We’re Rethink Priorities. Ask us anything!

I think donating to the EA Funds is a very good thing to do, but I don't think every donor should do this. I think for donors who have the time and personal fit, it would be good to do some direct donations on your own and support organizations to help those organizations hedge against idiosyncratic risk from particular funders and help give them more individual support (which matters for showing proof to other funders and also matters for some IRS stuff).

I don't think any one funder likes to fund the entirety of an organization's budget, especially when that budget is large. But between the different institutional funders (EA Funds, Survival and Flourishing Fund, OpenPhil, etc.), I still think there is a strong (but not guaranteed) chance we will be funded (at least enough to meet somewhere between our "Low" and "High" budget amounts). Though if everyone assumed we were not funding constrained, than we definitely would be.

My other pitch is that I'd like RP, as an organization, to have some direct financial incentive and accountability to the EA community as a whole, above and beyond our specific institutional funders who have specific desires and fund us for specific reasons that don't always match what the community as a whole wants or needs.

Lastly, if you trust us, we also value unrestricted funds highly (probably 1.5x-2x per dollar) because this allows us to start new research areas and programs that have less pre-existing proof/traction and get them to a point where they are ready to show bigger funders.

So there are a bunch of questions in this, but I can answer some of the ops related one:

• We haven't had ops talent bottlenecks. We've had incredibly competitive operations hiring rounds (e.g. in our most recent hiring round, ~200 applications, of which ~150 were qualified at least on paper), and I'd guess that 80%+ of our finalists are at least familiar with EA (which I don't  think is a necessary requirement, but the explanation isn't that we are recruiting from a different pool I guess).
  • Maybe there was a bigger bottleneck in ~2018 and EA has grown a lot since or reached people with more ops skills since?
  • We spend a lot of time resources on recruiting, and advertise our jobs really widely, so maybe we are reaching a lot more potential candidates than some other organizations were?
• Management bottlenecks are probably our biggest current people-related constraint on growth (funding is a bigger constraint).
  • We've worked a lot on addressing this over the summer, partially by having a huge internship program, and getting a lot of current staff management experience (while also working with awesome interns on cool projects!) and sending  anyone who wants it through basic management training.
  • My impression is that we've gotten many more qualified applications in recent manager hiring pools.
• Bypassing bottlenecks
  • In general, I think we haven't experienced these as much as other groups (at least so far)
  • We tend to hire ops staff prior to growth, as opposed to hiring them when we need them to take on work immediately (e.g. we hire ops staff when things are fine, but we plan to grow in a few months, so the infrastructure can be in place for expansion, as opposed to hiring ops staff when the current ops staff has too much on their plate, or something).
  • We do a ton of prep to ensure that we are careful while scaling, thinking about how processes would scale, etc.
  • The above mentioned intern program really stress-tested a lot of processes (we doubled in size for 3 months), and has been really helpful for addressing issues that come with scaling.
• Downsides to hiring quickly
  • I'd say that we've seen a mild amount to the downsides to growing in general, though it hasn't necessarily been related to speed of hiring - e.g. mildly more siloing of people, people not sure what other people are working on, etc. and we've been taking a lot of steps to try to mitigate this, especially as we get larger.

This comment sounds like it's partly implying "RP seems to have recently overcome these bottlenecks. How? Does that imply the bottlenecks are in general smaller now than they were then?" I think the situation is more like "The bottlenecks were there back then and still are now. RP was doing unusually well at overcoming the bottlenecks then and still is now."

The rest of this comment says a bit more on that front, but doesn't really directly answer your question. I do have some thoughts that are more like direct answers, but other people at RP are better placed to comment so I'll wait till they do so and then maybe add a couple things. 

(Note that I focus mostly on longtermism and EA meta; maybe I'd say different things if I focused more on other cause areas.)

In late 2020, I was given three quite exciting job offers, and ultimately chose to go with a combo of the offer from RP and the offer from FHI, with Plan A being to then leave FHI after ~1 year to be a full-time RP employee. (I was upfront with everyone about this plan. I can explain the reasoning more if people are interested.)

The single biggest reason I prioritised RP was that I believe the following three things:

1. "EA indeed seems most constrained by things like 'management capacity' and 'org capacity' (see e.g. the various things linked to from scalably using labor).
2. I seem well-suited to eventually helping address that via things like doing research management.
3. RP seems unusually good at bypassing these bottlenecks and scaling fairly rapidly while maintaining high quality standards, and I could help it continue to do so."

I continue to think that those things were true then and still are now (and so still have the same Plan A & turn down other exciting opportunities). 

That said, the picture regarding the bottlenecks is a bit complicated. In brief, I think that: 

• The EA community overall has made more progress than I expected at increasing things like management capacity, org capacity, available mentorship, ability to scalably use labor, etc. E.g., various research training programs have sprung up, RP has grown substantially, and some other orgs/teams have been created or grown.
• But the community also gained a lot more "seriously interested" people and a lot more funding.
• So overall the bottlenecks are still strong in that it still seems quite high-leverage to find better ways of scalably using labor (especially "junior" labor) and money. But it also feels worth recognising that substantial progress has been made and so a bunch more good stuff is being done; there being a given bottleneck is not in itself exactly a bad thing (since it'll basically always be true that something is the main bottleneck), but more a clue about what kind of activities will tend to be most impactful on the current margin.

I'll let Peter/Marcus/others give the organizational answer, but speaking for myself I'm pretty bullish about having more RP-like organizations. I think there are a number of good reasons for having more orgs like RP (or somewhat different from us), and these reasons are stronger at first glance than reasons for consolidation (eg reduced communication overhead, PR). 

