Keiran’s intro [00:00:00]
Keiran Harris: Welcome to 80k After Hours. I’m Keiran Harris — producer of the show, and the star of Home Alone 9.
This is probably our most niche After Hours episode to date — all about the experience of running an effective altruism university group, the impact such groups can have, and the challenges involved in making them good.
So if you’re not already fairly involved or interested in the effective altruism world this might not be for you — but 80k After Hours is a place where we’re comfortable producing content for smaller audiences, and hopefully we’ll have something up your alley soon.
In the outro I’ll have an announcement about the fact that we’re looking for another audio editor to join our podcast team, so stick around or skip to that if it might be relevant to you.
Alright, here’s Rob and Kuhan.
Kuhan’s journey [00:00:47]
Rob Wiblin: Today, I’m speaking with Kuhan Jeyapragasan, someone I’ve regularly been asked to interview over the last year or two. From 2015 to 2020, Kuhan did an undergrad and then a master’s in maths and computer science at Stanford. And during the six years they were there, they did a lot to organise and improve the effective altruism group on campus, something a few listeners have been keen to hear their thoughts on. So welcome to the show, Kuhan.
Kuhan Jeyapragasan: Thank you so much for having me. Super excited to be on the podcast.
Rob Wiblin: Is it possible for you to put maybe some meat on the bones of that little intro there? What have you generally been up to since 2015?
Kuhan Jeyapragasan: Yeah. Wow. 2015. So I started out at Stanford as a pre-med student, planning on becoming a doctor. And throughout my first year, where I was largely taking pre-med classes, I had this nagging suspicion that being a doctor wasn’t the most impactful way to help people, and had a mini quarter-life crisis over the summer after my first year. I still had no idea what I wanted to major in — since in the US you get to choose your major while you’re in university — but had no idea what I wanted to do with my career. I thought, “I should probably have a better sense of what I want to do after graduating to decide how I spend the next three or more years of my education.” So, I actually Googled “most impactful careers,” or something along those lines, and found 80,000 Hours.
Rob Wiblin: Fantastic.
Kuhan Jeyapragasan: I remember when I first opened up 80,000 Hours, I was like, “This is exactly what I’ve been looking for. This is incredible.”
Rob Wiblin: The system works.
Kuhan Jeyapragasan: Yeah. The SEO really did its job. So I learned about effective altruism, and really liked the ideas around moral circle expansion; being really consistent and rigorous in how to apply these intuitions I had around, say, every human’s life being of equal value; caring about sentience more broadly; and wanting to be more rigorous about how I applied my moral intuitions to my life. Then I started going to a few Stanford EA meetings early on, but very regrettably, I unfortunately didn’t really get more involved for a good three-ish years for a few reasons.
Kuhan Jeyapragasan: So one was, I think frankly, Stanford EA at the time was not the most intro-friendly group. I think there were some really cool discussions going on…
Rob Wiblin: I had visited the Stanford EA group a handful of times, and they were very intelligent, and really on the cutting edge in terms of the ideas that they would chat about at typical meetings. I think it was not the median university group.
Kuhan Jeyapragasan: Yeah, for sure. And then I had just read the 80,000 Hours career guide at the time. So you know, no LessWrong exposure, hadn’t read Poor Economics, didn’t know what an RCT was. I think I could largely follow along with the discussions, but I did feel pretty out of place.
Kuhan Jeyapragasan: And perhaps more importantly — while I thought the discussions were really intellectually stimulating, and it was quite clear that the people in these discussions cared a lot about being really rigorous in their thinking and cared a lot about impact as someone who was quite new — I think it didn’t feel like going to meetings was helping me figure out how I personally could do the most good. And also I did feel a bit just intimidated by how much I didn’t know and how quickly the conversations were moving. And as a result, I think the meetings were Sunday afternoons, so I would often have homework due on Monday and just kept having excuses for not going to meetings.
Kuhan Jeyapragasan: But this all changed in 2019, when I was luckily invited to a North American group organisers’ retreat run by the Centre for Effective Altruism. I actually was only able to go because another Stanford EA member couldn’t attend, so I got pretty lucky. And from there, I learned about the Community Building Grants Program that CEA was running and thought, “Oh wow, this sounds like a way better use of my time than the teaching assistantship I had planned for my master’s degree.” So I applied to the CBG Program, and got a grant to do 20 hours a week of organising for Stanford EA. And that’s what actually got me into community building and taking EA really seriously.
Rob Wiblin: I see. And that was around 2019?
Kuhan Jeyapragasan: Yeah. So my grant started in September 2019, but I think I was approved for that back in April or so.
Rob Wiblin: Right. OK, we’ll return to some of those themes of the challenges of making a local group appealing and accepting of everyone. It’s definitely difficult to have a culture that doesn’t bounce anyone, so it’s good to try to be relatively expansive. What concrete stuff were you doing in 2020, 2021, and up until now in terms of trying to grow the Stanford EA group?
Kuhan Jeyapragasan: I started officially running the group around September 2019, but I think even before that I had been doing a lot of planning over the summer, which actually involved a lot of just reading about cause prioritisation, longtermism, existential risk, community-building strategy from resources that other organisers had written up, talking to successful EA group organisers in the past — so like past Stanford EA organisers, Oxford, Cambridge, et cetera.
Kuhan Jeyapragasan: And then once the school year started, lots of individual messages to people, emails, Facebook posts, lots of advertising — I wanted to make sure that anyone who could plausibly be interested in EA with a one-sentence description of it would know about the group, and know how they could learn more and how to get more involved. I remember we had a big talk from Will MacAskill that happened to be on the same day that there was this societies fair for service-oriented groups. So we were aggressively advertising the Will MacAskill talk that was happening later that day, and would make flyers and cardboard signs and all that stuff.
Rob Wiblin: What drew you to getting involved? It sounds like you had some reservations about the local group, or at least you didn’t feel like you quite fit in. But nonetheless, you decided to become president of it and take it over. What pushed you to do that?
Kuhan Jeyapragasan: I remember even back in 2016, right after reading the 80,000 Hours career profile, I remember thinking, “Wow, these ideas are so compelling. Why don’t more people know about this? So many people I imagined would be as excited as I am if they just saw these arguments.” So I think the intuition around wanting to share these ideas, because they seem so compelling and correct — or very compelling based on my moral intuitions; I had that intuition from an early age.
