouldn’t foreground it.

Arden Koehler: If it turns out that we should consult a crystal ball in order to find out if that’s the best way, then we should do that?

Ben Todd: Yeah.

Arden Koehler: Okay. Yeah. So again, very abstract: whatever it is that turns out to be the best way of figuring out how to do the most good.

Ben Todd: Yeah. I mean, in general, you have this just big question of how narrow or broad to make the definition of effective altruism and it is a difficult thing to say.

The rigorous argument for effective altruism [00:28:35]

Arden Koehler: Do you want to just talk about, why is effective altruism important? Why is this project important? And we’ve sort of gestured at a couple of reasons, but what’s the rigorous argument?

Ben Todd: Yeah. And it’s a bit alarming in a way that there isn’t just a rigorous argument for effective altruism written up somewhere, because that seems like that’s the central thing that people interested in effective altruism care about. So yeah, I was thinking a bit about what that argument should actually be recently. Yeah, I think there’s like three main premises. And yeah, but before we get into those, just what is actually the claim. And so I see the claim as something like, if you care about the common good, so we’re just taking that as a given and starting from that point, then it’s a mistake not to pursue the project of effective altruism, where the project I defined earlier as trying to seek the very best ways of helping others, rather than just the typical ways.

Arden Koehler: Okay.

Ben Todd: And then also the next clause, it’s more of a mistake the greater to the extent to which each of the three premises hold. So there’s kind of a spectrum of how important effective altruism is, and it might be more or less of a mistake to not be into effective altruism.

Arden Koehler: All right. And those are the three premises that you’re about to give us?

Ben Todd: Yes.

Arden Koehler: Okay, so we can come back to it.

Ben Todd: Yeah. There’s a lot to hold into your head.

Arden Koehler: Yeah. Let’s come back to the spectrum thing after we have the premises, to make it a bit more clear.

Ben Todd: Cool.

Arden Koehler: All right, so what are the premises? How do we get to this conclusion?

Ben Todd: Yeah, so I think one premise is, I call it the spread premise, which is just that there are big differences about how much different actions contribute to the common good. Second one is identifiability. So it’s that we can actually find these unusually high impact actions with moderate effort. And then the third one is novelty. So that the things that we could find that are really good are different from the ones that common sense would get us to focus on normally.

Arden Koehler: So if there are these big differences in how much good you can do, and you can figure out what they are, and they aren’t what you’re already doing, then it’s important to that extent to try to figure out what the best ways of doing good are.

Ben Todd: Yeah, to focus on those actions when you’re trying to do good.

Arden Koehler: Okay.

Ben Todd: Yes. And you’ve said, “Well, this is starting from the point where yes, you do care about the common good.” And so it’s kind of saying, “Well, if you care about the common good, some ways of helping the common good are way better than others. So to not do those things, you’re just failing in your stated goal.”

Arden Koehler: Okay. So maybe in order to see how these premises add up to the conclusion, we can talk about each premise and what it would mean if it was false. And then some reasons to think it could be false, but why you aren’t persuaded by them. So let’s take the first one. So, if it were not the case that there are these big differences in how much actions contribute to the common good, I guess that would mean that it’s not that important to figure out the best ones, because they’re only a little better than the worst ones anyway. Is that the idea?

Ben Todd: Yeah, exactly. Though it’s maybe worth saying, again, it’s good to think of this as a spectrum. If some of the best things you can focus on are like 1,000 times better than others, then that means effective altruism is enormously important. But even if it’s only like twice as good, well, say in one career you could save one person’s life and another career you could save two people’s lives, well, that’s a whole person’s life. That’s a huge deal. Even then it might be pretty important to really be attentive to these differences in scale.

Arden Koehler: Yeah. So I guess it would defeat the argument if it was like, “Oh, well actually it’s only 1.005 times better.”

Ben Todd: Yes.

Arden Koehler: So it’s like, you make somebody’s life a tiny bit better and you’re like, “Well…”

Ben Todd: Yeah. I mean, I think they are, again, more likely to kind of be defeated by the combination of things. Like if you’re more pessimistic about being able to find the best things, and they’re a bit less novel, and there’s a lower spread, then you end up being like, “Oh, effective altruism is kind of meh.”

Arden Koehler: I was thinking it actually could be defeated by any of them being false. So, even if you could find them and they weren’t what you were already doing, if they were only a tiny bit better at least effective altruism wouldn’t be very important.

Ben Todd: Yeah. No, so I definitely agree with that.

Arden Koehler: Okay.

Ben Todd: If that first premise is just false, let’s say the best things are only like 1% better than the average thing, then it’s already not important. But in practice, no one seems to hold that position. The thing that’s more realistic is to just be a bit unenthused about all three together.

Arden Koehler: Okay. And that can sort of add up?

Ben Todd: Yeah.

Arden Koehler: Yeah, so I guess just to make it ultra clear, and maybe we’re belaboring this slightly and our producer can decide to cut this if this is too much belaboring… So the second premise, if you can’t figure out what it is that is the best thing then, or you can’t do it without spending maybe so many resources that you’re bankrupt by the time you figure it out, then it’s not worth it.

Ben Todd: Yes.

Arden Koehler: And there might be some great actions out there, but you just can’t figure out what they are. And then the third one I guess is like, “Well, if the third premise is false, then in some sense we’re already doing effective altruism. So it’s not important to do it on purpose because we’ll do it on accident anyway.”

Ben Todd: Yeah. It’s almost not that effective altruism is wrong, it’s just that it’s uninteresting.

Arden Koehler: Right.

Ben Todd: But then, yeah, it wouldn’t need to be its own thing. It’s just, yeah, we’re already doing it.

Arden Koehler: Cool. So maybe hopefully someday it will be the case that the third premise is false.

Ben Todd: Yes.

Objections to effective altruism [00:33:42]

Arden Koehler: So with these, it does seem like this is a valid argument to me. So I guess if somebody was going to object, they’d have to reject one of the premises.

Ben Todd: Yes.

Arden Koehler: So what are some ways that people could reject these premises?

Ben Todd: Yeah, well maybe just to clarify, when you’re rejecting effective altruism, you can reject the core ideas, which is what I’ve been trying to cover here. You could also just say, “Well, I agree with the core ideas, but EA as it exists in practice is not actually pursuing the thing that matters.” Mayb e it’s too biased, or all kinds of reasons. And so, yeah, there’s a whole category of objections to effective altruism in practice, which might be good ones. But we’re just here talking about, if you really want to get at the meat, the very, very core fundamental underlying idea, what do you need to do?

Arden Koehler: Right. So I guess if somebody objected to effective altruism in practice, but they bought this argument that we’ve been talking about, maybe the thing for them to do would be to start their own effective altruism movement with a different name that gets it right, instead of going off.

