Update from Open Philanthropy’s Longtermist EA Movement-Building team

post by ClaireZabel · 2022-03-10T19:37:27.283Z · EA · GW · 19 comments


  What’s happened so far 
    Metrics of impact
  Other changes that seem important to me
  Our mistakes
  Mistakes I’m worried we will make
  Looking forward


For much more detail on all of this, see the rest of the post.


This post is a report and update on the Open Philanthropy Longtermist Effective Altruism Movement-Building team’s thinking and goals. It’s written by me, Claire, and mostly represents my perspective. I’m writing this in my role as an Open Phil staff member, but I take sole responsibility for the angsty commentary near the bottom. 

 Our team currently consists of me, Asya Bergal, Bastian Stern, and Eli Rose. We are supported by Open Phil’s “longtermist budget” (funding to support projects motivated by the longtermist view), but unlike the other longtermist cause areas, we aren’t aiming to make progress on longtermist priorities directly. Instead, our goal is to grow and support the pool of people motivated and well-positioned to work on longtermist priority projects (e.g. reducing existential risk and aiming to improve the far future). [1]

I think our team’s grantmaking has high expected value, because (1) in my experience, most of the relevant object-level longtermist projects are bottlenecked by the dearth of aligned people who are good fits (so our goals are aimed at a core problem), (2) there’s a lot of funding to direct (which is bottlenecked by the number of grantmakers working to direct it), and (3) we have a reasonably high number of potential grantmaking projects we are working as fast as we can to implement (i.e. there’s a feeling of traction) and could implement faster if we had more capacity on the team. Not coincidentally, we are currently hiring for several roles.  

What’s happened so far 

I took over this area from Nick Beckstead (who was working on it part-time) in early 2019. Bastian started working with me and Eli joined in 2020, and Asya joined in 2021. 

I think a lot has changed since then, and a lot of important changes are ongoing. We committed funds equaling ~$17M in the area in 2019, ~$26M in 2020, and ~$60M in 2021. So far in 2022, we’ve already committed >$65M (though some key aspects of the relevant grants are still TBD[2]), so we are on track to continue to vastly increase our giving. I hope and believe that if we hire more strong grantmakers, we can double giving in this area several more times (i.e., there’s sufficient funder interest and are/will be worthy opportunities). 

As I’m going to discuss a bit below, I’ve shifted further away from focusing on “money moved” figures, and I think they can be misleading proxies for impact. Even among funding that is “above the bar” from a financial perspective[3], I think the top decile is at least an order of magnitude more impactful (per dollar) than the bottom decile grantmaking that we’re doing. In other words, seeing that more money has been moved doesn’t tell you much unless you have a sense of where it falls along the cost-effectiveness spectrum (or time-effectiveness spectrum). 

Right now, the time of aligned longtermists working on high-priority projects seems to be the scarcer resource and perhaps the more useful metric to focus on. Still, “money moved” is one of the easier figures to report, and I reported it above because I think it mostly tracks the more meaningful but less measurable growth in projects and funding opportunities my team is working with. 

Over the last few years, my thinking about my role and the role of other longtermist grantmakers has shifted significantly. In the past, I spent a lot more time working on the question: “How can I tell if a funding opportunity in the longtermist meta space meets the bar?” Nowadays, we spend less time on evaluation and more time creating new funding opportunities to achieve our main goal (growing and supporting people who help with longtermist priority projects).[4] 

There were a few reasons for this change:

Metrics of impact

The above points led me to think that in fact, going forwards, grants of the kind we were making will likely be substantially “above the (new) bar”. 

That updated my views in several ways: 

So, we’re shifting to prioritize our and our grantees’ time more highly. And we’ve been creating different kinds of open calls[4] for people to request different kinds of fairly short-term support[5] (in contrast to grants where we support existing organizations).  For these, we’ve started thinking in terms of how much quality-weighted longtermist output we think they’ll produce per hour we put into them, rather than focusing primarily on output per dollar. (In an ideal world, we’d have a conversion factor we trusted between longtermist time and dollars and be able to get an aggregate longtermist resource cost estimate; this is more of a heuristic about which factor will tend to dominate for the kinds of decisions we’re making, given the situation we find ourselves in). 

On the one hand, I think giving in this category (the short-term support, including for less EA-engaged individuals) tends to be less impactful per dollar compared to many other outreach activities aimed at less EA-engaged people.[6]  But, I think it can have more positive impact per hour of EA (grantmaker and grantee) labor used. 