Kuhan Jeyapragasan: And when I first encountered the arguments, I did talk to lots of friends about EA, even though I wasn’t involved with the group as much. I think the idea around the multiplier effect argument really stuck with me: we may only have approximately 80,000 hours in our career, but it probably takes way, way less than that to convince others to counterfactually switch their career plans to something much higher impact or directly addressing the world’s most pressing problems.
Kuhan Jeyapragasan: Then also I was just thinking about what a unique situation university is for getting career changes. Like where and when else do you see such a large concentration of really caring, dedicated, driven, talented, ambitious people who are figuring out their values, and are actively trying to figure out what they want to do with the rest of their life? And they’re pretty open-minded and flexible to making pretty big changes, or don’t even have plans yet, so are much more open to making tentative plans — they’re really starved for good advice to help them find some way that they can spend over half of their waking hours after graduating on something that’s both exciting to them and intellectually fulfilling, but also helps them achieve their goals and values. I think for many of them, impact and doing good is a pretty strong motivation.
Rob Wiblin: Yeah. How straightforward did you find it to get people excited about effective altruist ideas at Stanford? Were there a lot of people who were immediately engaged once they heard about it, or was it a slightly difficult sell?
Kuhan Jeyapragasan: I think many people would agree that doing more good is better than doing less good, and I think a lot of intuitions that effective altruism is built off of are fairly uncontroversial, at least to many people. I think there are tradeoffs that different people feel more inclined to make or different intuitions that feel stronger to certain people than others.
Kuhan Jeyapragasan: So I found, for me, really valuing consistency and rigour in thinking and argumentation, and taking arguments to their logical conclusions, and acting accordingly — I think these are values that, for whatever reason, differentially motivate different people, or feel a lot more important to some people than others.
Kuhan Jeyapragasan: There are also other things around how far do you go with certain moral intuitions or logical arguments. Like different people are much more willing to bite the bullet for, say, the Repugnant Conclusion, or how far we extend this caring about wellbeing in sentience thing when we’re thinking about expansive altruism. I think openness to ideas definitely seems pretty important, and depending on how strong your anti-weirdness or anti-what’s-considered-normal prior is, definitely adoption of various ideas can be pretty difficult.
Kuhan Jeyapragasan: And then there are also other things, like how easy it is to, say, not pursue the highest-paying career option, depending on your financial background or family circumstances. Or other considerations, like are you an American or an international student? There are all these other things.
Rob Wiblin: Yeah, how much flexibility do you have in general.
Kuhan Jeyapragasan: Exactly. Yeah.
Rob Wiblin: I imagine when you got involved and actually started organising things, there were probably some aspects of the Stanford EA group that you wanted to shift or improve. What’s one of those that you had in mind?
Kuhan Jeyapragasan: Yeah. I will say that there were lots of aspects of the old Stanford EA group that felt really, really important that I wanted to maintain, especially around rigour in thinking and discussion, and having a culture of respectful disagreement like Aumann’s agreement theorem: when you have two rational agents with similar values or something, and there’s a disagreement, you should be able to resolve that disagreement and come down to some difference in observations or beliefs or maybe values or priors. So this culture of, if we’re two rational agents with the same goals, it’s interesting if we have very different ideas about an answer to a question like “How to do the most good?” or “Which causes seem most important to work on?” So that culture of respectful disagreement I really liked.
Kuhan Jeyapragasan: But also I think a thing I really wanted to emphasise — largely based on my initial experience with Stanford EA — was thinking really hard about how I wanted people who are involved with Stanford EA to be thinking really carefully about how they can do the most good with whatever resources they want to give towards doing that. And thinking really hard about how to make involvement with the Stanford EA group help people make progress on their career plans, hopefully make some long-lasting friends, raise ambition among members, work together on increasing our productivity levels and our ability to actually get things done and achieve our goals, improve our thinking, and actually make progress on our plans to do the most good.
Kuhan Jeyapragasan: So I wanted to put a lot more emphasis on career planning sessions, figuring out our own cause prioritisation, why we think what’s most important is most important, what the implications of these beliefs are — and then how to apply these to how we spend our time. So figuring out summer plans, figuring out career plan options, figuring out good people both inside and outside the community to talk to, what opportunities to pursue, and then also creating a really fun social environment to be a part of.
‘Bad Omens in Current Community Building’ [00:13:00]
Rob Wiblin: Since there have been local groups, there’s been disagreement between people about what kinds of events they should be running, what sort of culture they ought to be cultivating. So there’s all kinds of different topics that we could tackle there. I was keen to bring up some that recently appeared on the EA Forum a couple of weeks ago in this blog post, “Bad omens in current community building.” It was mostly positive, saying that local groups are great in a lot of ways and it’s important to get them right. But it suggested that local groups have perhaps gone too far in some particular directions, and could be turning off some people who should be interested in effective altruism if they really understood it properly and it was getting presented in a good light.
Rob Wiblin: One concern they had is that a lot of local groups measure their success in terms of how many heavily involved or highly engaged people who identify as part of the effective altruism community they’re causing to exist by informing other students about these ideas and encouraging them to come along to events. When you have that as your key metric, then you tend to get into this very persuasiveness mode where you’re just like, “All right, we got to get people into workshops so we’re going to hit them with a message; we’re going to figure out what is most likely to convert people” — and you start thinking about it in terms of conversion rather than sharing information.
Rob Wiblin: Do you have any thoughts on this general issue of downsides you might get from the idea of just persuading people and “creating EAs,” as some people will call it?
Kuhan Jeyapragasan: Yeah. I think I resonated with a bunch of the ideas in the post. I think having really strong epistemics is one of the integral parts of running a really great EA group or related group. One thought is, man, it’s really hard to run an EA group while being a full-time student, perhaps being involved with other extracurriculars, maybe having jobs on the side to help pay for tuition or other expenses. I think because of that, and maybe other intuitions around how it’s just a lot more exciting when your group has lots of members.