Ben Todd: Yeah, I mean an objection to effective altruism in practice could just be something like, “The whole idea of a movement is a bad thing because it ruins people’s epistemology.”

Arden Koehler: It makes them biased because they want to promote their movement or something.

Ben Todd: Yeah, or it means people then they’re too influenced by the social norms in the community rather than thinking for themselves or something. But yeah, you should want to still be an effective altruist in some form, just not with the current community.

Arden Koehler: Got it. Okay. So objections to the premises. How might somebody go about rejecting one of these?

Ben Todd: Yeah. So there’s a lot of different options. So I think the way I most like to see it is, there’s a spectrum of how important effective altruism is, which is basically determined by how much better the best actions are compared to what you would have done otherwise.

Arden Koehler: And how easily we can figure them out and how different they are from what we’re doing already.

Ben Todd: Yes. Okay, yeah. I mean, though, if you were going to boil it down, it would be expected value of best thing you can find with effective altruism, versus expected value of what you would have done without effective altruism. And the difference between those two is kind of like the additional impact you have due to effective altruism. And so basically, how much of a mistake it is not to be an effective altruist depends on how big that thing is.

Ben Todd: And so there’s basically a bunch of forces that push that thing out to seeming very big, but there’s also a bunch of forces that push it back towards the best thing to do is closer to common sense or what people would have done otherwise. And which of those things you think dominates determines how important or how much of a mistake it is not to pursue the project of effective altruism.

Arden Koehler: Okay.

Ben Todd: Maybe just one thing that pushes the differences out towards being bigger, I think is… So neglectedness as a rule of thumb is something in this area, where if you think that all else equal, something that’s more neglected, you can have more impact because there’s been less diminishing returns.

Arden Koehler: So fewer people are working on it, so it’s less the case that an additional–

Ben Todd: The best things have already been done.

Arden Koehler: Yeah.

Ben Todd: And then naively it seems like there’s very big differences in neglectedness. Some priorities have had over 1,000 times more invested in them than others. And so if those priorities seemed roughly similarly important and similarly tractable like the other two dimensions, then additional resources for the more neglected one would achieve 1,000 times as much than that.

Arden Koehler: So just to get slightly more concrete here, when we talk about priorities, what are some examples?

Ben Todd: I was mainly thinking of different global problems you could work on. Though yeah, you could think about it in terms of what interventions you want to do as well.

Arden Koehler: Okay. But you’re thinking about things like education in the US versus global health, versus trying to get representation for future generations.

Ben Todd: Yeah. So education in the US gets around $1 trillion spent on it each year, I suppose to benefit… I don’t know how many kids there are actually in American schools each year. Probably, what, a few tens of millions? Whereas, global health… So all international aid I think is around $140 billion as of a couple of years ago. So there’s between five and 10 times as much being spent on US education as is being spent on all international aid. That’s not a completely fair comparison because there’s other resources going to the global pool besides international aid.

Arden Koehler: Right. I mean, like the resources of their own government and economies presumably.

Ben Todd: Yeah.

Arden Koehler: Yeah. Okay.

Ben Todd: But still, I think even if you take account of that, probably more is being spent on US education, and it’s like we’re also talking about a billion people rather than a few tens of millions of people. So there seems to be a really big difference between those two.

Arden Koehler: Although I guess the number of people, that’s more on the importance category than the neglectedness category.

Ben Todd: Yeah, that’s true, but ideally we’d be comparing problems of similar importance and then showing the big neglectedness differences.

Arden Koehler: I guess you could talk about, like, the same number of people as there are kids in the US, how much money is going to improving their health? People who are underprivileged around the world.

Ben Todd: Yeah. Well, I mean, maybe a clearer example is between different existential risks. So climate change has hundreds of billions spent on it per year and is widely agreed by people who want to have a positive impact to be perhaps the most pressing global priority. Whereas as we’ve covered on this podcast, we think how AI goes, especially from a longtermist perspective, seems to be really crucial as well. But even today, having become much more popular as an issue, it’s still under $100 million a year spent on research into AI safety and longtermist AI policy. So that’s at least a factor of 1,000 difference between the two in terms of how neglected they are.

Arden Koehler: So that suggests very big differences in working to try to address AI safety versus working to address climate change.

Ben Todd: Yeah. And where climate change is already a really pressing global priority compared to just more everyday priorities that we might have.

Arden Koehler: Okay. So what’s a factor that pushes in the other direction, that somebody who is skeptical of effective altruism or the importance of effective altruism, might bring up?

Ben Todd: So I think one is epistemic humility. This is the idea that, what should we believe about a difficult question like how much of a pressing priority is AI safety? One answer to that is to just try and figure it out for yourself. Another answer to that is just go with what other sensible people think and take an average across them. And some evidence that that kind of approach works is like in forecasting. So people have tried to measure, what’s the best way to predict geopolitical events, which are super hard to predict. And it turns out that getting like 100 people who are really good at forecasting, seeing what they think will happen, and then averaging that together is a really good way of doing it. And it’s usually better than any individual person.

Arden Koehler: Also, I mean, I feel like more basically it just seems like the community as a whole will often be better at knowing stuff than a single person because they just have a lot more brain power between them.

Ben Todd: Yeah. And you can think of it as like, if someone else has a radically different view of what the biggest priorities are from you, and you’re in a similar position epistemically, why would you think you’re right and they’re wrong? Unless you have some really clear explanation of like, they’re making this mistake and you haven’t made some other mistakes that they’ve figured out and you haven’t figured out. So there’s a kind of symmetry argument for going with what other people think.

Ben Todd: And yeah, it seems like if you put a lot of weight on that kind of argument, then you should be focusing on the priorities that a sensible coalition of other people would think are the biggest priorities. And that almost brings you back to the most common sense priorities by definition.

Arden Koehler: Just to clarify, this seems like more of a reason to think that, in fact, we’re already doing the things that are best for contributing to the common good, because common wisdom has come to think whatever priorities we’re prioritizing are the best things, and so we should probably defer to that.

Ben Todd: Yes, if you really buy that in a strong form.

Arden Koehler: Okay.

Ben Todd: Yeah. It might still mean you should change your career a lot. The thing you should probably defer to is, what people in general think from an impartial perspective, what would most benefit others is the best thing. Which is not what most people do with their lives. So, for instance, it probably means you should work in climate change, because that seems to be what… When people are thinking about it from a global perspective, what most matters, often climate change is the answer. And obviously, most people don’t spend their careers fighting climate change. So, in practice you–

Arden Koehler: Yeah, thanks for that clarification.

Ben Todd: Yeah. You’d probably still change your life, but in a much less extreme way than if it’s like, “No, AI safety is the main thing.”