1. The EA movement has a strong appetite for research consultancy work, and RP is far from sufficient for meeting all the needs of the movement. 
2. RP clones situated slightly differently can be helpful in allowing the EA movement to unlock more talent than RP will be able to.
   1. For example, we are a remote-first/remote-only organization, which in theory means we can hire talent from anywhere. But in practice, many people may prefer working in an in-person org, so an RP clone with a physical location may unlock talent that RP is unable to productively use.
3. We have a particular hiring bar. It's plausible to me that having a noticeably higher or lower hiring bar can result in a more cost-effective organization than us. 
   1. For example, having a higher hiring bar may allow you to create a small tight-knit group of supergeniuses pursuing ambitious research agendas
   2. Having a lower hiring bar may allow you to take larger chances on untapped EA talent, is maybe better for scalability, and also I have a strong suspicion that a lot of needed research work in EA "just isn't that hard" and if it's done by less competent people, this frees up other EA researchers to do more important work.
4. More generally, RP has explicitly or implicitly made a number of organizational decisions for how a research org can be set up,  and it's plausible/likely to me that greater experimentation at the movement level will allow different orgs to learn from each other. 
5. Having RP competitors can help keep us on our toes, and improve quality via the normal good things that come from healthy competition. 
6. Having an RP competitor can help spot-check us and point out our blindspots.
   1. I'm pretty excited about an EA red-teaming institute, and maybe a good home for it is at RP. But even if it is situated at RP, who watches the watchmen? I think it'd be really good for there to be external checks/red-teaming/evaluation of RP research outputs.
      1. Right now, the only org I trust to do this well is Open Phil. But Open Phil people are very busy, so I'd be really excited to see a different org spring up to red-team and evaluate us.
7. AFAICT when doing very rough BOTECs on the expected impact of RP's research, the EV of RP work is massively cost-effective (flag: bias). If true, I think there's a very simple economics argument that marginal cost (including opportunity cost) should equal marginal revenue (expected impact), so in theory we should be excited to see many competitors to RP until marginal cost-effectiveness becomes much lower.

One other thing I'd add to Linch's comments, adapting something I wrote in another comment in this AMA:

If anyone feels like maybe they're the right sort of person to (co-)found a new RP-like org, please feel free to reach out for advice, feedback on plans, pointers to relevant resources & people! I and various other people at RP would be excited to help it be the case that there are more EA-aligned orgs scaling rapidly & gracefully. 

Some evidence that I really am keen on this is that I've spent probably ~10 hours of my free time over the last few months helping a particular person work towards possibly setting up an RP-like org, and expect to continue helping them for at least several months. (Though that was an unusual case and I'd usually just quickly offer my highest-value input.)

Quick clarifying question: is 

one specialist organisation in a field

referring to RP, or more field-specific organizations like e.g. CSET or an (AFAIK, hypothetical) organization focused on answering questions on medical approaches to existential biosecurity. 

Put another way, is your question asking about larger RP vs RP + several RP clones, or RP + several RP clones vs. RP + several specialist organizations? 

Mostly yes. It definitely is the case that, if we were given more cash than the cash that we already have, we could meaningfully accelerate our longtermism team in a way that we cannot do with the cash we currently have. Thus funding is still an important constraint to scaling our work, in addition to some other important constraints.

However, I am moderately confident that between the existing institutional funders (OpenPhil, Survival and Flourishing Fund, Long-Term Future Fund, Longview, and others) that we could meet a lot of our funding request - we just haven’t asked yet. But (1) it’s not a guarantee that this would go well so we’d still appreciate money from other sources, (2) it would be good to add some diversity from these sources, (3) money from other sources could help us spend less time fundraising and more time accelerating our longtermism plans, (4) more funding sooner could help us expand sooner and with more certainty, and (5) its likely we could still spend more money than these sources would give.

It’s been hard for me to make five year plans, given that we’re currently only a little less than four years old and the growth between 2018 when we started and now has already been very hard to anticipate in advance!

I do think that RP could be 2x as large in five years. I’m actually optimistic that we could double in 2-3 years!

I’m less sure about how much funded work we’d do - actually I’m not sure what you mean by funded work, do you mean work directly commissioned by stakeholders as opposed to us doing work we proactively identify?

I’m also less sure about impact per dollar. We’ve found this to be very difficult to track and quantify precisely. Perhaps as 80,000 Hours talks about “impact-adjusted career changes”, we might want to talk about “impact-adjusted decision changes” - and I’d be keen to generate more of those, even after adjusting for our growth in staff and funding. I think we’ve learned a lot more about how to unlock impact from our work and I think also there will have been more time for our past work to bear fruit.

I think we cover this in our 2021 Impact and 2022 Strategy update!

By its reputation, output, and the quality and character of management and staff, Rethink Priorities seems like an extraordinarily good EA org.

Thanks Charles for your unprompted, sincere, honest, and level-headed assessment. 

Your check will be in the mail in 3-7 business days. 

Thanks for the question and the kind words. However, I don’t think I can answer this without falling back somewhat on some rather generic advice. We do a lot of things that I think has contributed to where we are now, but I don’t think any of them are particularly novel:

• We try to identify really high quality hires, bring them on, train them up and trust them to execute their jobs.
• We seek feedback from our staff, and proactively seek to improve any processes that aren’t working.
• We try to follow research and management best practices, and gather ideas on these fronts from organizations and leaders that have previously been successful.
• We try to make RP a genuinely pleasant place to work for everyone on our staff.

As to your ideas about the possibility of RP’s success being high founder quality, I think Peter and I try very hard to do the best we can but I think in part due to survivorship bias it’s difficult for me to say that we have any extraordinary skills others don’t possess. I’ve met many talented, intelligent, and driven people in my life, some of whom have started ventures that have been successful and others who have struggled. Ultimately, I think it’s some combination of these traits, luck, and good timing that has lead us to be where we are today.

(This other comment of mine is also relevant here, i.e. if answering these questions quickly I'd say roughly what I said there. Also keen to see what other RP people say - I think these are good questions.)

I think to better expand Rethink Priorities, we need Rethink Priorities to be bigger and more efficient.

I think the relevant constraints for "why aren't we bigger?" are:

(1): sufficient number of talented researchers that we can hire

(2): sufficient number of useful research questions we can tackle

(3): ability to ensure each employee has a positive and productive experience (basically, people management constraints and project management constraints)

(4): ops capacity - ensuring our ops team is large enough to support the team

(5): Ops and culture throughput - giving the ops enough time to onboard people (regardless of ops team size), giving people enough time to adapt to the org growth ...that is, even if we were otherwise unconstrained I still think we can't just 10x in one year because that would just feel too ludicrous

(6): proof/traction (to both ourselves and to our external stakeholders/funders) that we are on the right path and "deserve" to scale (this also just takes time)

(7): money to pay for all of the above

~

It doesn't look like (1) or (2) will constrain us anytime soon.