For example, when we fund e.g. 80,000 Hours, we (amongst other activities) support their full-time advisors to advise interested people about how to have more impactful careers. With our scholarship programs, we’re also trying to cause people to spend more time on more impactful activities. But rather than do this via the 80k advisors, our scholarship programs use money “directly” (without much intermediating EA labor) to try to make impactful careers more accessible and attractive. In general, we think we get less impact per dollar from interventions that consume money “directly” like this. Since EA labor is the scarcer resource in many contexts, these types of interventions can make sense for grantmakers to prioritize.  

I think it’s good for people starting projects of various kinds to think through not just monetary costs, but also the amount of aligned EA labor required to make a project work well. However, I expect most important longtermist projects to consume a ton of EA labor (including high opportunity cost labor), and am worried many newer EAs are already too hesitant to ask for support and advice because of personal and professional underconfidence, so it’s confusing. 

Other changes that seem important to me

I’m not going to try to justify these now, just share my impressions sans evidence or explanation.

Our mistakes

By which I mostly mean “my mistakes”, given the relative recency of my teammates joining and getting up to speed, and my responsibility for final calls about team strategy and direction. 

I think:

On a meta level, I think most of my mistakes revolve around being unnecessarily slow to reorient around a change, so I’m trying to address that pattern by, when I notice myself having the thought that we might be erring slightly in some direction, trying to more quickly evaluate the hypothesis that we might actually be erring substantially and that fixing it should be a top priority.

The other, weaker pattern I noticed was being bottlenecked on emotional pain tolerance and reputational concerns (e.g. related to advocating for very uncertain grants for which I have little evidence when they have a reasonable probability of going poorly, or making riskier hiring decisions which might end in mutual unhappiness). 

Mistakes I’m worried we will make

Looking forward

Over the next few years, I expect us to spend more time on projects engaging with high-school students (largely for the reasons listed here [EA · GW]) as well as working more directly with community-building efforts aimed at undergraduates. 

If we found the right people (we’re hiring!) I could also imagine us spending tens of millions of dollars more on the following projects, which could easily end up seeming similarly cost-effective as our previous grantmaking:

  1. ^

     Sometimes, our team supports projects that aren't directly aimed at these priorities, often because we think their value from a movement-building perspective is sufficiently high that it justifies supporting them (i.e. in those cases we might have different motives for supporting a project than the people who work on it have for working on it.)

  2. ^

    Two caveats about this though:

    1. A relatively small fraction of the funding is for different regranting programs, and so has not in fact yet “bottomed out”, and will absorb more grantmaker labor before that happens (and might be reported by another entity as part of their money moved in the future).
    2. This is somewhat driven by unusually large outliers. However, grants that were previously outliers in terms of size are becoming more common. I’m left pretty uncertain about how much we should expect to give this year.
  3. ^

     I.e. in expectation a better use of funding than the longtermist last dollar. See here for a discussion about last dollars in the global health and wellbeing space.

  4. ^

    So far, this includes our RFP for outreach projects, course development program, early-career funding, undergraduate scholarship, with more to come. The FTX Foundation Future Fund also currently has an open round

  5. ^

     I’d love to have a better name for this category, suggestions welcome

  6. ^

     There are also programs that support highly EA-engaged individuals, which I think can be really impactful per dollar and hour, but there’s a limited number of such people and so only so much financial support to provide.

  7. ^

     Occasionally, spending more time can lead one to realize that a grant is actually really promising or really net negative, but I think that’s pretty rare.


Comments sorted by top scores.

comment by Akash · 2022-03-12T00:01:07.939Z · EA(p) · GW(p)

Thank you for this write-up, Claire! I will put this in my "posts in which the author does a great job explaining their reasoning" folder.

I noticed that you focused on mistakes. I appreciate this, and I'm also curious about the opposite:

  • What are some of the things that went especially well over the last few years? What decisions, accomplishments, or projects are you most proud of?
  • If you look back in a year, and you feel really excited/proud of the work that your team has done, what are some things that come to mind? What would a 95th+ percentile outcome look like? (Maybe the answer is just "we did everything in the Looking Forward" section, but I'm curious if some other things come to mind).
Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-15T03:14:41.983Z · EA(p) · GW(p)

Thanks Akash. I think you're right that we can learn as much from successes and well-chosen actions as mistakes, and also it's just good to celebrate victories. A few things I feel really pleased about (on vacation so mostly saying what comes to mind, not doing a deep dive): 

