I frequently hear concerns about burnout in EAs who are unsustainably pushing themselves in pursuit of the greater good. And I get it. We’re racing a marathon, not a sprint, and rest is vital. Yet, for people striving for causes they care so passionately about, it can also be hard to slow down.

What if there was a way to balance both?

In today’s interview, I’m joined by Niel Bowerman, an AI policy specialist at 80,000 Hours. Among other topics, Niel shares how he implements 80,000 Hours’ bout system, which alternates sprints on a focused goal with more relaxed planning periods.

Starting in the interbout period, you plan your goals and take things slower. The regular period for reflection ensures this vital step doesn’t get ignored, hence saving you time and effort later. This is also the period where you might go on vacation or catch up on admin tasks, so they don’t get perpetually pushed aside.

Next, 80,000 Hours has nailed how to set a concrete goal with accountability. You set clear metrics for the goal during the bout, with both a realistic and stretch goal. For example, Niel’s 25 call goal and 35 call stretch goal discussed below. During the process of setting the goals, you get feedback on your goals from a few people to help robustly prioritize.

Finally, you sprint to execute goals for a time-bound period of five to eight weeks, before you repeat the whole process again. During the sprint, you’re in heads-down execution mode to make as much progress on your top goal as possible.

Avoiding decision paralysis, you don’t rethink or strategize on your goals unless you get new information that would have changed your initial plan. Last but not least, the 80,000 team has a culture that encourages supportive accountability, including a norm where teammates can post their daily goals and successes or failures on Slack.

You can read the full transcript below for other ways Niel uses social accountability to boost his motivation, plus:

• His goal-setting system from career decisions to day-to-day prioritization
• The single biggest factor influencing his productivity
• The questions he asks to stay efficient
• How he increased his focused work hours
• His best weird tips for productivity

Highlights

“When there's a task that I feel I've been procrastinating on – especially send this email to this important person – I just put 10 bucks down on someone's desk, and I'm like, ‘if I don’t send this email in 20 minutes the money is yours’ or an hour-and-a-half or however long it is. And so as soon as I noticed this feeling of, ‘Oh this is taking longer than it should’ I'll often do that money down thing. It feels really good. Because then I get it done.”

“I think that sleep is the single biggest determinant of my productivity, and I prioritize it a lot.”

“Key focus hours became a big deal for me when we were tracking my productivity at 80,000 hours. It seemed like there were, basically, three intervention points. It was Niel could work more hours, Niel could work on his key priorities for more of the hours he works, or Niel could get his key priorities done faster. And of those three points of leverage, it felt like the second one was probably the one I wanted, that I could most push up most easily. So a focus of mine has been pushing up the fraction of my hours that I spend on key focus hours. I think I've already pushed that up by maybe six or seven hours a week on average, just by focusing on it a bit more.”

“I normally start with an outcome. So I'm like, ‘What is the outcome I actually care about here?’ Because often there are faster ways of getting to an outcome than the product design. And so I'll be like, ‘What might be an outcome?’ Like, ‘These people have meetings scheduled with me at EA Globals or something’. Then I’ll just think through the things that need to happen for that outcome to happen, and just break it down into those pieces. I'll often do a thing like, ‘If I only had an hour to complete this project, what would I do?’, and then think about what it would look like in that case. Then I think about like, ‘Is it even worth doing more than that?’ If there's any doubt in my mind, then I'll just sprint the hour version and then reassess from that point. So that's what happened for EA Global prep, for example. We ended up just hiring a VA to do a lot of the tagging and screening of people and things, and then used templates to email everyone.”

Full Interview

Note: This transcript was edited for clarity and length.

Lynette Bye: Okay, so one of the big areas that I'm asking about is prioritization. You have quite a detailed system. I'm wondering if you could explain some of how you go about prioritizing what things to focus on?

Niel Bowerman: Sure. Hi future Lynette. Okay, Niel's prioritization system. So, I can start at the highest level. The highest-level privatization system is Niel thinks about his goals in the world, and then he writes them down in a sort of life plan type thing. Niel's goals are to maximize utility, and that hashes out as reduce X-risk, and that hashes out as reduce AI risk. Within that, things get a bit more complicated. But in my current role, that hashes out as maximize IASPCs, impact adjusted significant plan changes. And then that gives rise to annual goals. I used to have decadal and three-yearly, but they changed so much that now I just go for annual goals. And then that cuts down into bout targets. So 80,000 Hours splits its year up into five or six bouts and we have bout targets. And then...