Arden Koehler: Right. And in particular, it would mean that the project of effective altruism, which involves doing all of this relitigating of what really is the best way to do good, is less important, because in fact, we all can kind of just defer to each other or the common wisdom about…

Ben Todd: Yeah. Yeah, but I mean, we can see that this argument in a strong form becomes really implausible, because it’s like no one can ever discover a new global priority because it’s already just everyone knows what the best thing is.

Arden Koehler: Yeah.

Ben Todd: And I guess I would also say that I think in practice, if we got loads of experts together and said, “What were the most pressing global problems”? Probably some of them would still agree we should put some effort into AI safety. So it’s not actually not being epistemically humble to invest a little bit into AI safety. And so I’m not actually even sure that our current priorities conflict with epistemic humility. But what I actually want to do here is start what I see as a real debate about what the actual core ideas of effective altruism are. And I’m just saying, one kind of interesting avenue for attack seems to be a strong form of epistemic humility.

Arden Koehler: Okay. So what are some other interesting avenues for attack?

Ben Todd: Well, so if we go back to this kind of, there’s some things that make us more contrarian or further away from common sense and there being big differences there, and other things that make us go back towards common sense, yeah, another factor that pushes us back is regression to the mean.

Arden Koehler: Yeah, so what’s that mean?

Ben Todd: Okay, so you could come along and you’d be like, “Well, I did this estimate and I found that there’s this amazing global priority that no one else is working on and it’s like 100 times better than what everyone else is doing.” So then the question is, should you just trust that, or should you figure that you’ve probably made a mistake somewhere? And because your calculation has said there’s this thing that’s amazing compared to what everyone else is doing, most likely you’ve made an error in the direction of it being better than it actually is. And when you figure out all the errors in your argument, it will make that thing end up seeming a lot worse than your naive calculation suggests.

Arden Koehler: So I guess to sum this up is, it’s something like, “There’s this phenomenon where usually when you do more investigation on things that seem like outliers, they will tend to be closer to the average upon further investigation. So you should think that’s true of whatever problem you think is really an outlier when it comes to how good it is to work on”.

Ben Todd: Yeah, that’s the kind of broad character of things. Regression to the mean is a particular thing in statistics, which also has a formal meaning, which is pretty related to what we’re saying here, but not exactly the same. And so then, yeah, the question is, how big a correction should you make? If you should make a huge correction, generally then, again, effective altruism becomes less interesting because we’re the people who are being like, “Oh, this is the thing that everyone else is wrong about.” Whereas if it’s more modest, then, well, maybe the things that seem best to us aren’t quite as good as they first seem, but they could still be much higher impact than what we would have focused on any way.

Arden Koehler: So I guess this is another one where if you took it to the really extreme, it would be implausible, right? Because it’s like, the really extreme version would be like, “Well, you should expect everything to be exactly the same in terms of how good it is to work on”, right?

Ben Todd: Yeah. Though one thing is, regression to the mean happens more for like speculative, uncertain estimates than for really robust things.

Arden Koehler: Okay.

Ben Todd: So it could end up meaning you focus on the most robust ways of doing good you can think of which then, I don’t know, maybe that would lead you back to something like global health.

Arden Koehler: And so by robust, you mean like, you’re really unlikely to be wrong that, for instance, distributing treated malaria nets… Sorry, insecticide-treated bed nets for malaria prevention is in fact not saving lives. That seems like really unlikely because we have good evidence that it is. So you might want to go with that instead of the crazy speculative stuff.

Ben Todd: Yeah. Or maybe you think you should just help friends that you know, because you know them so well that you can be guaranteed that it helps them a little bit.

Arden Koehler: I see. Yeah.

Ben Todd: It’s actually really hard to know what makes people happy even those that you know.

Arden Koehler: You should just meddle in your friends’ lives and assume you know what’s best for them.

Ben Todd: Yeah. I suppose if it’s like you need to help them move house or something, you’ve at least saved them some time, so that’s a benefit.

Arden Koehler: It would be weird if that turned out not to be good for them.

Ben Todd: Yeah.

Arden Koehler: Okay, yeah. So regression to the mean, epistemic humility. Yeah, are there any others that are worth bringing up?

Ben Todd: Yeah. Okay, so just to name a few other things maybe a bit more quickly. But I think there’s something in the area of people’s personal passions and interests being important, where if you take the more extreme perspective, the best things are likely to be these things that are… Well, if we just focus on neglectedness, the most effective things are just the things that are important but no one else is working on. And so those are very unlikely to be things that you happen to have developed a passion for, because you will have never stumbled across them before until you actively seek them out.

Ben Todd: And then that does make it harder in some ways to work on them. A way this commonly comes up is someone’s like, “Okay, what about working on biorisks?” And then someone’s like, “Oh, I’ve never heard of that before. I’m not passionate about that at all. It sounds totally outside of my normal–”

Arden Koehler: “And I have no experience that would make me good at working on it because I hadn’t heard of it.”

Ben Todd: Yeah. It’s harder to get career capital in these more neglected issues as well, whereas maybe they were already passionate about climate change because that’s something that lots of their friends have talked about and they’ve worked on before, and it would be kind of easier for them to be motivated to work on climate change. And if you think effects like that are really big, then again, that seems to push us back more towards, just do something that’s generally in the air and common sense and that you’re motivated by, and don’t try to seek out these really unusual novel priorities that would make effective altruism the most important.

Arden Koehler: Yeah, so it seems like this is sort of a way of directly pushing back on the neglectedness argument. So you say like, “Okay, yeah, more neglected problems would be better to work on, all else equal, but, in fact, when something is neglected, it also means that people are going to be worse at working on them.”

Ben Todd: Yes.

Arden Koehler: Or so they’ll sort of maybe cancel out or roughly cancel out or something like that. Another, I guess related argument would be, if something is really neglected, it’s hard to make progress because nobody listens to you. You’re sitting there being like, “Let’s all work on AI safety,” and everyone’s like, “We’re not going to give you a grant because we don’t think that seems important,” and therefore, in fact, it’s really hard to make progress.

Ben Todd: Yeah. So if we think impact is how pressing the cause is times how good your opportunity is for helping times how skilled you are at that thing, often there seems to be some anti-correlation between how good the opportunities are you can get versus pressingness. Where in an area that’s really neglected, it’s just hard to do something really, really big, because there’s just fewer resources around.

Arden Koehler: Okay. Yeah. So wait. No, now I feel confused, because we started that off by saying, “Well, if something is not neglected, then it’s more likely that only some not very good opportunities will be left for you to take”. But now you’re saying, “In fact, if something is more neglected, that means that there will be worse opportunities”. So how do those things interact?