Kuhan Jeyapragasan: These problems seem extremely important, and for many of them there are just so few people working on them — or at the very least far fewer than the ideal number — often due to simple reasons like lack of awareness of these ideas, or perhaps the ideas not being presented in the most compelling way. I think the intuition that we should get as many people as possible is very understandable.
Rob Wiblin: Totally.
Kuhan Jeyapragasan: But at the same time, I think it’s really important to have good epistemics and show how rigorous thinking in the community can be, and why it’s so important to have really rigorous thinking, given how complex the problems we’re trying to solve are. If they weren’t so complex, it’s a lot more likely that they would’ve been solved already. So I definitely agree that it’s really important to both do your research, think through arguments, steelman counterarguments; and think of what you really think, what the best responses are, how uncertain you are about various claims, and ways to reduce your uncertainty.
Kuhan Jeyapragasan: It’s also just really good to be honest with people you’re talking to about how much you know, how confident you are, what your uncertainties are. Showing that you’re a really thoughtful, caring person, and that the other community members around you and your group are as well, I think is — for the people who could really resonate with the EA ideas and take these ideas really seriously and contribute a lot to the world’s most pressing problems — just a lot more compelling than something that’s pretty clearly a sales pitch or something that isn’t trying to be rigorous or isn’t properly caveated or qualified.
Rob Wiblin: Reading that post, it occurred to me that many people might think that promoting effective altruist ideas on campus is a pretty accessible thing to do, which in a sense it is. But if you want to go and talk to really bright people about these very complex problems and very challenging conceptual issues that come up, as soon as you start trying to prioritise the world’s problems, in order to do that well, you have to not only just be able to parrot things that you’ve heard other people say, but you have to understand them well enough to present them faithfully.
Rob Wiblin: Then when people push back and they have objections — so they’re like, “What about this empirical fact that I know that you don’t know?” — either just be happy saying, “Look, to be honest, I don’t know. I’m going to have to ask someone else about this because I’m kind of new to this whole thing as well,” or be sharp enough and knowledgeable enough in order to provide a useful response that will either address people’s objections or point them in the right direction in order to learn more. Which is easier said than done.
Kuhan Jeyapragasan: For sure, yeah.
Rob Wiblin: So I can totally understand — especially with student groups, where many of the members are going to be undergrads churning over pretty quickly; a quarter or a third of them leave every year — many people are going to be fairly new. It’s very natural to repeat things that you’ve heard without necessarily having invested years yet into researching them yourself and independently deciding which parts you agree with and which parts you don’t. Do you have any thoughts on that?
Kuhan Jeyapragasan: Yeah. As I said earlier, really having time to think through these arguments, very rigorously reading all the relevant material, thinking through all the counterarguments, all the things I mentioned earlier — that’s just super time consuming. And it’s very understandable that as a student who’s trying to figure out what you want to do with your career — which I imagine for most EA group members, isn’t continuing to do university community building — there’s always more you can read or learn or talk to people about.
Kuhan Jeyapragasan: So I definitely sympathise with students who are like, “Man, you’re telling me I should think through all the arguments so that I can speak well to people who might bring up reasonable objections that people with a background in the problems we’re hoping to get more people to work on might have?” But what can you do?
Rob Wiblin: We’re all muddling through most of the time, yeah.
Kuhan Jeyapragasan: Yeah. As I said earlier, showing that you’re really thoughtful, care a lot, and care a lot about doing what’s right or what seems best given the information available to you and our limited resources and cognitive abilities and whatnot. I think I’m much more of a fan, at least for EA groups particularly, to have the mindset of, “I want group members to be thinking really rigorously about how they can do the most good, and then actually do it,” rather than, “My goal is to get people to work on X, and I’m going to try to say whatever things maximise the likelihood of people working on X.”
Kuhan Jeyapragasan: But that being said, for many of the problems we’re trying to solve, it’s not like the only people who can make progress on these problems are people who are super bought into the EA thing or longtermism thing or what have you. There are lots of reasons to work on these issues. Like in The Precipice, in the second chapter, Toby Ord talks about a bunch of different reasons why people might want to work on existential risk reduction, as one example. It’s not like everyone needs to agree with these 10 different important core premises or something to do lots of good.
Rob Wiblin: Totally. Yeah. Another issue that article on the EA Forum raised was that some people — and in the view of this author, potentially some of the best people — get put off by a style of advocating EA that is about persuasion rather than just exploring ideas and informing people of stuff that they might find useful. Do you have any thoughts on the balance between explicitly making a case and arguing for what you believe versus just exploring ideas together?
Kuhan Jeyapragasan: One thought I have is that there are definitely ways to present arguments in a more compelling way. And it’s, I think, a sad fact about how humans work that how true or logical an argument is is only part — and maybe sometimes a small part — of what leads us to actually update our beliefs or change our minds and change our behaviour accordingly. So I definitely sympathise with the idea that there are ways to present ideas and arguments in a way that’s more likely to get people to seriously engage with them. For example, I think it’s a lot easier to believe an argument on, say, an emotional level — even if you maybe agree with it on an intellectual level, when it’s not super costly to act in accordance with the logical implications of the argument and there are clear next steps or actions to take that actually seem exciting.
Kuhan Jeyapragasan: But that being said, I definitely resonate with the idea that when you’re trying to make a sales pitch, we have pretty good gut intuitions around, “I feel like someone’s trying to trick me or not let me actually think through what I believe and why.” I think that that’s definitely bad.
Rob Wiblin: And if I raise an objection, they just do some slick move where they move onto the next point.
Kuhan Jeyapragasan: Right, exactly. No, it’s pretty easy to catch that or sense that something is wrong.
Rob Wiblin: Yeah, totally. One way that we hope to address this at 80,000 Hours is interviewing people who are from outside 80,000 Hours who know more about their area of specialty than we do. I wonder whether it might be better for student groups sometimes to do more guest speakers who actually have much more knowledge maybe than anyone in the student group does about their particular problem that they’re working on, and could seriously address objections some people might raise in the audience.
Kuhan Jeyapragasan: I think that would be great. Another thing that has worked well for me, and in conversations with people who know much more than me about certain topics that are relevant to what we’re discussing, is saying, “I have uncertainty here. I don’t really know what’s best. Here are arguments that I’ve heard, or here are thoughts I have that might be pointers to check out more resources on a topic or to think through ideas more.” Rather than saying, “Oh, I think you’re wrong. And even though I can’t really explain why, trust me: people who have thought about it more are pretty sure.”