  • My sense is that our (published and unpublished) research has been useful for clarifying my picture of the meta space, and helpful to other organizations (and led to some changes I think are pretty promising, like increased focus on engaging high schoolers who are interested in longtermist-related ideas, and some orgs raising salaries), though I think some of that is still TBD and I wish I had a more comprehensive picture.
  •  We've funded just a bunch of new initiatives I'm quite excited about, and I'm happy we were there to find worthy projects with funding needs and encourage founding new projects in the space, and to support their growth. My best guess is that projects we fund will lead to a substantial increase in the EA/longtermist community. 
  • When I look back at both my portfolio of grants made, and anti-portfolio (grants explicitly considered but not made), I mostly feel very satisfied. As far as I can tell were far more false positives (grants we made that had meh results) than negatives (grants I think we should have made but didn't), but roughly similar false-negatives-that-seem-like-big-misses to false-positives-that-were-actively-meaningfully-harmful (the sample size in both of those categories is pretty small). 
  • I like and respect everyone on my team, they are all sincerely aimed at the real goals we share, and I think they all bring different important focuses and strengths to the table. 

If you look back in a year, and you feel really excited/proud of the work that your team has done, what are some things that come to mind? What would a 95th+ percentile outcome look like? (Maybe the answer is just "we did everything in the Looking Forward" section, but I'm curious if some other things come to mind.)

A mixture of "not totally sure" and "don't want to do a full reveal" but the "Looking Forward" section above lists a bunch of components. In addition: 

  • We or other funders seize most of the remaining the remaining obvious-and-important-seeming opportunities for impactful giving (that I currently know of in our space) that are lying fallow. 
  • We complete a few pieces of research/analysis I think could give us a better sense of how overall-effective EA/LT "recruiting" work has been over the last few years and how it compares to more object-level work (and we do indeed get a better sense and disseminate it to people who will find it useful). 
  • We gather and vet more resources for giving grantees that want it more non-financial support (e.g. referrals for support for various kinds of legal advice, executive and management coaching.)
comment by Miranda_Zhang (starmz12345@gmail.com) · 2022-03-11T02:39:48.213Z · EA(p) · GW(p)

After your talk at the SERI Conference, I really enjoyed reading this more detailed write-up of your recent updates. I'd be keen to see an update on how the Longtermist EA Movement-Building team ends up trying to address the concerns you're worried about!

In particular, I share concerns around the possibility that grant evaluation could become increasingly affected by more visible signals like certain names or reputations. To me, this is not only by default a concern considering the small size of the EA community, but also seems more of a risk with longtermist causes or other projects where EA alignment seems especially important and it is cheaper to rely on reputation (rather than investigating whether an unknown applicant is sufficiently aligned).

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-17T00:47:19.317Z · EA(p) · GW(p)

Thanks Miranda, I agree these are things to watch really closely for. 

comment by weeatquince · 2022-03-15T16:29:03.143Z · EA(p) · GW(p)

Hi Claire,

Thank you for the write-up. I have a question I would love to hear your (and other people's) thoughts on. You said:

I should have hired more people, more quickly. And, had a slightly lower bar for hiring in terms of my confidence that someone would be a good fit for the work, with corresponding greater readiness to part ways if it wasn’t a good fit.

This is really interesting as goes against the general tone of advice that I hear that suggests that being cautious about hiring. That said I do feel at times that the EA community is perhaps more cautious and puts more effort into hiring than other places I have worked.

I wondered if you had any elaboration, such as: advise on how someone at an EA org can tell if they are being too cautious? When you felt you should have taken more risks? What things it is worth taking risks on and what things it is not worth taking risks on? How you plan to change wat you do going forward?

No worries if nothing to add but it would be helpful to hear (I am involved right now in hiring decisions at a few EA orgs).

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-17T05:32:07.356Z · EA(p) · GW(p)

So to start, that comment was quite specific to my team and situation, and I think historically we've been super cautious about hiring (my sense is, much moreso than the average EA org, which in turn is more cautious than the next-most-specific reference class org). 

Among the most common and strongest pieces of advice I give grantees with inexperienced executive teams is to be careful about hiring (generally, more careful than I think they'd have been otherwise), and more broadly to recognize that differences in people's skills and interests leads to huge differences in their ability to produce high-quality versions of various relevant outputs. Often I find that new founders underestimate those differences and so e.g. underestimate how much a given product might decline in quality when handed from one staff member to a new one. 