Ben Todd: Well, so one term is kind of how efficient extra resources are in producing good. So that’s what we mean when it’s a really pressing area. It’s like, an additional dollar goes a long way. But then there’s, how much leverage can you bring to bear on that problem? And it seems like being a bit less neglected is helpful on that one. So for instance, if you’re a scientific researcher, you could get bigger grants to work in a more mainstream priority.

Arden Koehler: Like climate change?

Ben Todd: Yeah. Or if you’re able to become a leader in the climate movement, you’ll be leading more people than if you become a leader of a much more niche issue. And so it’s a way in which it becomes a bit more tractable to work on things that are a bit less neglected. Yeah, exactly where you want to account for that in the framework can go different ways. But I think there is some effect like this.

Ben Todd: And then, I mean, my personal view is like there’s probably reasons not to work on very, very crazily most neglected things. But I also doubt that the optimal place to be is on just common sense. That it’s somewhere in between the two, but there’s kind of a sweet spot of neglectedness, where it’s not so neglected that you can’t do anything and everyone thinks you’re crazy and you’re probably going to be wrong about it, but it’s neglected enough that you have a edge on just doing what everyone else is doing.

Arden Koehler: Yeah. Interesting. I think my intuition says that even if something were so neglected that literally nobody else in the world was working on it, I would still think that even though it’d be really hard to make progress, you’d probably still make a really big difference. Because I’m imagining the very first person to talk about climate change or something… I mean, I actually don’t know the history, but I guess I’m imagining that even if nobody listened to them for their entire life, the fact that there’s one paper out there that the next person could cite 30 years later, turned out to maybe accelerate the movement in a couple of decades and plausibly was really impactful even though it sounds like a frustrating existence.

Ben Todd: Yeah, I think there’s something to that. Yeah. One thing that I think slightly helps to explain that is also value of information, where being the first person ever to work on a cause helps us figure out a lot about whether that cause is worth investing in. So yeah, you get those benefits right at the start.

Arden Koehler: Okay.

Ben Todd: Yeah, I agree this could even end up just turning out that, yeah, you should still focus on the very most neglected things.

Arden Koehler: Okay.

Ben Todd: In which case, effective altruism is more important.

Arden Koehler: So one other argument that I’ve heard you talk about before is an objection to, I guess, the third premise. I think Tyler Cowen argues that economic growth is in fact the best thing we can do for the common good, and that economic growth is basically best served by people just going about their business, trying to make a living for themselves. (In our interview with Tyler Cowen, it’s not clear that he thinks economic growth is a better option than work to prevent existential risks – Ed.) And that’s what people do anyway. And so, if that were the case, then everyone would just already be doing what is the best action for making the world a better place.

Ben Todd: Yes.

Arden Koehler: What do you think about that argument?

Ben Todd: Yeah. I mean, I think that’s a good style of objection to effective altruism, is to try and show the things we’re already working on are in fact the best things from an impartial perspective. As to whether Tyler Cowen’s one works, I mean, I guess I don’t… Yeah, I think existential risks are more pressing from a longtermist perspective than economic growth, so that would be one difference I’d have. And then, I still feel pretty unsure about even if economic growth was the… Well, it’s actually the economic growth rate is what he says is most important.

Arden Koehler: Yeah.

Ben Todd: It’s the most key thing. Whether that is actually best served by everyone going about their daily business… I mean, in our podcast with Tyler Cowen, he kind of agrees that this advice doesn’t necessarily apply to very talented people. Because he’s kind of like, “Well, if you go and become an economics researcher or work at a think tank that’s promoting good policies for growth, that seems a lot better than just having a regular job, in terms of how much you speed up economic growth.”

Ben Todd: And it would mean things like we should really be focusing on innovation, because if you want to change the economic growth rate, you basically have to change the rate of innovation in society. And most jobs don’t seem to contribute that much to innovation, at least directly. They kind of do by increasing GDP, and then some of that goes into innovation later. But I think it would be more like, do the thing you can do that most supports economic innovation. And that could end up being pretty different.

Arden Koehler: Yeah.

Ben Todd: And I think at the very least people should be donating things to… They should be investing in startups. They should just be giving grants to interesting scientists.

Arden Koehler: Okay, yeah.

Ben Todd: That would be the highest impact thing they could do.

Arden Koehler: All right. So it might mean it’s a bit closer to what we do automatically, but it wouldn’t be just, in fact, the things people are inclined to do already.

Ben Todd: I mean, that’s my suspicion, but it’s still an interesting question.

Why do we think that effective altruism is actually important [00:53:04]

Arden Koehler: Okay. We’ve talked a bunch about reasons to reject this argument, but obviously, you and I, in fact, do not reject this argument, at least provisionally, we think that effective altruism is really important. Why think that these three premises that imply that it’s important are true?

Ben Todd: Yeah. Well, one is that they might just seem obvious to you. But putting that aside, I think probably the best way to demonstrate the premises is just by giving specific examples of things we can do that have much more impact than what people would normally do. And yeah, that is kind of the whole subject of 80,000 Hours. We try to just be kind of constantly giving examples of those types of things. But this is now where I would place something like the arguments we talked about earlier for why to donate to global health charities. Because if you’re saying, well, you could save a life equivalent for roughly one and a half thousand dollars with GiveWell currently… Well, that would mean that, say, so typical college graduates, I think in the US, make like $70,000-80,000 a year on average. That’s the mean. And then so suppose you give 10% of that, that would be $8,000 a year. And so then that’s, what’s that? Roughly four or five lives saved per year. And so then over a 40 year career, that’s 40 times five, which I think is 200 lives, right?

Arden Koehler: Yes.

(Editor’s note: ‘saving a life equivalent’ isn’t the same as ‘saving a life’, so this doesn’t mean you can literally save a life for $1,500. GiveWell’s best estimates move around regularly, but as of the date of publication they estimate that donating around $3,400 to Malaria Consortium can save a life – which would more than halve the possible number of lives saved in the above calculation.)

Ben Todd: And then that’s an incredible amount of good. And then it seems like people don’t normally do that. In fact, you can even just have the same job that you would have had otherwise. It’s not even a big change to your life, but you’ve also saved 400 lives.

Arden Koehler: 200.

Ben Todd: Sorry, 200.

Arden Koehler: But yeah, that is pretty incredible. If you think about it, if you save 200 people in one event, you would think, “I’m a hero”, but this is a way that you can save 200 people over the course of your life.

Ben Todd: Yes. But then the point is not that that’s what effective altruism says everyone should do, this is just a proof of concept. We’ve identified at least one way to do way more good than we kind of normally think is possible. But effective altruism is important just if we can find any convincing example, we just need to be able to show this one thing.

Arden Koehler: And it’s more important if we could find something that was doing even more good than that.