Kuhan Jeyapragasan: I feel pretty optimistic about getting people excited to learn more or even just entertaining the possibility that this claim does seem pretty wild, or big if true — but it seems important to figure out if it is true, because if so, that would have pretty large or wild implications.
Lessons from Stanford EA [00:22:35]
Rob Wiblin: Yeah. What’s a lesson or two you learned from being involved with Stanford EA for a few years?
Kuhan Jeyapragasan: Maybe the most important lesson or concept that feels a lot more salient to me now is that I remember reading about the idea of heroic responsibility and really resonating with it. The idea that when I really care about something, there’s no excuse: when it comes to the thing I care about happening or not happening, trying really hard, or assuming that other more qualified people have things handled, or telling others about the problem or something isn’t an excuse when I really care about something happening or care about the outcomes.
Kuhan Jeyapragasan: I think our intuition that we only have limited responsibility for failure or for not achieving our goals comes from how that’s how things usually work in society. Our teachers and parents and other people usually just hold us responsible to a limited degree, and we’re often not blamed for some failure if we tried. But I think taking heroic responsibility is about remembering that we don’t just care about avoiding social blame for a bad outcome: we actually care about preventing the bad outcome. And when that’s the thing you care about, just doing your part or what’s plausibly reasonable to expect of you isn’t good enough. The thing that matters is the thing you care about actually.
Rob Wiblin: So it’s the idea that a lot of the time, we potentially go through the motions or something; we do whatever is reasonable in other people’s eyes in order to try to accomplish some goal. Because maybe we don’t care about the goal that much in itself, we just care about being seen to do what’s reasonable. But if you actually care about getting a particular outcome, sometimes that requires you to go beyond what other people would judge is reasonable and just to do whatever is going to work in that specific case. It strikes me as a little bit stressful of an attitude to have. I’m not sure that I’d take that attitude to everything in my life. Sometimes I just want to do what feels safe and reasonable.
Kuhan Jeyapragasan: It is a pretty intense mindset, and not one I embody all the time to be sure. But as an example of how this applied to university group organising, as I got more involved with community building, I learned that lots of things that seemed pretty clearly good to do just weren’t being done, and that this was often due to just a lack of capacity and the number of people working on these issues full-time and really devoting a lot of attention to it being surprisingly low. At least back when I started out; I think things have expanded very rapidly, which is both really exciting and comes with some different issues and growing pains, as were just discussed in that post.
Kuhan Jeyapragasan: But I found it pretty empowering to see that despite not really having much of an organiser or community-building background, just reading a bunch, talking to people, and just thinking hard about how I could provide value and what things seemed most impactful to do — and actually putting effort into executing on these ideas and getting others around me excited about doing the same — could go a really long way.
Kuhan Jeyapragasan: I think it’s pretty wild that a bunch of students who had only really been involved with the community for around a year were, within less than a year, organising one of the few programmes on existential risk through the Stanford Existential Risk Initiative, and one of the largest internship programmes a year later. So I think just caring a lot, being rigorous, and trying hard can get you pretty far.
Rob Wiblin: Yeah, I guess it makes sense when something is run on a volunteer basis or who-shows-up basis, that determines what gets done — having a sense of, “Well, if I don’t do it, it’s probably not going to happen. So if I want it to happen, then I just have to put in the hours and I need to become the champion for that project.” And that’s a really useful attitude to have when you are involved in a student group.
Kuhan Jeyapragasan: The second most important lesson, at least for me, was realising the salience of opportunity cost. What isn’t happening or could be happening in an ideal world isn’t clear, and you don’t really get much feedback on, “If you had just thought a bit harder or been more strategic, things could be 10 times better” or something — we don’t really get that kind of feedback from the world often. Obviously if things could be better, it’s way easier to think of concrete mistakes you make when you actually execute something than it is to think about strategic mistakes or how much better things could be.
Kuhan Jeyapragasan: So one way I try to get around this is to regularly brainstorm lots of ideas for projects, and then see if any of them might actually be useful and do some prioritising. It’s this regular process of brainstorming how things could be better, what considerations I might be missing, and that sort of thing that led to the Stanford Existential Risks Conference, which we’ve run for the last two years, and also noticing opportunities for perhaps outsized impact.
Kuhan Jeyapragasan: With university group organising, I realised the beginning of the year just seems extremely important — way more important than the rest of the year, since that’s when people are deciding what clubs they want to get involved with and what their time commitments are going to be. That’s often when a lot of their strongest friendships form, that’ll last for their whole time during university; that’s when classes aren’t particularly busy yet, and people actually have time to explore and consider different options. And noticing this, I also realised that Stanford actually starts three weeks later than most US universities, so a bunch of Stanford EA organisers went to a few different US universities to help out with their beginning-of-year outreach and programming, which I think I actually went quite well.
Kuhan Jeyapragasan: So just always thinking of opportunities for outsized impact and then actually acting on them, and considering a wide set of possibilities and being open to doing maybe seemingly crazy things — like flying to the other side of the country for a few weeks and helping out with tabling or advertising the EA group or running intro presentations or retreats — just that out-of-the-box thinking I think has been really helpful.
Rob Wiblin: What’s one of the projects that you took on while you were involved in Stanford EA that you think had really good outcomes? Can you flesh out what it involved?
Kuhan Jeyapragasan: Yeah, one example of this is what we called it the “Stanford EA residency programme” — it was kind of an unofficial thing with a bunch of students that could have been much better planned, but I think was a great idea that can be iterated on for future years. Another example was that over the pandemic, realising that since we’re going to have to run all of our programming online anyway, this seems like a really unique opportunity to open up our programming to a much broader audience.
Kuhan Jeyapragasan: So basically, even our first iteration of our Intro to EA Program or fellowship we opened up globally. And I think that was pretty successful. We had attendees from Brazil and the Philippines and the United Arab Emirates and all over the world, which was really exciting.
Rob Wiblin: Nice.