My guess is that (3) is our current most important constraint but that we are working by experimenting with directly hiring managers and by promoting people into management internally. We rolled out management training this summer and also used our internship program, in part, to train management capacity. From a project management perspective, we recently hired a manager and have rolled out Asana across the team and we will continue to focus on the Asana processes we’ve built and make sure they are working before scaling more.

For (4), this will occasionally become a constraint from time-to-time but we solve this by proactively identifying ops bottlenecks and hiring for them well in advance. So far this has gone well.

For (5), I think this will be our next biggest constraint once we solve (3). I think this is best solved just with time to let the current level of growth become normal as well as listening to staff and their concerns. We just launched our biannual staff survey and we are awaiting important staff feedback before hiring more.

For (6), I think also comes with time and probably can be seen in combination with (5).

For (7), I do think we are funding constrained right now - we have room for more funding and definitely need to get money from somewhere in order to continue our work. I’m optimistic that we can get money from our current institutional sources because we haven’t tried too recently to ask them for money and I think they still like us and want us to continue to succeed. But I think, as I’ve mentioned elsewhere, we’d still like other people to support our work to enable us to diversify our funding sources, give us more flexible unrestricted funding that is 1.5x-2x as valuable per dollar to us, and to build us more sustainability / flexibility in the face of idiosyncratic risk.

Sorry that was seven things instead of 2-3, but I think it helps to communicate the full picture.

Thanks so much for this question!

We have learned a lot during our Fellowship/Internship Program. Several main considerations come to mind when thinking about running a fellowship/internship program.

• Managers’ capacity and preparedness – hosting a fellow/intern may be a rewarding experience. However, working with fellows/interns is also time-consuming. It seems to be important to keep in mind that managers may need to have a dedicated portion of time to:
  • Prepare for their fellows/interns’ arrival, which may include drafting a work plan, thinking about goals for their supervisees, and establishing a plan B, in case something unexpected comes up (for example, data is delayed, and the analysis cannot take place)
  • Explain tasks/projects, help set goals, and brainstorm ideas on how to achieve these goals
  • Regularly meet with their fellows/interns to check in, monitor progress, as well as provide feedback and overall support/guidance throughout the program
  • Help fellows/interns socialize and interact with others to make them feel included, welcomed, and a part of the team/organization.
• Operations team capacity and preparedness – there are many different tasks associated with each stage of the fellowship/internship program. It’s crucial to ensure that the Operations Team has enough capacity and time to hire, onboard, support, and offboard fellows/interns, especially when the program is open to candidates worldwide. For example, we work with an international employment organization that acts as a proxy employer in each of the countries our staff and fellows/interns are based. Taking into account the amount of coordination needed between international employment organization – staff internally – fellows/interns is important (the amount will vary significantly between adding 2-3 vs. 10 fellows/interns to the team).
• Internal processes – capacity is one thing, but having strong, internal processes developed beforehand appears to be equally vital. This refers to hiring and candidate selection procedures, establishing reasonable timelines, setting up check-in structures with both fellows/interns and managers, as well as organizing relevant professional development and social opportunities.
• Hiring internationally and remotely – it may be worth considering where most of the team members are located. If most of the staff are in the US time zones, then it may make sense to think how that could affect candidates from completely different time zones (e.g., Australia and Oceania). Will they be able to communicate with their managers easily? Will they have enough opportunities to interact with other fellows/interns and colleagues?

In summary, any fellowship and internship program may be truly beneficial to the organization running it. Most importantly, however, the questions are how to make the program beneficial to fellows/interns, and how will it impact their future education paths and careers.

Good question! Please enjoy me not answering it and instead lightly adapting an email I sent to someone who was interested in running an EA-aligned research training program, since you or people interested in your question might find this a bit useful. (Hopefully someone else from RP will more directly answer the question.)

"Cool that you're interested in doing this kind of project :)

I'd encourage you to join the EA Research Training Program Slack workspace and share your plans and key uncertainties there to get input from other people who are organizing or hoping to organize research training programs. [This is open only to people organizing or seriously considering organizing such programs; readers should message me if they'd like a link.]

You could also perhaps look for people who've introduced themselves there and who it might be especially useful to talk to.

Resources from one of the pinned posts in that Slack:

• See here for a brief discussion of why I use the term "research training programs" and roughly what I see as being in-scope for that term. (But I'm open to alternative terms or scopes.)
• See here for a collection of EA Forum posts relevant to research training programs.
• See here for a spreadsheet listing all EA-aligned research training programs I'm aware of.

You might also find these things useful: 

• Michael's quick notes on RP's internship, RP's processes, how RP picks research projects, etc. [for SERI etc.]
• Collection of collections of resources relevant to (research) management, mentorship, training, etc.
• Improving the EA-aligned research pipeline

I'd also encourage you to seriously consider applying for funding, doing so sooner than you might by default, and maybe even applying for a small amount of funding to pay for your time further planning this stuff (if that'd be helpful). Basically, I think people underestimate the extent to which EA Funds are ok with unpolished applications, with discussing and advising on ideas with applicants after the application is submitted, and with providing "planning grants". (I haven't read anything about your plans and so am not saying I'm confident you'll get funding, but applying is very often worthwhile in expectation.) More info here:

• https://forum.effectivealtruism.org/posts/DqwxrdyQxcMQ8P2rD/list-of-ea-funding-opportunities
• https://forum.effectivealtruism.org/posts/4tsWDEXkhincu7HLb/things-i-often-tell-people-about-applying-to-ea-funds

[...] ...caveat to all of that is that I know very little about your specific plans - this is basically all just the stuff I think it's generically worth me mentioning to people interested in EA running research training programs.

Best of luck with the planning, and feel free to send through specific questions where I could perhaps be useful :) 

Best,

Michael"

Thanks for the question! I think describing the current state will hint at a lot on what might make us change the distribution, so I’m primarily going to focus on that.

I think the current distribution of what we work on is dependent on a number of factors, including but not limited to:

1. What we think about research opportunities in each space
2. What we think about the opportunity to exert meaningful influence in the space
3. Funding opportunities
4. Our ability to hire people

In a sense, I think we’re cause neutral in that we’d be happy to work on any cause provided the good opportunities arise to do so. We do have opinions on high level cause prioritization (though I know there’s some disagreement inside RP about this topic) but I think given the changing nature of marginal value of additional work in any given the above considerations, and others, we meld our work (and staff) to where we think we can have the highest impact.