Lynette Bye: You want to explain what a bout is?

Niel Bowerman: So we have sprints, which are bouts and then we have interbout periods, where you're allowed to doss about a bit more and go on holiday and catch up on smalls. Whereas in your bouts, you're supposed to sprint on a key priority. We are currently in a bout, hence why I have like 22 meetings in two days. But if EAG was in an interbout period than I'd probably go a little bit easier.

Lynette Bye: And there are like six weeks in a bout?

Niel Bowerman: It really varies. Some of them are eight weeks, some of them are five weeks. Some of them are getting longer, I think. We’re currently coming up to the end of these five weeks. In the bout, I have a very clear metric goal. For example, in this bout my goal is to have 25 calls, where I pitch a new job opportunity to people over the course of five weeks. And my stretch goal is to have a number larger than that, which is seven times five, which is 35.

Niel Bowerman: So then that comes down to weekly goals, where I'm figuring out what I'm going to achieve that week. We have two systems for that. One is that I make a weekly commitment that I give to the 80,000 Hours team, and I got to hit it with 90% confidence. And then the other thing is goals that I set for the week, and then sort of see if I hit those. Then they get brought down into daily goals and each day, I try to hit my daily goals. And I use Asana prioritize amongst those. And that is the Niel productivity system.

Lynette Bye: So fleshing a couple of these details, how many daily and weekly goals do you set?

Niel Bowerman: Generally like two to six daily goals. And then the weekly goals are, more like two or three or something.

Lynette Bye: Cool. And about how many hours do you work in a day on these top priorities?

Niel Bowerman: There is data for this because I track everything in Toggl. And my key priorities, my goal is to spend 17 hours a week, my stretch goal is to hit 20 hours a week. And I think in practice it varies a lot, but I rarely go below 20. I need to update these goals. But it's often around 22, maybe average something like that.

Lynette Bye: And what is the limiting factor here? Do you get tired, does your marginal productivity decrease? Do you need to answer emails?

Niel Bowerman: Yeah, so the limiting factor is I only want to work about 55 hours a week or something like that at the most. And so maybe I'll experiment in pushing that up to 60. But somewhere between 45 and 60 hours a week. That is hours in the office, and that transfers into Toggl hours at a rate of something like 75% to 85% just because of pee breaks and chatting to people and eating and stuff like that. And so that translates into around 40 Toggl hours, like 35 to 40 Toggl hours a week. And then, then the fraction of those that are spent on key priorities. The rest of it is meetings with people and emails, which are the two biggest sucks of time. Also internal comms, which is checking Slack or recording my goals and stuff like that.

Lynette Bye: And I'm curious why this is your goal for the week? What stops it from being more?

Niel Bowerman: Right. Yeah. So the reason I set those goals was because it seemed like historical averages of what I was doing. And then what stops it from being more. One way of answering that question would be "Niel not saying no to more things." Another way of answering it is, "Niel being very helpful with things that are not his day job."

Niel Bowerman: I think in some weeks, it is a lot more. City visit weeks are the weeks when I tend to hit the highest number of hours. So a week like this one, where I have 22 back-to-back meetings this weekend. And then I had a bunch of stuff crammed in for after, I could easily hit 40 plus, sometimes 50 hours of key focus time, where I'm just in productive meetings, with people on useful topics. But in my normal week, I allow more non-key priority meetings into my schedule.

Lynette Bye: Cool. And jumping back a little bit at the top of your back chain, from your goals down to what you do. There was X-risk, AI safety, etc. When I first introduced back chaining to people who aren't already doing this, they sometimes start somewhere up here and get so overwhelmed that they can't make any progress. How do you go through actually deciding what these top ones are? Till you get back to down to something like impact adjusted plan changes, the thing that you're optimizing for?

Niel Bowerman: Yeah, so the first thing to says is maximizing utility is only my goal during my working hours. And the rest of my time my goal is to have fun. Have a nice family and life and friends and all of those things. And so that only applies for those hours. Then how do you come to the decision to maximize utility with your career is like. So the next step down I think is picking a cause. And so EAs tend to end up picking like "I'm doing to do X-risk reduction", or "I'm going to do animals or poverty or meta". Maybe you care about global health, maybe you care about climate change or environment.