Kuhan Jeyapragasan: We had students from universities or cities or countries without a local EA group being able to participate. And doing this was at least somewhat instrumental in that kind of evolving into what’s now the Effective Altruism Virtual Programs. I think Oxford EA also did a similar thing with their in-depth programme, opening that up globally and going quite well. So just noticing how to make the most of a very unfortunate situation with COVID, and using that as an opportunity to make EA content more accessible to a broader audience is yet another example of a project that felt pretty cool.
Kuhan Jeyapragasan: We had 300 participants in our intro fellowship programme one term during the pandemic. That was pretty chaotic to organise and get enough facilitators for, but it was also pretty cool at the same time. And I think it got me more into the mindset of thinking more ambitiously about how to scale up programming I’m doing, how to help it reach a wider audience, how to make sure that the lessons I’m learning from running Stanford EA can extend to university groups and local groups more broadly. Just trying to always think about how we can maximise impact better.
Rob Wiblin: Yeah, nice. I get the sense that it might be a little bit easier for local groups now than it was in, say, 2015. Because back in 2015, people had the impression I would say that it felt like we were just kind of saying the same thing again and again and again. I think the reason was just that effective altruism as a community — and especially as a research community — was tiny back then. And so the amount of people actually generating new content and novel ideas that we’d never discussed before, you could count them on one hand. So the talks tended to remain the same year to year, and it was maybe hard to find exciting new things to say.
Rob Wiblin: But these days, the Effective Altruism Forum has so much content, and there’s so many more people doing research into specific problems and specific solutions, that it feels like every month there’s exciting new ideas coming out and maybe that’s helpful.
Kuhan Jeyapragasan: Oh, for sure. Yeah, I think there are just way more resources now and a lot more infrastructure to support people getting involved with the community, and having concrete opportunities to engage with these ideas full time over an internship, or getting a grant to do some self studying or skilling up, or the number of internships that are available in organisations that you can join, and resources for careers outside of EA organisations — like in policy or technical research or community building, entrepreneurship, all the things.
Kuhan Jeyapragasan: Also the development of EA hubs, and these offices and coworking spaces that now exist in a bunch of EA hubs — like groups of friends getting together to live together or at least meet regularly. The number of opportunities to get more integrated and involved — and form close friendships and mentorships and partnerships and potential collaborations — is just a lot more accessible now.
Rob Wiblin: Makes things easier as an organiser.
Kuhan Jeyapragasan: Yeah, for sure.
Rob Wiblin: Speaking of things that might make things a bit easier these days, it sounds like the Centre for Effective Altruism is now offering some training to people who are involved in organising local groups and maybe some other resources as well. What kinds of things do they have?
Kuhan Jeyapragasan: Yeah. One programme that I’m pretty excited about is the University Group Accelerator Program that Jessica McCurdy is running. The programme takes new organisers, or organisers for newer or younger groups, and helps them run an intro to EA fellowship programme; and offers advice on intro to EA events, how to run one-on-one conversations or intro to EA coffee chats well. And there are also a bunch of other resources that are on the EA Forum. The CEA Groups team is also creating a new website; we can find the EA Groups Resources page and maybe attach it to the transcript later.
Kuhan Jeyapragasan: I think there’s some other exciting opportunities as well. So the Global Challenges Project has created a university group organiser handbook with a bunch of resources on programmes to run, with all the resources you need to advertise the programme and run it. And recommended reading for people who are interested in a bunch of topics that you might expect group members to be interested in or interested students to want to learn more about. There are a bunch of resources there.
Kuhan Jeyapragasan: Then I think a bunch of community members and group organisers are also very happy to give their time to new organisers — to offer advice, lessons they’ve learned, resources they’ve found particularly useful; and to share what’s worked well and what hasn’t worked well, and what doesn’t exist but could for people who are interested in taking on projects.
Rob Wiblin: Yeah. Do you think there’s room for a lot more people to go into local group organising? On the one hand, I’ve heard that there’s been so much excitement about this over the last years that maybe people have crowded into it a little bit too much, and they’ve gotten too excited about this versus contributing in other ways.
Rob Wiblin: On the other hand, I saw this survey last week, which suggested that only 7.5% of NYU students had ever heard of effective altruism at all. Most of the people who haven’t heard about it wouldn’t be interested, but nonetheless, it might be good to reach that other 92.5% so that they can decide whether it’s for them or not. That suggests that there is a lot of low-hanging fruit left, potentially. What do you reckon?
Kuhan Jeyapragasan: So one thing I’ve been thinking recently is there are lots of meta EA and community-building efforts outside of university and local group organising that I’d be excited to see way more people in. Cause-area-specific field building and community building I think could be really good, so creating resources for people interested in getting up to speed on the most important considerations and research for specific cause areas.
Kuhan Jeyapragasan: For example, Cambridge is running some global programmes like the AGI Safety Fundamentals programme for people to get a pretty strong foundation in a lot of main considerations for working on technical AI safety research, and I think a governance programme was recently released as well that SERI organisers helped create. So more programmes like that I think could be really good, and more infrastructure for mentorship and connections to form for people who are interested in skilling up in a particular cause area and learning about, say, summer opportunities, early-career opportunities, how to learn more, applying for funding to skill up.
Kuhan Jeyapragasan: I think doing independent research can be perhaps a lot easier to do or more motivating to do with peers who are in a similar position. So finding these people who are also interested in similar things and maybe working together with them, and maybe even living together with them. I’ve found living with fellow Stanford EA organisers over the pandemic was really motivating for me — just the conversations around dinner being really good and leading to lots of idea generation.
Kuhan Jeyapragasan: There are definitely tonnes of things outside of just local group and university group organising that seem really exciting. And I think things that can scale well seem quite exciting — like resources that can be relevant for all group organisers or for all people interested in a particular career path — I’d be excited to see a lot more focus on that.
Rob Wiblin: Some people react negatively to hearing about people who are enthusiastic about effective altruism, even undergrads, all living together in group houses and so on. I definitely get worried when I hear about people closing their friendship circle, so they don’t have friends who are interested in other topics. But I don’t think that we invented “birds of a feather flock together” among the undergraduate crowd. I remember when I was a student, all kinds of different people would meet like-minded folks and they’d want to hang out all the time at college. That’s actually one of the great things about an undergrad experience.