In general, though this is fairly generic and high level, were we to come to think our in a given area wasn’t useful or the opportunity cost were too high to continue to work on it, we would decide to pursue other things. Similarly, if the reverse was true for some particular possible projects we weren’t working on, we would take them on

[This is like commentary on your second question, not a direct answer; I'll let someone else at RP provide that.]

Small point: I personally find it useful to make the following three-part distinction, rather than your two-part distinction:

• Academia-like: Projects that we think would be valuable although we don't have a very explicit theory of change tied to specific (types of) decisions by specific (types of) actors; more like "This question/topic seems probably important somehow, and more clarity on it would probably somehow inform various important decisions."
  • E.g., the sort of work Nick Bostrom does
• Think-tank-like: Projects that we think would be valuable based on pretty explicit theories of change, ideally informed by actually talking to a bunch of relevant decision-makers to get a sense of what their needs and confusions are.
• Consultancy-like: Projects that one specific stakeholder (or I guess maybe one group of coordinated stakeholders) have explicitly requested we do (usually but not necessarily also paying the researchers to do it).

I think RP, the EA community, and the world at large should very obviously have substantial amounts of each of those three types of projects / theory of change.

I think RP specialises mostly for the latter two models, whereas (for example) FHI specialises more for the first model and sometimes the second. (But again, I'll let someone else at RP say more about specific percentages and rationales.)

(See also my slides on Theory of Change in Research, esp. slide 17.)

The short answer is that a) none of our past hires in longtermism (including management) had substantive  biosecurity experience or biosecurity interest and b) no major stakeholder has asked us to look into biosecurity issues.

The extended answer is pretty complicated. I will first go into why generalist EA orgs or generalist independent researchers may find it hard to go into biosecurity, explain why I think those reasons aren't as applicable to RP, and then why we haven't gone into biosecurity anyway.

Why generalist EA orgs or generalist independent researchers may find it hard to go into biosecurity

My personal impression is that EA/existential biosecurity experts currently believe that it's very easy for newcomers in the field to do more harm than good, especially if they do not have senior supervision from someone in the field. This is because existential biosecurity in particular is rife with information hazards, and individual unilateral actions can invoke the unilateralist's curse. 

Further, all the senior biosecurity people are very busy, and are not really willing to take the chance with someone new unless they a) have experience (usually academic) in adjacent fields or b) are credibly committed to do biosecurity work for a long period of time if they're a good fit. 

Since most promising candidates are understandably not excited to commit to doing biosecurity work for a long period of time without doing some work on it first, this creates a chicken-and-egg problem.

(Note again this is my own impression. Feel free to correct me, any biosecurity experts reading this!)

Why RP in particular may be a good place to start a biosecurity career anyway.

I think RP is institutionally trusted by the major groups enough to be careful if we were able to wade into biosecurity. In particular, we would be careful to not publish things that we think are potentially dangerous without running it by a few more experienced people first, and also we are credibly very willing to take down things quickly if we get a "cease and desist" from more experienced parties first (and then carefully reassess offline whether this was the correct move to do). 

On an individual level, I have a number of contacts with some of the key biosecurity people in EA, both through covid forecasting before joining RP, and socially.  In addition, I believe I can credibly pull off "non-expert making useful and not-dangerous contributions to biosecurity" as my covid forecasting and cultured meat analysis experiences have at least somewhat demonstrated an ability to provide value via reading, disseminating, and evaluating fairly technical work in adjacent domains (as a non-expert). 

So I'd maybe be excited to do biosecurity projects within my range of capabilities if stakeholders reached out to us with sufficiently important/interesting projects, or (more plausibly) advise colleagues/interns/contractors who can provide enough technical expertise while I provide the less technical guidance.

Why we haven't gone into biosecurity anyway

As you may have already inferred from past sentences, the biggest reason* is that none of our hires have had biosecurity experience or even strong interest. This is another chicken-and-egg problem. We haven't done biosecurity work because we don't have strong biosecurity hires, but we don't have strong (enough) biosecurity candidates applying because they don't see us doing biosecurity work. 

One of my planned ways around this was trying to get a biosecurity intern last summer, in the hopes that having public outputs in biosecurity by an intern would be a smooth way for us to both scale up our institutional biosecurity knowledge and also demonstrate our interest in this arena. The idea is that interns with the relevant backgrounds (eg math bio, or epidemiology) can provide the technical backgrounds while RP complements their skillsets with the relevant EA contacts, discretion, and analytical ability. 

I did try nontrivially hard to make this happen smoothly. I asked some promising biosecurity people to apply.I got verbal agreement from some FHI bio people to be a co-advisor to our biosecurity-interested interns if we had any. And some of the questions in our (blinded) intern assessment process should have differentially been easier for people with bio backgrounds.

But ultimately our strongest intern candidates last round neither had the relevant academic backgrounds nor were particularly interested in biosecurity. 

Next steps

RP's longtermism team is currently going through a hiring round. It seems plausible we might just have a strong biosecurity hire this round, in which case they'd lead our future biosecurity efforts in 2022 and this discussion is moot.

It also seems plausible to me if unlikely (~20% in the next 6 months?) for us to end up prioritizing biosecurity even without a strong biosecurity hire, whether due to internal cause prioritization or external stakeholder requests.  

At any rate, if you or others reading this want to support future RP biosecurity efforts, the best way to do this is encouraging strong biosecurity people you know to apply in future rounds! Funder interest is also helpful, but substantially less so.

*we also have internal disagreements about whether it makes sense for us to be more proactive about doing biosecurity work, a) given that we're already scattered pretty thin on many projects, b) focus is often good, and c) we internally disagree about how important marginal biosecurity work by people without technical expertise is anyway. I'm just presenting my own view.

As much as I like to imagine it's my own work (in longtermism), I think the clearest institutional comparative advantage of RP relative to the rest of the EA movement is the quality of our animal-welfare focused research. To the best of my knowledge, if you want to focus on doing research that directly improves the welfare of many animals, and you don't have a  long-chain theory/plan of impact (e,g. by shifting norms in academia or having an influential governmental position), RP's the best place to do this. This is just my impression, but my guess is that this is broadly shared among animal-focused EAs. 