Ben Todd: Yes. And so, yeah, there’s lots of other ways of demonstrating this. And I think we have content about why we think some global problems are maybe 10 or a 100 times more pressing than others. And then again, that would mean if you can work on one of the more pressing ones and if it’s really a 100 times more pressing, then that means that in your 40 year career, you’d achieve as much as what would normally take 4,000 years of work. Again, that would mean at least in terms of contribution to the common good, a really huge difference. And if that’s true, then it would mean that taking this seriously is really important.

Arden Koehler: Right. That’s if the problems we think are the most pressing are a 100 times more pressing than the problem you would have worked on otherwise.

Ben Todd: Yes. That is a bit of a complication. Because if you would have done something like climate change otherwise and that’s already pretty good, maybe it’s hard to find something that’s a 100 times better than climate change. But I would still say that some of our most neglected things are maybe 10X better.

Arden Koehler: Right. And presumably lots of people are working on things where we think the most pressing issues are a 100 times better than, or more pressing.

Ben Todd: Yeah. Though one thing pushing back is… It is important, you could switch to a more pressing issue, but then if that means that, say, you end up in a career with worse personal fit, you need to take into account that as well. And again, that might make the differences a bit smaller, but I still think there can often be large differences.

Arden Koehler: Yeah. It also seems like even if you don’t have a good personal fit, for instance, like AI safety, among the set of top 10 or something most pressing problems, probably there’s something that you could do a pretty good job at for most people.

Ben Todd: Yeah, and you could always donate to the most pressing thing. In a way, that’s a good baseline is earning to give and then donating whatever you could give to the most pressing thing.

Arden Koehler: Right. That’s kind of a version of the ‘just donating 10% example’, but maybe you even decide to try to donate more by taking a higher earning career.

Ben Todd: Yeah. And some things seem much higher impact than what we would’ve done otherwise. And yeah, I think there’s maybe just loads of other examples. Just to give a quite randomly different one is in Doing Good Better and on Brian Tomasik’s blog, which we can link to, he does an analysis of how much animal suffering you avoid by cutting out different meat products. And he actually shows that there’s roughly a 1,000 times difference in animal suffering between a pound of beef versus a pound of chicken or fish. And so that would actually mean that you could kind of cut out 99% of the impact of your diet on animal suffering just by giving up chicken and fish, which again is not a commonly held thing to do. Most people are like “You should either be vegan”, or just “Whatever, eat meat: it’s fine”.

Arden Koehler: Yeah. I guess in the current context, maybe the comparison is between somebody who sort of chooses randomly what meat to give up, because they’re thinking, “Look, I want to reduce my contribution to animal suffering so I’m going to give up, let’s say chicken”. And they sort of pick it out like that.

Ben Todd: Well yeah, most people actually do the opposite where they give up beef and red meat first and then they keep eating fish.

Arden Koehler: Yeah. Red meat just seems meatier somehow, right? Whereas this says, better to give up chicken or fish.

Ben Todd: Oh, and then people keep eating eggs as well, which is actually on the bad end of the spectrum.

Arden Koehler: Yeah. Okay. Those are some sort of reasons to believe in the first premise, any sort of–

Ben Todd: Well no, I think those actually show all three premises because you’re just demonstrating, taking this perspective, gives you this big boost in impact so all three have to be true for them to work.

Arden Koehler: Yeah. Okay. No, that’s helpful. Because we had to be able to figure out that in fact some problems are, or at least we think we’ve provisionally figured out that they are more impactful to work on in expectation. And also it’s not just what everyone’s doing otherwise. No, thank you.

Ben Todd: Yeah. But then yeah, if you would disagree with our analysis and you disagree for every case that we might bring, then that’s another way of objecting to effective altruism.

Arden Koehler: Cool. That makes sense. All right. Is there anything more to say on why believe this argument for the importance of effective altruism?

Ben Todd: Yeah. I think there’s some qualitative arguments to help make it plausible because we kind of come in with, well, what should our prior be? Should we expect it’s possible to have way more impact? Do way more to help the common good than we might’ve first thought? Or should we be skeptical of these examples that I just brought up and whether we might be able to find other examples besides them? And I think there’s a couple of things there. One is just the question of what your priors should be.

Arden Koehler: That’s just the beliefs you come in with before you think about it.

Ben Todd: Yeah. And it seems like my prior is basically that some actions will do more good than others and some actions will be more costly than others. And how cost-effective an action is, is kind of the multiple of two: it’s one divided by the other. And so there’s big differences between both of those two factors, then we should expect just big differences in the effectiveness of different actions.

Arden Koehler: I guess one reason to think the opposite would be if you thought there was sort of an efficient market for doing good. If you thought, “Well, if something really did do more good than other things and everyone would have already started doing that, and then have taken all the opportunities and so practically everything that’s available to me is going to be roughly equal in impact”.

Ben Todd: Yeah, exactly. Another kind of big way of having the debate about effective altruism is just the question of how efficient are our current institutions at doing good? And so, yeah, I guess there I would say my kind of theoretical perspective is our current institutions in some ways do quite a good job of furthering the common good, but they have lots of important gaps in them. And the existence of these gaps is another reason why we should expect effective altruism to be important. And you can think of one whole framing of effective altruism is what we’re trying to do is find the biggest, most pressing gaps in our existing institutions and then fill those.

Arden Koehler: Yeah. I don’t think I know that many people who really believe that the sort of market for doing good is super efficient. Although actually, maybe my dad is an exception and sorry dad for calling you out. But I feel like, I remember showing him the Peter Singer article ‘Famine, Affluence, and Morality’ and he was like, “Nah, but like all the things you try to do are going to have these crazy side effects and they’ll all kind of end up maybe equal”. Okay. I’m probably misrepresenting it. But I think maybe that was one way of thinking maybe that it was sort of efficient. If there was a great opportunity that didn’t have crazy side effects that made it super unpredictable, then somebody would have done it.

Ben Todd: Yeah. Exactly. I think there’s some plausibility in that. Another way you could get to kind of reasonable efficiency quite easily is, so we tend to think, well, I’m a bit worried about this getting too technical; if we think different actions, how much impact they have, is a log normal distribution. A normal distribution is kind of all the different actions are quite clustered around the mean.

Arden Koehler: It’s a bell curve.

Ben Todd: Yeah. It’s a bell curve. Log normal distribution looks a bit the same, but then it has what’s called a heavy tail. There’s like a small number of things that are way better than the median thing. And so income is log normally distributed where the highest earning people earn hundreds of times more than the median person. But if you have one of these log normal distributions, because that tail is only a minority of the things, you only need that to be some kind of effectiveness-minded people to go and take the tail. Maybe like Bill Gates, he’s just getting all the good opportunities. And then once the tail is gone, everything else is kind of back at the median. Then you could end up being kind of just back at the median again, even if there’s only a few percent of people that are really kind of impact minded.