Kuhan Jeyapragasan: Oh yeah, totally. I mean, many of my friends, maybe even most of my friends, are living with other friends now. I think that’s becoming increasingly common as a living situation for recent graduates, especially for people who want to live in cities that are quite expensive, and it doesn’t seem unique to the EA community, at least from what I can tell.
Rob Wiblin: What’s one of the most common reasons for someone who should or could easily get into effective altruism to kind of bounce off of a local group and lose interest?
Kuhan Jeyapragasan: I think the EA Forum post you mentioned covered some pretty good considerations. Other potential reasons: I think it’s really valuable as a group organiser to think, “How can I make group members’ experiences of being involved with this group valuable?” And “valuable” can mean a bunch of different things. It can mean making progress on your career plans. It can mean making progress on your biggest uncertainties relating to how you want to have an impact. It can be socially valuable, forming really great friendships and having really great conversations and having fun together. Not feeling valuable is definitely one of the main ways that people just stop coming to events.
Rob Wiblin: Not quite worth the time.
Kuhan Jeyapragasan: Not worth their time, exactly. I think a mistake I made with Stanford EA early on — and I think I still haven’t maybe fully corrected, but trying to figure out how to find the right balance here — is that I think I was really compelled by the arguments around “These problems are so, so important and we really need really dedicated, sharp, talented people working on these problems.” I think I kind of tried to make Stanford EA this “factory of impact” — where we’re just running all these really big programmes and motivating people to the importance of these problems we’re trying to work on as seriously as they deserve, and giving them the respect they deserve.
Kuhan Jeyapragasan: As a result, I maybe leaned too hard into doing as much outreach as we can, running these really big programmes, spending as much time trying to introduce these ideas to people, helping them figure out their career plans, running socials, all these things. I think at a point I maybe — especially for the most involved members, the people who cared a lot about these ideas and cared a lot about doing the most good — was somewhat sacrificing what would’ve been most helpful for the most engaged members at the expense of doing more intro-level or top-of-the-funnel outreach, rather than prioritising helping our most engaged.
Rob Wiblin: Rewarding the people who are most engaged already.
Kuhan Jeyapragasan: Yeah. And given the nature of a lot of the problems we’re trying to solve, I think probably most of the impact will come from people who are focused and thinking really critically and rigorously about how to really tackle these problems.
Kuhan Jeyapragasan: A mistake I see both in myself and in other organisers is not taking into account how heavy-tailed impact might be, and acting accordingly — and not, say, back chaining from thinking about which problems seem most important to solve and how we’re actually going to solve them. Thinking about what the key bottlenecks are now and what they’re likely to be, and then determining what kinds of community-building work to put the most effort into, or what outcomes would, from a group-organising perspective, actually lead to the most progress on these problems.
Rob Wiblin: Yeah. If someone came to you and said, “I’ve decided to take a leadership role in my local EA group. What’s a mistake that you can help me avoid based on your experience?”, what might you tell them?
Kuhan Jeyapragasan: One thing that comes to mind is that it can be easy to get into the mindset of, “I’m running a student group. Therefore, I should do the standard things student groups do like X, Y, and Z” — rather than, as I mentioned before, doing this kind of back chaining from, “What problems am I trying to solve? What am I trying to achieve with this group?” And then thinking through, “If I’m trying to solve problem X, what are the bottlenecks? How can we best address these bottlenecks? What can community-building or group-organising work do to help?”
Kuhan Jeyapragasan: I also think maybe the ideal vision for an EA group that I have in mind is pretty different from how most student groups operate, at least as far as I’m aware. Lots of uncertainty, all the caveats, but I think the kind of group I’d be really excited to see has a very rigorous intellectual environment — where group members are improving each other’s thinking. And as I mentioned before, having this culture of respectful disagreement, and knowing that these disagreements or confusions — like, “Oh, that reasoning doesn’t seem to be consistent for X, Y, Z reasons” — is coming from a place of, we’re really trying to do the most good, and we’re all on the same page about that.
Kuhan Jeyapragasan: So creating this kind of rigorous intellectual environment. Having group members encouraging each other to become more productive and getting better at achieving our goals. Stanford EA, for example, has an accountability buddy system, which I think can be really great for setting up accountability mechanisms to do things — like, say, weekly reviews, or making a list of what your biggest bottlenecks are and actually implementing plans to address them. So I think systems like that could be really great. Brainstorming projects to execute on, potentially together with other group members, and becoming a tight-knit group of friends.
Kuhan Jeyapragasan: And of course doing lots of career planning, figuring out what to do over the summer and what opportunities to pursue while in school. Maybe considering taking a term off to explore potential career options. I think an environment that encourages this kind of out-of-the-box thinking and considering probably a wider set of actions than most students consider could be really valuable. And that seems like a big reason for a lot of the success that the Stanford EA group has had.
Rob Wiblin: OK, we’ll push on from local student groups in a second, but our last question is: What people in the audience should maybe consider this as a way to have impact? Getting involved in organising events and improving the community in their city or on their campus or whatever it is?
Kuhan Jeyapragasan: One of my biggest uncertainties, back when I was in my final year at university and deciding what to do next, was that community building felt like a very nebulous career or long term option. There were a few orgs — like 80,000 Hours, of course, and CEA and a few others — that were focused on community building. But the number of options for full-time work in this space felt pretty limited, or there was just a lot of uncertainty. Obviously a lot of this comes from the EA community itself being quite young; there aren’t too many examples of 60-year-old role models to —
Rob Wiblin: Yeah, I’m getting there, but not quite 60 yet.
Kuhan Jeyapragasan: But I do think the space is evolving quite rapidly and I’d be surprised if it didn’t continue to do so. The number of funding opportunities, opportunities to start new organisations or projects, and just the number of existing projects and opportunities to contribute full time, all keep growing.
Kuhan Jeyapragasan: I’d be happy to chat with people who are interested in considering community building as an option but don’t know where to start. I think writing on the EA Forum and LessWrong are some pretty great ways to have your ideas shared with a wider audience — and also to put yourself on the radar of other organisers and community members, which could lead to exciting opportunities like internships or collaborations, or maybe being reached out to about getting funding for ideas that you might pursue.