The main exception I could think of is Open Phil, but they're not hiring.

I also get the impression that our survey team is very good, probably the best in EA, but I have less of an inside view here than for the animal welfare research. 

Our longtermism and global health work are comparatively more junior and less proven, in addition to having fairly stiff competition.

Research, especially EA-aligned research done based on an explicit theory of change.

Asked differently, why are you so cool, both at the RP level and personally?

This is a great question! Thank you so much!

At Rethink Priorities we take an employee-focused approach. We do our best to ensure that our staff have relevant tools and resources to do their best work, while also having enough flexibility to maintain their work-life balance. Staff happiness is a high priority for us and one of our strategic goals. 

Some aspects of our employee-centered approach include:

• Competitive benefits and perks – we offer unlimited time off, flexible work schedule, professional development opportunities, stipends etc., which are available to full- and part-time staff, as well as our fellows/interns.
• Opportunities to socialize, make decisions, and take on new projects – for example, we have monthly social meetings, we run random polls to solicit opinions/ideas from staff, and create opportunities for employees to participate in various initiatives, like leading a workshop.
• Biannual all staff surveys – we collect feedback from our staff twice a year. The survey asks a series of questions about leadership, management, organizational culture, benefits and compensation, psychological safety, amongst others. The results are thoroughly analyzed and guide our decisions about how to improve our culture, moving forward.
• Positive environment – we foster an inclusive and welcoming environment in which we encourage individuals to pose their questions, provide feedback, share thoughts, and raise concerns; additionally, we practice transparency at RP with regards to all aspects of our operations (e.g., decision-making, salary).
• Internal processes – we continuously revise and/or develop internal processes and practices to ensure equity across the entire organization (e.g. we have recently audited our hiring procedures to increase equity and reduce bias when selecting candidates). 
• Reflection – we reflect on how we do our work, how we interact with one another, what culture we aspire to develop, and implement necessary changes.

We’re working right now on a values and culture setting exercise where we are figuring out intentionally what we like about our culture and what we want to specifically keep. I appreciate Dominika's comment but I want to add a bit more of what is coming out of this (though it isn't finished yet).

Four things I think are important about our culture that I like and try to intentionally cultivate:

Work-life balance and sustainability in our work. Lots of our problems are important and very pressing and it is easy to burn yourself out working hard on it. We have deliberately tried to design our culture for sustainability. Sure, you might get some more hours of work this year if you work harder but it isn’t worth burning out just a few years later. We want our researchers here for the long haul. We’re invested in their long-term productivity.

Rigor and calibration. It’s very easy to do research poorly and unfortunately easy to do bad research that misleads people because it is hard to see how the research is bad. Thus a lot of work must be done by our researchers to ensure that our work is accurate and useful.

Ownership. In a lot of organizations, managers want their employees to do exactly what they are told and follow processes to the letter. At Rethink Priorities, we think the ideal employee instead seeks to understand the motivation behind the assignment and how it fits into our goals and notices if there is a better way to achieve the same goals or even if the project shouldn’t be done.

Working on the right things. There are a lot of problems that we need to solve, so we must prioritize them. Selecting the right research question can often be more impactful than answering it.

We'll have something more finished at a later date!

One major factor that makes some research questions more suited to academia is requiring technical or logistical resources that would be hard to access or deploy in a generalist EA org like RP (some specialist expertise also sometimes falls into this category). Much WAW research is like this, in that I don't think it makes sense for RP to be trying to run large-scale ecological field studies.

Another major factor is if you want to promote wider field-building or you want the research to be persuasive as advocacy to certain audiences in the way that sometimes only academic research can. This also applies to much WAW research.

Personally, I think in a most other cases academia is typically not the best venue for EA research, although the latter considerations about field-building and the prestige/persuasiveness of academic research recurs sufficiently commonly that I think the question of whether a given project is worth publishing academically recurs fairly commonly even within RP.

Commenting on our public output, particularly if you have specialized technical expertise, can often be somewhere from mildly to really helpful. RP has a lot of knowledge, but so does the rest of the EA community and extended EA network, so if you can route our reports to the relevant connections, this can be really valuable in improving the quality of our reasoning and epistemics.

One thing the EA community can help us with is by encouraging  suitable candidates to apply to our jobs. (New ones will be posted here and announced in our newsletter.) Some of our most recent hires have transitioned from fields which, at first sight, would seem unlikely to produce typical applicants. But we're open to anyone proving us they can do the job during the application process (we do blinded skills assessments). I think we're really not credentialist (i.e. we don't care much about formal degress if people have gained the skills that we're looking for). So whenever you read a job ad and think "Oh, this friend could actually do that job!", do tell them to apply if they're interested.

More importantly, I think EA community builders in all geographies and fields can greatly help us by training people to become good at the type of reasoning that's important in EA jobs. I particularly think of reasoning transparency,  expressing degrees of (un)certainty and clarifying the epistemic status of what you write. Furthermore, probabilistic thinking and  Bayesian updating. Also learning to build models and getting familiar with tools like Guesstimate and Causal. Forecasting also seems to be a valuable skill to train (e.g. on Metaculus). I think EAs anywhere in the world can set up groups where people train such skills together. 

I recently spent ~2 hours reflecting on RP's longtermism department's wins, mistakes, and lessons learned from our first year[1] and possible visions for 2022. I'll lightly adapt the "lessons learned for Michael specifically" part of that into a comment here, since it seems relevant to what you're trying to get at here; I guess a more effective person in my role would match my current strengths but also already be nailing all the following things. (I guess hopefully within a year I'll ~match that description myself.)

(Bear in mind that this wasn't originally written for public consumption,  skips over my "wins", etc.)