Arden Koehler: Interesting. Yeah, that’s really interesting because then that would be a way of saying, “Okay, I agree with the first premise that some things do a lot more good than others and I think we could figure out what they are, but I just think Bill Gates has already figured it out and so like the options available to me are, in fact, all relatively similar in terms of impact.”

Ben Todd: Yes. And I think this did happen a little bit when people went into global health, because one of the things that seems best is childhood vaccinations and Bill Gates, plus a bunch of governments and so on, who have all banded together to do this, have kind of done all of the easy opportunities to do that around the world. And so that stuff was all taken by Bill Gates and that stuff seems more effective than the malaria nets.

Arden Koehler: I see. Interesting. Yeah. Cool. Okay.

Ben Todd: Yeah, there is a bit of that happening, but then the question is, is it so much that it takes away all the good things?

Arden Koehler: Right. And maybe it takes away all the good things in a particular area perhaps, or sorry, by good of course we mean better than the median. The median might still be good. But then, can you move to another area and find some more exceptional opportunities there?

Ben Todd: Just to kind of wrap up on efficiency, we can kind of think, well, our existing institutions, we kind of have the free market and that’s kind of incentivizing us to do the things that most, well kind of the things that people are most willing to pay for in the market. And so there’s a lot of incentives to do that. And then we have our governments, which try to… Obviously the free market leaves out lots of things as market failures, public goods, things like that and then we try to use the government to take those things. But then it seems clear to me that those two things don’t cover everything that might matter, especially because governments mainly focus on what their citizens need and they focus less on global public goods. They focus on public goods within their own countries. And then also they have a lot of issues with short-termism, especially focused on the election cycle. We should kind of think that providing fixing market failures in 30 years’ time, we might think there’s not much being done about that currently in our society.

Arden Koehler: Just maybe kind of climate change might be an example of a sort of longer term market failure.

Ben Todd: And it’s also global again where if the US spends loads to stop climate change, most of the benefits of that accrue to other countries rather than the present US citizens.

Arden Koehler: Right. And then we shouldn’t expect, of course, for the sort of consumer market to provide all of the goods because it’s mostly responding to, it’s skewed by the fact that some people have a lot more resources than others, even though it’s not because their interests are more important.

Ben Todd: Yes. And then all the kind of market failures and coordination failures, the kind of classic economic studies.

Arden Koehler: Okay. So that’s efficiency and why we should probably not expect the world to be already doing all of the things to make the world a better place that are the best.

Ben Todd: We, in theory, then do have the nonprofit sector to try and fill in these gaps, but that’s only a few percent of GDP and it seems to have plenty of its own issues. This would though be the framing is just to try to think, “Well, we’ve got academia, nonprofits, government, free markets. What’s not being covered by those?” And if there are those gaps, then that would probably be what effective altruism would focus on.

Arden Koehler: Yeah. And I think it would take a pretty optimistic person to be like, “Nope, no gaps. It’s totally covered”.

Ben Todd: Yeah. And I guess there’s another big argument for that is, longtermism is a big argument for that because just, well, there’s all these future generations who just don’t really have any say in these institutions. And so there’s no reason to expect that our existing institutions are a tool doing what’s best from a longtermist perspective.

Arden Koehler: Right. Okay. Is there anything else to say about why people might find this argument persuasive or unpersuasive?

Ben Todd: Well so there’s actually two others that have been taken by people. One, I don’t want to cover too much now, but it’s the kind of argument from human irrationality where it’s just like, “Well, we have loads of cognitive biases. There’s no reason to think that our intuitions will be good at finding the things that do the most, so we need to take the systematic approach to finding them”. And then the one that I’ve kind of put more–

Arden Koehler: Oh, I see. That’s an argument for, that’s the reason to support the argument for effective altruism. I thought you were going to say the opposite but it’s, “Oh, we’re so bad at figuring stuff out that, in fact, we can’t ever figure out what does the most good because we’re too irrational.

Ben Todd: Yeah. So this would, if you see effective altruism as the claim of something like, “Well, what we need to do is apply systematic reasoning to doing good”. Then you could be like, “Currently we don’t apply systematic reasoning and that doesn’t work. But if we apply systematic reasoning, then we’ll do way better, so therefore effective altruism”. That’s kind of how that argument works. That’s a bit of a different way of, that’s putting a lot of emphasis on the evidence and reason bit, which I actually wanted to de-emphasize.

Ben Todd: Yeah. Anyway, the thing I focus on more is this idea that we’re in an unusual time in history. And so if you think over the last 200 years, people who were educated and living in rich countries, we’ve become way wealthier than almost anyone in human history. Whereas, well, a lot of people are still kind of living at subsistence level and just being in that really privileged position should mean that there are things we can do to help others to a big degree. And it’s not just the kind of income inequality in the present generation, but also it seems like things we do today potentially have these really long term effects. There’s also a kind of longtermist version of this argument, where if I was a medieval peasant, I’m not causing climate change, I can’t help to start a nuclear war. I can’t elect a totalitarian government because these things are just not technologically possible.

Ben Todd: Whereas as citizens of nuclear powers today, and countries that are emitting lots of CO2 and doing factory farming and all these things, our actions can have these really massive kind of global and long term effects. And all this technology and wealth basically gives us this unusual amount of power that we’ve never had before in human history. And we should now expect that that power will enable us to, well, either do much more harm than we could have ever done before or much more to benefit others than we could have done before. And so I kind of like to see it as the stakes have got bigger over time for this reason. And if you were a medieval peasant, then probably yeah, the best thing you can do from a moral perspective is just don’t cheat your neighbors, be kind of hardworking, look after your family, that’s the key stuff.

Arden Koehler: And that might still be true of lots of people today who are living at subsistence level. But for those of us who have the privilege to be able to.

Ben Todd: Have spare resources and yeah. Then now it’s really important for us to think about what exactly we can do with them that’s best. I don’t know if people are picking up that, but it’s starting a thunderstorm here.

Arden Koehler: Yeah. It’s starting to rain and thunder. Hopefully it’s not too loud, but maybe gives a little atmosphere. Here’s something somebody might say to that argument. They might say, “Okay. Yeah, sure. I’m richer than a medieval peasant, but so are the powers that be. Such that in fact, my power to change things, it hasn’t really changed very much because I have to go up against these huge corporate interests or huge governments that have all of these resources if I want to make change. And so in fact, I’m no more powerful than a medieval peasant would be.” Is there any response to that?