Kuhan Jeyapragasan: The infrastructure is a lot better right now for people who are interested in getting more involved, and I’d be really excited to see more people figure out a problem that isn’t currently being addressed optimally in the community and take agency. Obviously talk to lots of people and make sure that your idea is good and implementation plans make sense, but I definitely encourage people to take an agentic mindset of looking for ways to provide value and then making it happen.
Rob Wiblin: Yeah, nice. Where can people email you if they’d like to check in and see whether they have what it takes, or what you think of their plans?
Kuhan Jeyapragasan: My Stanford email is pretty good, so [email protected].
Stanford Existential Risks Initiative (SERI) [00:45:46]
Rob Wiblin: We can put that on the post associated with this episode. OK, let’s push on briefly to talk about the thing that you’re currently doing. So you’ve graduated from this master’s — you went from medicine to computer science and maths, a fairly big jump. And you finished your master’s and now you are working full time as the Program Director at the Stanford Existential Risk Initiative, is that right?
Kuhan Jeyapragasan: Yeah.
Rob Wiblin: What kind of programmes does SERI operate?
Kuhan Jeyapragasan: SERI is running a number of different programmes right now. We’ll be running a Summer Research Fellowship — this will be our third iteration, so that’ll be starting in a few weeks. That’s an opportunity for people interested in existential risk research — particularly risks from transformative artificial intelligence, synthetic biology and pandemic prevention, nuclear risk, extreme climate change, and miscellaneous other related work like field building, forecasting, et cetera — to do research for 10 weeks with a mentor on a topic relevant to existential risk mitigation.
Kuhan Jeyapragasan: We also run a virtual Existential Risks Conference. That’s been run for each of the past two years to provide a broader, more globally accessible educational and networking opportunity to people who are interested in learning about existential risk reduction — there are talks, workshops, a platform where people can message other attendees and set up meetings. We also try to provide resources for next steps, like how to learn more if you’re interested in careers in this space, checking out 80,000 Hours, we send people copies of The Precipice if they want a book on existential risks. So yeah, we run the conference.
Kuhan Jeyapragasan: We have collaborations with other relevant organisations. I mentioned earlier that some SERI members helped develop the curriculum and helped run the AI Safety Fundamentals programme that EA Cambridge ran. We’ve also been doing field-building work, helping seed relevant institutes or student groups. For example, I think the SERI conference last year was a big part of the inspiration for starting the Cambridge Existential Risk Initiative, so that was a really exciting outcome. Some collaborations from discussions at last year’s conference also led to a US Policy Careers Speaker Series. I think writeups of takeaways from that are on the EA Forum.
Kuhan Jeyapragasan: We also recently ran a postdoc hiring round to get some postdoc researchers for next year. And our professors run a class at Stanford called Preventing Human Extinction to introduce first-year students to existential risk. And then there’s the standard programming to get students excited about these issues.
Rob Wiblin: Are you part of the university?
Kuhan Jeyapragasan: Yeah. I’m hired by Stanford University to work on SERI.
Rob Wiblin: I see. It seems a little bit hard to classify exactly what this is. It’s like, it’s a local group, but it’s available to everyone and it’s focused on existential risks — broadly speaking, all different kinds of ones — and you run events, and you’re also doing other stuff…
Kuhan Jeyapragasan: Yeah, SERI got started basically at the beginning of the pandemic with a pretty open-ended mandate of wanting to promote existential risk education and research at Stanford. It was pretty open-ended about what moving forward with that could look like. I think things are still pretty experimental. It’s hard to tell if you found your niche or how things could be better — going back to the opportunity cost discussion — but it’s definitely been a great learning experience.
Rob Wiblin: Is it just you or are there other people involved?
Kuhan Jeyapragasan: I’m currently the only full-time employee for SERI. Our Professor Directors, Stephen Luby and Paul Edwards, also spend obviously a lot of time teaching the Preventing Human Extinction course, and dealing with a high-level strategy for SERI and other administrative work. And then it’s mostly student organisers now. I think we’ll be hiring more full-time organisers, more staff members shortly. But yeah, right now it’s me.
Rob Wiblin: Very cool. What’s the reaction been to a university course on human extinction?
Kuhan Jeyapragasan: Actually quite positive, I believe. Stanford has these first-year requirements and one of these requirements is a programme called Thinking Matters — where every first-year student has to take a Thinking Matters class — and Preventing Human Extinction is one of these Thinking Matters options. I believe for the past two years — I don’t know about this year, but at least for the past two years — it was the Thinking Matters class with the highest enrolment. So that was pretty exciting.
Rob Wiblin: Wow. So Thinking Matters is trying to get students to maybe do some practical, interdisciplinary subject?
Kuhan Jeyapragasan: Yeah, I think so. I regrettably don’t have a great understanding of what actually links all the Thinking Matters courses, but I think it’s something about integrating academic ideas into how they’re relevant to society at large.
Rob Wiblin: Makes sense. Thinking back to as an undergrad, what would I do, I think the human extinction course would be a pretty appealing option. It sounds like it’s going to be all over the place and kind of exciting.
Kuhan Jeyapragasan: For sure. I mean, definitely a very provocative title.
Rob Wiblin: Yeah. It sounds like, in order to get actually hired officially by the university, you must have buy-in from some academics or some academic department that is excited about existential risk and wants to make Stanford kind of a centre for thinking about this topic.
Kuhan Jeyapragasan: Yeah, for sure. So SERI is officially housed in the Center for International Security and Cooperation, which both of our professors are affiliated with. I think it’s just super exciting that there are professors who care enough about these topics to teach a whole course on it. I’m really excited about the future of university and academic engagement with these topics. And there are more and more professors who are taking these ideas really seriously and incorporating it into their research. So yeah, I’m excited for progress in this space.
Rob Wiblin: When you graduated, what other career next steps did you seriously consider?
Kuhan Jeyapragasan: I think one of my biggest lessons from my (unfortunately not particularly strategic) career exploration has been that exploring as much as possible early on would’ve been really valuable. I read the 80,000 Hours career guide, kind of just assumed I’m a reasonably mathy person, and found the arguments around longtermism and existential risk quite compelling. For the first three summers after getting into EA, I did different kinds of research — so I did some qualitative econ health policy research, and then the next two summers did more research then I did more quantitative stuff. So first statistics research and then machine learning or reinforcement learning research.