• "Focus more
  • Concrete implications:
    • Probably leave FHI (or effectively scale down to 0-0.1 FTE) and turn down EA Infrastructure Fund guest manager extension (if offered it)
    • Say no to side things more often
    • Start fewer posts, or abandon more posts faster so I can get other ones done
    • Do 80/20 versions of stuff more often
    • Work on getting more efficient at e.g. reviewing docs
  • Reasons:
    • To more consistently finish things and to higher standards (rather than having a higher number of unfinished or lower quality things)
    • And to mitigate possible stress on my part, [personal thing], and to leave more room for things like exercise
    • And to be more robust against personal life stuff or whatever
      • (I mean something like: My current slow-ish progress on my main tasks is even with working parts of each weekend, so if e.g. I had to suddenly fly back to Australia because a family member was seriously ill, I’d end up dropping various balls I’ve somewhat committed to not dropping.)
• Maybe trust my initial excitement less regarding what projects/posts to pour time into and what ideas to promote, and relatedly put more effort into learning from the views and thinking of more senior people with good judgement and domain expertise 
  • E.g., focus decently hard on making the AI gov stuff go well, since that involves doing stuff Luke thinks is useful and learning from Luke
  • E.g., it was good that I didn't bother to finish and post my research question database proposal
• Maybe pay more attention to scale and relatedly to whether an important decision-maker is likely to actually act on this
  • Some people really do have a good chance of acting in very big ways on some stuff I could do
  • But by default I might not factor that into my decisions enough, instead just being helpful to whoever is in front of me or pursuing whatever ideas seem good to me and maybe would get karma
• Implement standard productivity advice more, or at least try it out
  • I’ll break this down more in the habits part of my template for meetings with Peter
  • [I'm also now trying productivity coaching]
• Spend less time planning projects in detail, and be more aware things will change in unexpected ways
• Be more realistic when making plans, predictions, and timelines
  • (No, really)
  • Including assuming management will take more time than expected, at least given how I currently do it
• Spend more time, and get better at, forming and expressing hot takes
• Spend less time/words comprehensively listing ideas/considerations/whatever
• More often organise posts/docs conceptually or at least by importance rather than alphabetically or not at all
• Be more strict with myself regarding exercise and bedtime
• Indeed optimise a fair bit for research management careers rather than pure research careers
  • This was already my guess when I joined, but I’ve become more confident about it"

[1] I mean the first year of the current version of RP's longtermism department; Luisa Rodriguez previously did (very cool!) longtermism work at RP, but then there was a gap between her leaving (as a staff member; she's now on the board) and the current staff joining. 

Some ways someone can be more effective than me:

• I'm not as aggressive at problem/question/cause prioritization as I could be. I can see improvements of 50-500% for someone who's (humanly) better at this than me.
• I'm not great at day-to-day time management either. I can see ~100% improvement in that regard if somebody is very good at this.
• I find it psychologically very hard to do real work for >30h/week, so somebody with my exact skillset but who could productively work for >40h/week without diminishing returns would be >33% more valuable.
• I pride myself of the speed and quantity I write, but I'm slower than eg MichaelA, and I think it's very plausible that a lot of my outputs are still bottlenecked by writing speed. 10-50% effectiveness improvement seems about right.
• I don't have perfect mental health and I'm sometimes emotional. (I do think I'm above average at both). I can see improvements of 5-25% for people who don't have these issues.
• I'm good at math* but not stellar at it. I can imagine someone who's e.g. a Putnam Fellow be 3-25% more effective than me if they chose to work on the same problems I work on (though plausibly they'd be more effective because they'd gravitate towards much mathier problems; otoh ofc not all/most mathy problems are very important)
• Relatedly, obviously I'm not the smartest person in the world. I don't have a good sense of how much e.g. being half a standard deviation smarter than me would make someone a better researcher, anything from "not a lot" to "very high" seems plausible to me. ??? for quantitatively how much effectiveness this adds.

*Concretely, I did a math major in a non-elite liberal arts college, which wasn't too hard for me. I perceived both my interns last summer as probably noticeably better at math than me (one was a math major at Columbia and the other at MIT). Certainly they know way more math.

The person who replaces me has all my same skills but in addition has many connections to policymakers, more management experience, and stronger quantitative abilities than I do.

I've adjusted imperfectly to working from home, so anyone  who has that strength in addition to my strengths would be better. I wish I knew more forecasting and modeling, too. 

(less helpful answer, will think of a better one later) 

Hmm Rethink follows pretty reasonable management practices, and is maybe on the conservative side for things like firing unproductive employees. 

You're fired tomorrow...

So I can't really imagine being fired for ineffectiveness without warning  on a Saturday. The only way this really happens is if I'm credibly accused  of committing a pretty large crime or sexually harassing a RP colleague or maybe faking data or something like that. 

To the best of my knowledge I have not done these things.

...and replaced by someone more effective than you. What do they do that you're not doing?

Hmm since I haven't done these things, I must be set up to be falsely accused for a crime in a credible way. So the most likely way someone can replace me and be more effective on this dimension is by not making any enemies who's motivated enough to want to set them up for murder or something.

Quick clarifying question: 

Is the most important part of your question the "fired" part or the "more effective" part? Like would you rather I a) answer by generating stories of how I might be fired and how somebody can avoid that, or b) answer what can people do to be more effective than me?

Put differently, is there anything that you think aspiring hires might want to start working on to be better suited to join/support RP over the next few years?

I'll let my colleagues answer the object-level question/might answer it myself later if I get better ideas later, but broadly I would somewhat caution against having a multi-year plan to be employed at Rethink Priorities specifically (or at any specific organization). RP hiring is pretty competitive now and has gotten more competitive over time[1], and also our hiring processes are far from perfect so even very good researchers (by our lights) may well be missed by our hiring process. 

That said, some of the answers to James Ozden's question might be relevant here as well. 

[1] We're also scaling pretty quickly to hire more people, but EA community building/recruitment at top universities have also really scaled up since 2020, and it's unclear how these things shake out in terms of how competitive our applications will be in a few years.

I once told people in a programmer group chat what I was doing when I got my new job at RP. One of them looked into the website and gave like a $10 donation. 

To the best of my limited knowledge, this might well be our largest non-EA aligned donation in longtermism. 

It's a little hard to say because we don't necessarily know the background / interests of all donors, but my current guess is around 2%-5% in 2021 so far. It's varied by year (we've received big grants from non-EA sources in the past). So far, it is almost always to support animal welfare research (or unrestricted, but from a group motivated to support us due to our animal welfare research).

One tricky part of separating this out - there are a lot of people in the animal welfare community who are interested in impact (in an EA sense), but maybe not interested in non-animal EA things.

I'm particularly glad you  note this since the survey team's research in particular is almost exclusively non-public research (basically the EA Survey and EA Groups Survey are the only projects we publish on the Forum), so people understandably get a very skewed impression of what we do.