Ben Todd: I think compared to a medieval peasant, it’s just obviously wrong because at the very least you could donate 10% of your income and you could band together with a bunch of people and pay for corporate lobbyists to like work on your behalf or whatever the thing is that you think these powerful people are doing. I just think there is a some ability to, at the very least, convert our money into things that are helpful. But yeah, it doesn’t even depend on money. I think we have an analysis on the 80,000 Hours blog of the impact of voting in presidential elections. And we argue that, although your chance of making the difference in the election is absolutely tiny, you do have some chance, especially if you’re living in a swing state and if you do, it has such a big effect. You’re having basically a tiny influence over a massive thing. Those two things actually turn out to roughly cancel out and so each vote probably does in expectation have a significant impact.

Arden Koehler: Yeah. People should go check out that article on 80,000 Hours for the upcoming election. Go out and vote 80,000 Hours listeners, for the upcoming US election. Sorry, so US-centric, but it’s on the mind at the moment.

How to talk about effective altruism [01:10:51]

Arden Koehler: Okay. We’ve talked a bunch about the core ideas of EA and the argument for those ideas being important. But we said earlier that we’d talk a little bit about framing and how to talk about effective altruism, especially to people who don’t know very much about it. Do you have any sort of tips or ways you think it should be framed?

Ben Todd: Yeah. We opened the episode with me kind of suggesting we haven’t been explaining effective altruism that well, most people are engaging with not actually the core idea. And so, yeah, I think it would be great if we can start to develop some new ways of introducing effective altruism, which might avoid those misunderstandings or at the very least will then appeal to a different group of people. Because some people really like the global health evidence-based focused approach, but that’s not everyone that we want to appeal to. And so, yeah, I haven’t really… This is a big task. It’s kind of like developing a whole new way of introducing effective altruism, which I have not done, but I think there’s some kind of ideas to bear in mind, given the things we’ve been discussing.

Ben Todd: And so one thing I think is important is you can kind of see effective altruism as there’s two sides of things. One is effective altruism as a community, where that’s basically a group of people who are going around trying to do good in a certain way. And then there’s effective altruism as an intellectual project. We introduced both earlier. And I think most of our existing introductions tend to emphasize the community bit more than the intellectual bit, but actually both are really important and the nice thing about emphasizing the intellectual bit is it gets kind of more to the heart of effective altruism.

Ben Todd: And so, yeah, so an example of the kind of community approach is, “Okay, effective altruism is a group of people who give 10% of their money to charities such as those supported by GiveWell”. And that’s, again, very practical, ‘here’s what we do’ kind of explanation. Whereas the more intellectual project explanation would be, effective altruism is about the question of which ways to benefit others are most effective? And effective altruists like to do research into these questions and they try and tackle questions like which global problems are most pressing? And what even is the common good? And what implications might different definitions have? And which interventions are most effective? And should we help people now? Or try and invest and help people later?

Ben Todd: And so I like introducing effective altruism as a question and then giving examples of the types of questions that we think are really interesting and important. And then you can see if someone’s interested in one of those questions, then you can go more into, “Okay, well what do people say”? Some people say, “No, we should focus on the most evidence-based thing, because at least we can measure it.” Other people are like, “No, we should take a hits based approach so that we get the tail, the thing that’s most effective.” And that’s a debate within the community. And I think that type of approach is really good to appealing to someone who’re really interested in–

Arden Koehler: Intellectual debate.

Ben Todd: Yeah, intellectual debate in which, is at least… We don’t want only people who want to intellectually debate things within effective altruism because we do also want to do things, but we do need lots of people doing research as well as, so it’s great to be able to appeal to that type of person.

Arden Koehler: Yeah. Just to throw out a data point, which this is not contradict what you said, because what you said is we need both. But one thing that attracted me to effective altruism was, “Wow, these people are actually doing the stuff”. But I came from an unusual background. I was in academic philosophy, which is on the extreme side of the “Let’s have intellectual debates and not do that much stuff in the world”. Whereas I guess most people are not on that extreme.

Ben Todd: Yeah. And it is hard to say, in many ways, obviously, effective altruism is really intellectual and people aren’t good at practically doing things. It’s not totally obvious where we should push in that direction. But I think there is another big point in favor of the intellectual project approach of introducing effective altruism, which is that it makes it really clear that it’s not about particular answers. Whereas if you tried to do the practical introduction, then you’ll have to start talking about specific, practical things that people are doing. And then it’ll be very natural for people to then think, “Well, that is effective altruism”. Whereas, if you say, “Well, it’s like, people want to debate these types of questions”, it makes it really obvious that…

Arden Koehler: Yeah, I guess I feel like I would want to very slightly tweak the intellectual framing if it were me to be not so much about debate. Debate is important, but we want to emphasize that it’s about making progress and building on the knowledge of others. And in fact, actually we could debate all day about what it actually means to do the most good and maybe never feel satisfied fully with that answer. But the project is about coming to the best guess or what you think is actually the most justified course and building on the knowledge of others. And so it’s this kind of more, yeah, still always in the service of that practical project, and so I feel like I would want to emphasize that part.

Ben Todd: Yeah. And whenever you’re introducing effective altruism, the main question is what order do you say things in? And so I would definitely say, I would introduce the questions and be like, “This is what it’s about”. And then I would try to say fairly soon, “And there’s people actually putting this into action and that really matters because we can do more to solve pressing global problems.”

Arden Koehler: Yeah. And we have some provisional answers that are better than nothing. And we don’t need to start from scratch always.

Ben Todd: Yeah. I think I would definitely mention that in an introduction as well. It’s just, I would consider it not being the very, very first thing.

Arden Koehler: Yeah. That makes sense.

Ben Todd: I think one other thing we could work on more in when we introduce effective altruism is to rely a bit less on examples in global health. And that is very hard to do because those examples are so kind of concrete and quantifiable that it enables you to get across the ideas in a very solid way. But given that only say like 30%-ish of the resources of the community are working on that problem, I think it’s better to try to give a broader picture and especially because the global health stuff is already really out there. People kind of already think it’s more about that than it actually is. And it would be nice if we could slightly more emphasize some of the other cause areas more.

Arden Koehler: Yeah. That makes sense. Another reason to lead with global health is I do think it’s very central, it’s sort of the one to beat, you know what I mean? Well maybe this isn’t fair to certain other cause areas, but I think it’s oftentimes useful to think, is this thing better than–

Ben Todd: Just because it’s so quantifiable, you can use it as a baseline.

Arden Koehler: And it’s been studied a lot, it’s not quite as speculative.

Ben Todd: Yeah. Though I think if you’re a longtermist, then the baseline I might use would be avoiding CO2 emissions.

Arden Koehler: Oh interesting.

Ben Todd: Because we know that you can abate a ton of CO2 for about 10 euros, maybe one euro. And so that’s kind of something that anyone can do, is you could at least buy carbon credits or you could stop driving your car as much.