Kuhan Jeyapragasan: I don’t know, I got the sense that I was fine at research and it seemed reasonable, but I think in retrospect it was pretty clearly not the right fit for me. And I only happened to realise this because of, as I mentioned earlier, the retreat where I learned about the Community Building Grants. I hadn’t even intended, when I applied for the grant, for that to be a career option — that just wasn’t in my realm of possibility for me at the time, or I didn’t even realise it was an option. And it was only when I started seriously putting effort into community building, largely as a result of the Community Building Grant, that I realised, “Oh wow, this is what it’s like to really, really enjoy my work.” Cliches that I had heard about work that didn’t really feel like work, or the kind of thing that I think about for fun in the shower or talk to my friends about…
Rob Wiblin: It started happening to you.
Kuhan Jeyapragasan: Yeah. When I realised how exciting work could be — especially when I was really motivated by it, and also got the sense that I was maybe pretty good and doing quite well and actually having an impact — that feeling was really exciting for me. And it’s quite possible that if things had gone differently, I wouldn’t have found this thing. The same argument implies that maybe there’s something even better out there. But I wish I had explored more and been more intentional and deliberate about doing cheap tests to fit and following advice that the 80K career guide gave that I just maybe didn’t listen to enough.
Kuhan Jeyapragasan: In terms of specifically what led me to picking SERI over pursuing research more or other options, I think talking to quite a few more senior community members. Across the board, people said, “I think you should do community building. It seems like it’s going quite well. It seems like you’re able to generate some pretty cool insights and ideas for projects to run and are having some pretty good success, getting some other really, really great, sharp, dedicated community members involved.” And SERI seemed like a pretty unique opportunity to see how official university engagement could go. I was surprised at how basically every experienced community member I talked to said that they thought I should do full-time community building and felt excited about SERI in particular.
Rob Wiblin: People often give more conflicting messages than that, so I suppose it’s reassuring to have all the signals being positive. It sounds like SERI has tonnes of different balls in the air. What’s maybe the one that you are most excited about at the moment?
Kuhan Jeyapragasan: The thing I feel most excited about is improving infrastructure for the pipeline between a student who is intrigued by or excited about existential risk reduction and improving the long-term future — going from that point to concretely figuring out how they can meaningfully contribute full time to these issues.
Kuhan Jeyapragasan: I think there’s a lot of infrastructure missing in that process, and this infrastructure exists for many other career paths, at least to a much larger degree. You can imagine if you’re interested in software engineering careers, there’s pretty standard advice around reading Cracking the Coding Interview or doing some Lede code training, and all these resources on interview prep, and all these internships that are available, and even first- and second-year specific programmes for more junior students. In comparison, if you’re interested in existential risk reduction, the set of concrete actions to take is just much less clear.
Kuhan Jeyapragasan: And that was one of the main motivations behind starting the SERI summer internship two years ago — just wanting to make there be more concrete options for people who were really excited about these ideas, but didn’t know what concretely to do to, at the very least, get to try out research, be part of community with others, receive mentorship from more senior community members who have been thinking about these issues for longer. Making the option of tackling this problem as a real full-time career option.
Kuhan Jeyapragasan: So yeah, more things in that vein. More summer programmes, educational programmes like the fellowships that Cambridge is running, career advising, maybe formalising mentorship. And as I briefly mentioned earlier, making it easier to find peers and hopefully potential friends who are in a similar position, working with them and making progress together. Because I think it can be a lot less daunting when you’re on this journey together with friends, rather than trying to figure out all these big, scary, potentially very impact-relevant things on your own.
Rob Wiblin: Totally. It sounds like you’re going to have your hands extremely full in coming years. But I’m really glad you stumbled into that community-building event, and it’s pushed you towards a career that seems like it might be a really excellent fit for the long term.
Kuhan Jeyapragasan: Hopefully. And insofar as there are maybe even higher-impact options out there, I hope I can find them. One thing I continually try to do is ask myself, if I look back in two years or five years or what have you and think, “Oh man, how could I not have realised this really important consideration? Or that A was better than B?” What might those things be? And how could I have realised that sooner?
Kuhan Jeyapragasan: There’s an EA Forum post I really liked by Thomas Kwa called “Effectiveness is a conjunction of multipliers” that basically makes the argument that, say you want to be altruistic and think of something like, “Maybe I can give my friend some socks.” And then the post lists all these multipliers or considerations that could multiply your impact, say 10x or 40x or 2x or whatever.
Kuhan Jeyapragasan: And even just missing one of these considerations, if there are 20 or 10 to the optimal action, missing one 10x consideration out of 20 possible considerations means you’re only having 10% of the impact you could be having. And obviously that’s like an extremely high bar — I don’t know how reasonable it is to ex ante think of all the things, but I think that mindset just makes salient the importance of thinking rigorously about all the potential considerations. “Work smart, not hard” or “Work smart, then hard” maybe.
Rob Wiblin: Yeah. “Impact is a conjunction of multiplications” — we should stick it on a billboard. My guest today has been Kuhan Jeyapragasan. Thanks so much for coming on the 80k After Hours podcast, Kuhan.
Kuhan Jeyapragasan: Thank you so much for having me.
Keiran’s outro [00:59:01]
Keiran Harris: Just an announcement: we’re looking for another audio editor to join our podcast team.
We’re hoping to create a lot more content for both The 80,000 Hours Podcast and 80k After Hours, and we could really use some extra help to make that practical.
You’d be working with our lead audio engineer Ben Cordell on both shows, and the part-time role would include:
- Audio mastering
- Removing silences
- Taking out filler words
- And generally making our hosts and guests sound smarter than they actually are in real life — especially the guy who most regularly hosts the show
We’re looking for someone with at least one year of experience doing broadly similar work, and ideally someone who gets what we’re trying to do, and is excited about more of it existing.
So if you want to get paid competitive industry rates to listen to 80,000 Hours podcasts, or know someone we should reach out to — please let us know by sending an email to [email protected].
All right, audio mastering and technical editing for this episode by Ryan Kessler.
Full transcripts and an extensive collection of links to learn more are available on our site and put together by Katy Moore.
And I produce the show.
Thanks for listening.