Thanks! We'll make sure to get this changed going forward.

I would break this down into a) the methods for getting research in front of government orgs and b) the types of research that gets put in front of them.

In general I think we (me for sure) haven’t been optimising for this enough to even know the barriers (unknown unknowns). I think historically we’ve been mostly focused on foundations and direct work groups, and less on government and academia. This is changing so I expect us to learn a lot more going forward.

As for known unknowns in the methods, I still don’t know who to actually send my research to in various government agencies, what contact method they respond best to (email, personal contact, public consultations, cold calling, constituency office hours?), or what format they respond best to (a 1 page PDF with graphs, a video, bullet points, an in person meeting? - though this public guide Emily Grundy made on UK submissions while at RP has helped me). Anecdotally it seems remarkably easy to get in front of some: I know of one small animal advocacy organization that managed to get a meeting with the Prime Minister of their country, and I myself have had 1-1 meetings with more than two dozen members of the UK and Irish parliaments and United Nations & European Union bureaucrats (non RP work) with relative ease (e.g. an email with a prestigious sounding letterhead).

My assumption is government orgs are swamped with requests and petitions from NGOs, industry, peers, constituents. So we need some way to stand out from the crowd like representing a core constituency of theirs, being recommended to them by someone they deem credible such as an already established NGO, being affiliated with some already credible institution like a prestigious university, and proving to them we can provide them with policy expertise and legislative intelligence better than most others can.

On b) I think have a better sense of what content would be more likely to get in front of them. Niel Bowerman had some good insights on this in 2014, and the “legislative subsidy” approach Matthew Yglesias favours in the US context seems useful.There was an interesting study from Nakajima (2021) (twitter thread) which looked at what kinds of research evidence do policymakers prefer (bigger samples, external validity extends to the population in their jurisdictions, no preference on observational-v-experimental) so I think we can explore whether the topics on our research agenda fit within those designs.

Update: wanted to add in this post from Zach Groff:

1. Happily, evidence does seem to affect policy, but in a diffuse and indirect way. The aforementioned researcher Carol Weiss finds that large majorities (65%-89%) of policymakers report being influenced by research in their work, and roughly half of them strongly (Weiss 1980; Weiss 1977). It's rare that policymakers pick up a study and implement an intervention directly. Instead, officials gradually work evidence into their worldviews as part of a gradual process of what Weiss calls "enlightenment" (Weiss 1995). Evidence also influences policy in more political but potentially still benign ways by justifying existing policies, warning of problems, suggesting new policies or making policymakers appear self-critical (Weiss 1995; Weiss 1979; Weiss 1977).

2. There are a few methods that seem to successfully promote evidence-based policy in health care, education, and government settings where they have been tested. The top interventions are:

2a) Education—Workshops, courses, mentorship, and review processes change decision makers' behavior with regard to science in a few studies (Coburn et al. 2009; Matias 2017; Forman-Hoffman et al. 2017; Chinman et al. 2017; Hodder et al. 2017).

2b) Organizational structural changes—If an organization has evidence built into its structure, such as having a research division and hotline, encouraging and reviewing employees based on their engagement with research, and providing funding based on explicit evidence, this seems to improve the use of evidence in the organization (Coburn and Turner 2011; Coburn 2003; Coburn et al. 2009; Weiss 1980; Weiss 1995; Wilson et al. 2017; Salbach et al. 2017; Forman-Hoffman et al. 2017; Chinman et al. 2017; Hodder et al. 2017). A few other methods for promoting research-backed policies seem promising based on a bit less evidence:

2c) Increasing awareness of evidence-based policy—Sending employees reminders or newsletters seems to increase research-based medicine based on two high-quality review papers (Murthy et al. 2012; Grimshaw et al. 2012). . Similarly, all-around advocacy campaigns to promote evidence-based practices among practitioners achieves substantial changes in one randomized controlled trial (Schneider et al. 2017).

2d) Access—Merely giving people evidence on effectiveness does not generally affect behavior, but when combined with efforts to motivate use of the evidence, providing access to research does improve evidence-based practice (Chinman et al. 2017; Wilson et al. 2017).

2e) External motivation and professional identities— Two recent RCTs and a number of reviews and qualitative research find that rewarding people for using evidence and building professional standards around using research are helpful (Chinman et al. 2017; Schneider et al. 2017;Hodder et al. 2017; Forman-Hoffman et al. 2017; Weiss et al. 2005; Weiss 1995; Wilson et al. 2017; Weiss 1980; Weiss 1977; Matias 2017; Coburn 2005; Coburn 2003).

3. Interestingly, a few methods to promote evidence-based practices that policymakers and researchers often promote do not have much support in the literature. The first is building collaboration between policymakers and researchers, and the second is creating more research in line with policymakers' needs One of the highest-quality write-ups on evidence-based policy, Langer et al. 2016 finds that collaboration only works if it is deliberately structured to build policymakers' and researchers' skills. When it comes to making research more practical for policymakers, it seems that when policymakers and researchers work together to come up with research that is more relevant to policy, it has little impact. This may be because, as noted in point (1), research seems to influence policy in important but indirect ways, so making it more direct may not help much.

4. There is surprisingly and disappointingly little research on policymakers' cognition and judgment in general. The best research is familiar to the effective altruism community from Philip Tetlock (1985; 1994; 2005; 2010; 2014; 2016) and Barbara Mellers (2015), and it gives little information on how decision-makers respond to scientific evidence, but suggests that they are not very accurate at making predictions in general. Other research indicates that extremists are particularly prone to overconfidence and oversimplification, and conservatives somewhat more prone to these errors than liberals (Ortoleva and Snowberg 2015; Blomberg and Harrington 2000; Kahan 2017; Tetlock 1984; Tetlock 2000). Otherwise, a little research suggests that policymakers in general are susceptible to the same cognitive biases that affect everyone, particularly loss aversion, which may make policymakers irrationally unwilling to end ineffective programs or start proven but novel ones (Levy 2003; McDermott 2004). On the whole, little psychological research studies how policymakers react to new information.

If anyone reading this works at a governmental organization, we’d love to chat!

The first part I answered here.