Niel Bowerman: My metric used to be maximize the avoided emissions measured in tons of carbon dioxide equivalent, back when I worked on climate change. And that was the metric I was trying to maximize with my work. But once you've picked a cause area, then some of these have sub causes. So within X-risk production, maybe you want to pick AI. Within AI, maybe you want to pick AI policy or technically AI safety. Same similar things within global development. Maybe you want to work on global health. Within global health, maybe you want to work on malaria. Within malaria, maybe you work on bed net distribution. So you’re basically narrowing down from cause area to intervention.

Lynette Bye: And this is through some combination of your best guess at what's important and the best guess at what fits your skills?

Niel Bowerman: Yeah. And yeah, exactly. 80,000 Hours has lots of info on how to do this. And yeah, then once you have a sense of what piece of that you want to contribute to, then you want to start about thinking a bit about, "What do I have to offer?" And so for me, well, I could contribute in policy. That's the thing that I've thought long and hard about maybe doing, I ended up deciding to contribute in community building. I just find that more fun, I prefer talking to people as my day job. And so I ended up at 80,000 Hours doing this sort of work.

Lynette Bye: Okay. And once you've gone there, it sounds like you track a lot of things. And I'm trying to optimize for those. How do you choose which things to track? And how to set up these feedback loops?

Niel Bowerman: Yeah. So basically, I only track if it's useful.

Lynette Bye: But how do you decide that?

Niel Bowerman: Yeah. So, key focus hours became a big deal for me when we were tracking my productivity at 80,000 hours. It seemed like there were, basically, three intervention points. It was Niel could work more hours, Niel could work on his key priorities for more of the hours he works, or Niel could get his key priorities done faster. And of those three points of leverage, it felt like the second one was probably the one I wanted, that I could most push up most easily. And so a focus of mine has been pushing up the fraction of my hours that I spend on key focus hours. I think I've already pushed that up by maybe six or seven hours a week on average, just by focusing on it a bit more.

Lynette Bye: And how do you stay motivated as you're going about things? Do you find that just setting these goals is enough to keep making progress toward them or you ever need to do other things?

Niel Bowerman: Yeah, so then how do I stay motivated? I think part of it is about having daily goals that I need to report back to the team on. So every day I write up in Slack, “These are my goals for the day.” And at the end of the day, I report back to the team on how I did with my goals. And so a lot of us do this.

Lynette Bye: And do you check with each other? Would you notice if someone didn't follow up with them?

Niel Bowerman: Uh no, but people would write like, "I didn't meet my goals today". And fortunately within 80k culture, it's very warm and respected that no one always makes their goals, and we support people to be open about that fact. So how I stay motivated, that's one thing. Another thing is just having clear goals. Another thing is doing a thing I really enjoy.

Niel Bowerman: I think I still have some work to improve on, minute by minute. Like, I still find that I probably spend an hour a week during work time on news sites or something. And that's kind of an annoying thing that I'd love to cut out of my schedule. It might not be as high as that now; it might be half an hour week on news sites or something like that. I've just blocked all these sites on my phone, and that's been really helpful. Yeah. So I think it's a combination of having really good day-to-day motivations, being excited about the thing I'm trying to achieve. Having a packed schedule helps as well. And then yeah, having it be costly to get distracted.

Lynette Bye: What weird things do you find to make a really big impact on productivity?

Niel Bowerman: Sleep, it's not weird, but it's huge. So huge.

Lynette Bye: How much sleep do you normally need?

Niel Bowerman: I think I average about eight hours and 15 minutes, eight hours and a half. But like when I'm at EA Global, I carve out time to get about nine and a half. Just because this is fucking draining. But then during a normal week, it's sometimes more like eight. I think that sleep is the single biggest determinant of my productivity. And I prioritize it a lot. Yeah. What other weird things? I find Toggl weirdly motivating, just trying to hit targets on Toggl. I know that it's an input goal, not an output goal, but it still keeps you motivated to keep working.

Niel Bowerman: Oh, I do have a few things that make me go into turbo mode in terms of productivity. One is, basically, an important person asking me for a thing on a short deadline. That generally works quite well. One thing I'm working with at the moment is not ever doing life admin during work hours. Maybe that's too strong, avoiding wherever possible doing life admin during work hours. And yeah, postponing to the weekend, whenever possible, even if it means spending 10 pounds on delay fees or whatever it needs to be. And that forces me to do it early enough, like the previous weekend or doing it before I get to work or once I get home. And that's been pretty good. And then also, just not doing life admin until it's on fire has been surprisingly good. But I'm not sure I'd endorse it for everyone.