Arden Koehler: Yeah, that’s interesting. Okay. Yeah. Maybe that’s a good baseline too, but okay yeah. But in general, it does make sense that since effective altruists focus on lots of other things, maybe global health has been centered a bit too much in public presentations.

Ben Todd: Yeah. As we kind of gave the evidence for that at the start, which is that most people think that’s literally what the definition of effective altruism is. Well, many people do. So that, yeah, those are some thoughts on how we might get better at explaining effective altruism and I’m keen to hear more ideas on how we can do this better. Yeah, we’ve covered some of the arguments for effective altruism, what I see as the key ideas, and we’ve also through the course of the discussion, mentioned some factors that could be objections to effective altruism being really important. I think there’s a bunch of other objections that could be really interesting to have more work done into them and I’d be keen to see more research done. One that we’ve covered before on the podcast is cluelessness and how important that is. That’s a bit more of a general objection maybe to effective altruism because that’s almost saying that we don’t know how much good any actions do.

Arden Koehler: Yeah. Cluelessness is the idea that we’re just kind of clueless in the dark about what in fact has the best consequences, even though some things really do have better consequences than others.

Ben Todd: Yeah. Just a straight denial of the second premise. There’s a bunch of things about are there other moral values that are really important and how might they compare to the common good as we’ve defined it? And yeah, well this isn’t really an objection to effective altruism directly, but it’s the more you care about things that aren’t the common good, the less interesting effective altruism is. I think that’s another kind of area to respond… And yeah and I think there’s lots of others. And the reason we’re doing this episode is because when people do write critiques of effective altruism, I would love it if they were actually responding to one of the premises or something about overall, how big are the differences and impact between actions rather than just responding to a kind of particular suggestion like AI safety or donating to Against Malaria Foundation or earning to give, which are not the core idea of effective altruism, they’re just kind of ideas that people have had.

Arden Koehler: Although to be fair, it is useful when people object to those particular things because that helps contribute to the conversation.

Ben Todd: Yeah, that is effective altruism. That’s part of what I called earlier, the intellectual project of effective altruism. And it’s just, yeah, it’s a little bit annoying when people are like, “No, effective altruism is wrong, but then they’re actually doing effective altruism”.

Arden Koehler: They’re helping without even knowing.

Ben Todd: Yeah. To be more charitable to that, if you could kind of show that all the things that effective altruists have come up with are bad, then that would suggest that this isn’t a very useful project kind of indirectly.

Arden Koehler: Right. It might suggest that the second premise is false or probably false, “Okay, well theoretically we should be able to figure out what’s the best thing to do, but we keep failing so maybe in fact we can’t”. Yeah. You mentioned spread of impact being one thing people might push back on. What are some critiques you’re interested in there?

Ben Todd: Well, it’s almost not critiques, but if you just want to fundamentally research how important is effective altruism? One way of seeing the question is just how big are the differences in impacts between actions? And I would kind of break that down into how much more pressing are some causes or global problems than others? How much more effective are some interventions than others? And then kind of how big a difference is in personal fit between people? Those are three big topics, and it’d be really cool to see much more research into each of those three questions. And yeah, that research would have to grapple with many of the issues that we’ve brought up in this episode, like regression to the mean and epistemic humility and lots of other factors.

Arden Koehler: Yeah. And I guess that would be called global priorities research. At least it’s a subset of it.

Ben Todd: Yeah. It’s interesting that people do, I think, have a lot of interesting thoughts about how big these differences are and there is some work on it, but given that, in a way, that’s the most important idea, there’s an unfortunate lack of real published research on it directly.

Arden Koehler: Yeah. No, it is kind of strange that it does seem like there isn’t that much research that’s directly trying to address the question of how big are the differences in consequences between actions. Because it does feel like such a fundamental issue.

Ben Todd: Yeah. I think there’s almost a kind of thing where effective altruism either kind of seems obvious to people in which case that research is not that interesting or it just seems wrong, but in some really other different way. And so then, again, that research isn’t interesting. And so it kind of ends up being a bit neglected and maybe just to kind of zoom back to earlier, even if just one path open to you has twice as much impact as another, that might still be a really morally important thing. And so I wouldn’t want to kind of make it seem too much like, “Oh, it’s really, really uncertain whether effective altruism is a good idea”, but then I do still think it is very interesting whether the differences are a factor of three or a factor of 300. That makes a big difference to just how important effective altruism is.

Arden Koehler: Yeah. And I guess one kind of cool thing about that research is that it often serves sort of a dual role. One, it helps you figure out how important effective altruism is, but then you probably do that by looking at actual causes and in doing so you add something to the picture of which sort of issues should people be actually focusing on. People might be able to kill two birds with one stone.

Ben Todd: Yeah, totally. Though I think there are also a lot of interesting cross-cutting things like some of the things we brought up in this.

Arden Koehler: Yeah. The more abstract stuff like regression to the mean and stuff like that. Is there any last thing you want to leave the audience with, Ben?

Ben Todd: No, just to say, I hope people will run with trying to flesh out some of the things we’ve brought up and do more research into these questions.

Arden Koehler: All right. Thank you very much.

Rob’s outro [01:23:06]

That’s the second installment in a series of chats we’ll be releasing between Ben and Arden over the coming months.

Over the last week I’ve been enjoying reading your many responses to our user survey including a lot of commentary on this show.

We’re busy evaluating how the last twelve months at 80,000 Hours have gone, but don’t worry — we’ve still got new normal episodes in the pipeline for you, including one with Hilary Greaves and another with Russ Roberts.

If you’d like further reading on the topics in this episode, you could check out our recent article on Misconceptions about effective altruism, or Will MacAskill’s recent paper The Definition of Effective Altruism, both of which we’ll link to in the show notes.

And if you’d like to hear from Tyler Cowen himself on his views, I suggest checking out episode number 45 – Tyler Cowen’s stubborn attachments to maximising economic growth, making civilization more stable & respecting human rights

The 80,000 Hours Podcast is produced by Keiran Harris.

Audio mastering by Ben Cordell.

Full transcripts are available on our site and made by Zakee Ulhaq.

Thanks for joining, talk to you again soon.

About the show

The 80,000 Hours Podcast features unusually in-depth conversations about the world’s most pressing problems and how you can use your career to solve them. We invite guests pursuing a wide range of career paths - from academics and activists to entrepreneurs and policymakers - to analyse the case for working on different issues, and provide concrete ways to help.

The 80,000 Hours Podcast is produced and edited by Keiran Harris. Get in touch with feedback or guest suggestions by emailing [email protected]

Subscribe by searching for 80,000 Hours wherever you get podcasts, or click one of the buttons below:

If you're new, see the podcast homepage for ideas on where to start, or browse our full episode archive.