I think a major success for us would look like having achieved a large and sustainably productive research organization tackling research in a variety of disciplines and cause areas. I think we will have made a major contribution to unlocking funding in effective altruism by figuring out to fund with more confidence as well as increasing our influence across a larger variety of stakeholders, including important stakeholders outside of the effective altruism movement..

One recent experiment has been trying to get better at project management, especially at a larger scale. We’ve rolled out Asana for the entire organization and have hired a project manager.

Another recent experiment has been whether we can directly hire for “Senior Research Managers” (SRMs), instead of having to develop all our senior research talent in-house. We’ve hired two external SRMs and it has been going well so far, but it is too early to tell. We may try to hire another external SRM in our current hiring process.

If both these two experiments go well, it will unlock a lot of future scalability for our organization and for other organizations that can follow suit.

Our next experiment will likely involve hiring research and/or executive assistants to see if they can help our existing researchers achieve more productivity in a more sustainable way.

It has been said before elsewhere by Peter, but worth stating again:read and practice Reasoning Transparency . Michael Aird compiled some great resources recently here.

I'd also refer people to Michael and Saulius' replies to arushigupta's similar subquestion in last year's RP AMA.

For longtermist work, I often point people to Holden Karnofsky's impressions on career choice, particularly the section on building aptitudes for conceptual and empirical research on core longtermist topics .

I've also personally gained a lot from arguing with People Wrong on the Internet, but poor application of this principle may be generally bad for epistemic rigor. In particular, I think it probably helps to have a research blog and be able to do things like spot potential holes in (EA social media, EA forum, research blogs, papers, etc). That said, I think most EA researchers (including my colleagues) are much less Online than I am, so you definitely don't need to develop an internet argument habit to be a good researcher.

Making lots of falsifiable forecasts about short-term conclusions of your beliefs may be helpful. Calibration training is probably less helpful, but lower cost.

Trying to identify important and tractable (sub)questions is often even more important than the ability to answer them well. In particular, very early on in a research project, try to track "what if I answered this question perfectly? Does it even matter? Will this meaningfully impact anyone's decisions, including my own? Will this research build towards something else that will meaningfully impact decisions later?"

"Politely disagreeable" seems like a pretty important disposition. You benefit epistemically from being nice and open enough to other people's ideas that you a) deliberately seek out contrarian opinions and b) don't reject them outright, but also you need to be disagreeable enough that you in general shouldn't update on beliefs just because other (smart, respected, experienced, etc) people confidently believe it. 

Being very aggressively truth-seeking is a really important disposition. My belief is that most people are by default bad at this, including people who may otherwise make great EA researchers.

I also endorse Neil's comment.

I don’t think it is actually possible to 10x our impact with the same staff, funding, and other resources - hence our desire to hire and fundraise more. If it was possible, we’d certainly try to do that!

The best answer I can think of is Goodharting - we certainly could influence more total dollars if we cared less about the quality of our influence and the quality of those dollars. We also could exaggerate our claims about what “influence” means, taking credit for decisions that likely would’ve been made the same anyway.

Lazy semi-tangential reply: I recently gave a presentation that was partly about how I've used forecasting in my nuclear risk research and how I think forecasting could be better used in research. Here are the slides and here's the video. Slides 12-15 / minutes 20-30 are most relevant. 

I also plan to, in ~1 or 2 months, write and publish a post with meta-level takeaways from the sprawling series of projects I ended up doing in collaboration with Metaculus, which will have further thoughts relevant to your question.

(Also keen to see answers from other people at RP.)

We at Rethink Priorities definitely have made an increasingly large effort to include forecasting in our work. In particular, we just recently have been running a large Nuclear Risks Tournament on Metaculus. My guess is that the reasons we don’t have even more forecasting relates to not all of our researchers being experienced forecasters and it hasn’t been a sufficient priority to generate sufficiently useful and decision-relevant forecasting questions for every research piece.

We have not yet decided whether we will have internships / fellowships this summer - assuming you are referring to the Northern Hemisphere here. If we launch these internships, I imagine they will open in 2022 March. We are continuing to consider launching internships / fellowships for summer in each Hemisphere (as we launched an AI Governance and Fellowship for 2022 Jan-March for summer in the Southern Hemisphere).

Another thing we are considering in addition to, or in replacement of, internships this year is Research/Executive Assistant positions that focus more on supporting and learning the work of a particular researcher on the RP team. These roles would likely be permanent/indefinite in length rather than a few months like our internships have been.

I am also interested in future internship plans.  Specifically, how flexible are the dates and time commitments?

As someone based in Australia, seasonal descriptors (presumably from the Northern hemisphere) aren't ideal though I can convert them - specific months would be preferable :)  Also our university holiday periods are different, so I will need to work around that too.

Longtermism in its nascent form relies on a lot guesstimates and abstractions that I think could be made more empirical and solid. Personally, I am very interested in asking whether people at x time in the past had the information they needed to avoid later disasters that occurred. What kinds of catastrophes have humans been able to foresee,  and when we were able to but didn't, what obstacles were in the way? History is the only evidence available in a lot of longtermist domains and I don't see EA exploiting it enough. 

As is probably the case with many researchers, I have a bunch of thoughts on this, most of which aren't written up in nice, clear, detailed ways. But I do have a draft post with nuclear risk research project ideas, and a doc of rough notes on AI governance survey ideas, so if someone is interested in executing projects like that please message me and I can probably send you links. 

(I'm not saying those are the two areas I think are most impactful to do research on on the current margin; I just happen to have docs on those things. I also have other ideas less easily shareable right now.)

People might also find my central directory for open research questions useful, but that's not filtered for my own beliefs about how important-on-the-margin these questions are.

A note on why climate change is currently in our global health and development work rather than longtermism - the main reasons for this is that while we could consider longtermist work on climate change we do not think marginal longtermist climate change work makes sense for us relative to the importance and tractability of other longtermist work we could do. However, global health and development funders and actors are also interested in climate change in a way that does not funge much against longtermist money or talent, and the burden of climate change is felt heavily on lower and middle income countries. Therefore we think climate change work makes sense to explore relative to other global health and development opportunities.

Hi James, thanks for your question. The climate change work currently on our research calendar includes:

1. A look at how climate damages are accounted for in various integrated assessment models
2. A cost effectiveness analysis of anti-deforestation interventions
3. A review of the landscape of climate change philanthropy
4. An analysis of how scalable different carbon offsetting programs are