Lynette Bye: Do you want to explain what that means?

Niel Bowerman: Yeah. It's just, there are so many things in my life that I could spend some time on. Like, okay, here's a great example. I've noticed this thing, where maybe a quarter of life admin up and just disappears if you just leave it long enough. So minor things are like, there are these repairs in our flat that we've reported to our landlord, and he hadn't done anything about, and every day it would cross my mind “Maybe I should email him about it”, and I would just never really do it. And then just one day, the landlord came and fixed them. And all of that time I could have been hassling the landlord would have been wasted of time.

Niel Bowerman: Or, for example, I carried around superglue in my suitcase for two months to fix my shoes. And then my shoes were stolen in Mexico last week, and so fixing them was not useful. Also, often life admin gets done by someone else who's more keen to have it done or has a bit more context for something. That's another reason for delay. Yeah, it's a thing I've been trying on for size, and it's been working surprisingly well. The context is that I moved to England and had vast quantities of life admin. And found this to be a surprisingly fine way of prioritizing life admin.

Lynette Bye: Okay. How do you prioritize during the day? So you have some goals? But what how do you decide, what is urgent and important given limited time?

Niel Bowerman: Yeah. So I'm actually not sure that I do this that well. I can talk you through my current system. So my current system is I have a frog. And I eat the frog first thing in the morning. And the frog is generally my highest priority task that takes less than three hours. Then once I eat the frog, I do what in Asana is called morning start, and morning start involves clearing the new tasks in Asana and clearing Slack, then writing out my daily goals and brushing my teeth and doing my hair and taking creatine and things like that, that you need to get done every morning. And the last thing on my morning starts is clear your inbox, email inbox. So then I dive into my email inbox, clear that out.

Niel Bowerman: And then by this time, it's generally lunchtime, at least. And then I'll have a list of other things I needed to do that day that I do in the afternoon. Generally, lunchtime comes before I finished clearing my inbox, maybe after the frog. The frog sometimes gets a little out of hand. And so essentially what I'm thinking about is what is the most important thing to get done. And then the morning start keeps everything tidy and keeps my life on track. And then I go through them and I tend to use the afternoon for the more urgent things. That's generally how it works.

Lynette Bye: The animal welfare people might be horrified at how many people are eating frogs in the morning. So how do you choose what your top priority thing is? What's the process for "Here's the important thing?"

Niel Bowerman: Well, in my bout goals, I think about what is going to make the most difference on my top metric. In the bout goals, I define a bunch of stuff, and then that basically just translates down into day-to-day. Basically, my top thing each day is just what's going to make the biggest difference on my bout goal.

Lynette Bye: So it's like, “Here's the ultimate goal, what is the bottleneck or biggest thing you need to do to push that towards completion?”

Niel Bowerman: Yeah, in theory. In practice, it's like, "Oh, I said, I was going to give a talk at Oxford next week. I haven't written the talk yet. That's my frog." Or like, "Ah, EA Global's next week. I haven't finished reviewing the spreadsheet of all the people and inviting them to meetings, I'll do that today." And so it tends to be more driven by calendar than the description I gave. But when there's not a driven-by-calendar thing, then I also have this sort of running list of the next three big projects I want to do. And when I say big projects, I mean, things are going to take at least a day, that are within my bout targets. So then I work down this list.

Lynette Bye: Got it. And how do you scope out these projects? To get them done efficiently, while still capturing all the value that you can by doing them well?

Niel Bowerman: Yeah, so I normally start with an outcome. So I'm like, "What is the outcome I actually care about here?" Because often there are faster ways of getting to an outcome than the product design. And so I'll be like, “What might be an outcome?” Like, "These people have meetings scheduled with me at EA Globals or something". Then I’ll just think through the things that need to happen for that outcome to happen, and just break it down into those pieces.

Niel Bowerman: I'll often do a thing like, "If I only had an hour to complete this project, what would I do?", and then think about what it would look like in that case. Then I think about, “Is it even worth doing more than that?” If there's any doubt in my mind, then I'll just sprint the hour version and then reassess from that point. So that's what happened for EA Global prep, for example. We ended up just hiring a VA to do a lot of the tagging and screening of people and things, and then used templates to email everyone. So nobody got a personalized email.

Lynette Bye: But how long would it have taken you to personalized emails compared to what you did?

Niel Bowerman: Oh, way longer. Yeah, it wasn't worth it.

Lynette Bye: Okay, it sounds like you are motivated by social expectations, just other people knowing what you're going to do quite a lot.

Niel Bowerman: Yeah, totally.

Lynette Bye: Do you ever do more extreme things like financial policies, that kind of thing?

Niel Bowerman: Oh, yeah, I run the whole gamut. So the thing that really works for me unnecessarily well is just having my screen in the eyeline of my boss, that works super well. We rearranged our office, specifically at my request so that my screen could be in his eye line. And that's been useful at making me not dick around. Most people will find that insanely stressful. But fortunately, I really like my boss and trust him a lot so that works.

Niel Bowerman: Yeah, financial penalties. When there's a task that I feel like I've been procrastinating on – especially send this email to this important person – I just put 10 bucks down on someone's desk, and I'm like, “If I don’t send this email in 20 minutes the money is yours” or an hour-and-a-half or however long it is. And so as soon as I noticed this feeling of, “Oh, this is taking longer than it should,” I'll often do that money down thing.

Lynette Bye: How does that feel?

Niel Bowerman: Oh, really good. Because then I get it done. And then I feel good. And get my money back. I've actually never lost money with that method.

Lynette Bye: Impressive.

Niel Bowerman: Yeah, I think you just pick the time frame that is long enough, but you're definitely going to complete it in that timeframe. And then often on a thing you've been procrastinating on for a while, you know exactly what you need to do. Because your brains already mulled it over far more times than it needs to. And so you can just execute on it. For bigger projects, I'll use Beeminder and things like that. Writing my thesis or writing my articles, I'll track word counts and things like that in Beeminder and that helps.

Lynette Bye: Oh cool, yeah. For your bouts, do you have a list of projects on the back burner that might come up someday for a bout?

Niel Bowerman: Yeah, yeah, totally.

Lynette Bye: How do you mull those over and refine them until it's like, "Okay, I'm ready to commit to this one for however many weeks"?

Niel Bowerman: Yeah, I've got a doc that has possible project ideas. And then they start as headings, then, as they become a thing I want to think about more, they become paragraphs. And then if I want to think it about more than they became one-pagers, and two-pagers, and then not always. This is just like, especially...

Lynette Bye: The ones that make it up the list.

Niel Bowerman: Yeah, sorry, and not even all the ones that make it. Sometimes they just get fast-tracked. And I'm like, “This seems like a good idea, I'm doing it.” But for anything, that seems like a bit more contentious, then I write it out. And so for example, I still have not made a Slack group for people interested in policy. Because this is a little contentious as to whether it's a good idea to do. But there's now a one- or two-pager that a couple people have commented on. And that's the sort of thing that helps me prioritize is people comment on these things.

Niel Bowerman: For each bout, I propose what I want to do. And then my manager and a couple of other stakeholders take a look at the list of things I decided not to do and the list of things I decided to do. And they say if in their view any of them should be swapped around. And then we lock in a bout plan. And the bout plan is ideally locked in before the bout starts. But sometimes it doesn't get locked into the first week of the bout. Then you go and execute on that plan. You get held accountable to the commitments you made in the plan.

Lynette Bye: Yeah. Do you ever as you start executing the plan, learn more about it, find that your view of what it should be framed as changes?

Niel Bowerman: Oh, yeah, totally. Then I'll say to my manager, "Hey, I think I should not do this for this reason." And my manager is like, "Cool." And then when we do the bout review, we say like, "We planned to do this, we reassessed, we decided to change it over to this thing." Everyone's willing to change, actually, I should clarify that a bit more. It is frowned upon to dramatically change your bout plan mid bout because you're not supposed to be doing strategy work mid bout. But if you get new information that implies that this thing is a bad idea, then you're encouraged to change your bout plan.

Lynette Bye: Do you do any work pre-bout to vet the plan or test it out?

Niel Bowerman: A little bit, mostly just writing up the plan again, getting comments on it. Sometimes you do little tests but mostly not.

Lynette Bye: Okay, then you can reassess after a month or two?

Niel Bowerman: Yeah, exactly. The bouts aren't that long.