They'll say things like "oh, to learn [the answer to complicated question X] we'll have [random-seeming new person] research [question X]" in a way that feels totally insensitive to the fact that the question is difficult to answer, that it'd take even a skilled researcher in the relevant domain a lot of time and trouble, that they have no real plan to train the new person or evidence the new person is unusually gifted at the relevant kind of research, etc., and I think that dynamic is upstream of a lot of project failures I see. I.e. I think a lot of people have a kind of magical/non-gears-level view of hiring, where they sort of equate an activity being someone's job with that activity being carried out adequately and in a timely fashion, which seems like a real bad assumption with a lot of the projects in EA-land. 

But yeah, I think we were too cautious nonetheless. 

Cases where hiring more aggressively seems relatively better: 

  • The upside is large (an important thing is bottlenecked on person-power, and that bottleneck is otherwise excessively challenging to overcome) 
  • The work you need done is:
    •  Well scoped,
    • Easy to evaluate 
    • Something people train in effectively outside your org
    • Trainable
    • Has short feedback loops
  • You are 
    • An experienced manager 
    • Proficient with the work in question
    • Emotionally ready to fire an employee if that seems best 
  • This is taking place in a country where it's legally and culturally easier to fire people
  • Your team culture and morale is such that a difficult few months with someone who isn't working out is unlikely to deal permanent damage. 
Replies from: weeatquince
comment by weeatquince · 2022-03-17T08:48:47.485Z · EA(p) · GW(p)

Really helpful. Good to get this broader context. Thank you!!

comment by James Ozden (JamesOz) · 2022-03-11T01:53:04.552Z · EA(p) · GW(p)

Thanks for writing this up, I found the transparency around your perceived mistakes and future uncertainty incredibly refreshing and inspiring!

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-12T01:42:27.543Z · EA(p) · GW(p)

Thanks for the kinds words, James!

comment by michelle_ma · 2022-03-11T02:52:52.791Z · EA(p) · GW(p)

Thanks for posting! Your discussion of mistakes and rationality-and-epistemics-focused community-building reminded of this post [EA · GW], particularly Will's comment [EA(p) · GW(p)] about funding/supporting a red team to criticize EA/longtermism. Is Open Phil open to doing  something like this? 

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-12T01:41:57.788Z · EA(p) · GW(p)

Thoughtful and well-informed criticism is really useful, and I'd be delighted for us to support it;  criticism that successfully changes minds and points to important errors is IMO among the most impactful kinds of writing. 

In general, I think we'd evaluate it similarly to other kinds of grant proposals, trying to gauge how relevant the proposal is to the cause area and how good a fit the team is to doing useful work. In this case, I think part of being a good fit for the work is having a deep understanding of EA/longtermism, having really strong epistemics, and buying into the high-level goal of doing as much good as possible.

comment by MaxRa · 2022-03-11T17:30:35.592Z · EA(p) · GW(p)

Thanks for sharing your thoughts so transparently! :) 

I'm particularly interested in this point:

Sometimes, our team supports projects that are directly aimed at [making object level progress], often because we think their value from a movement-building perspective is sufficiently high that it justifies supporting them (i.e. in those cases we might have different motives for supporting a project than the people who work on it have for working on it.)

a) I have the impression that we urgently need more smart people working on longtermist issues, particularly AI safety, governance and strategy

b) What do you think about the idea of encouraging longtermist researchers in general, and AI researchers in particular, to see their impact more than currently in terms of growing a field vs. making direct object-level progress?

  • as you say, both direct progress and getting more people on board are far from mutually exclusive, but I'd be surprised if it wouldn't change what people are working on if we'd deliberatively prioritize the latter more
    • concrete examples: we might encourage them to do more things like contributing to course curricula, networking with and outreach to top CS departments, organize workshops, develop prizes and benchmarks
comment by Ben Pace · 2022-03-10T20:56:02.523Z · EA(p) · GW(p)

Great post.

I didn't quite parse this paragraph:

For example, when we fund e.g. 80,000 Hours, we (amongst other activities) support their full-time advisors to advise interested people about how to have more impactful careers. With our scholarship programs, we’re also trying to cause people to spend more time on more impactful activities. But rather than do this via the 80k advisors, our scholarship programs use money “directly” (without much intermediating EA labor) to try to make impactful careers more accessible and attractive. In general, we think we get less impact per dollar from interventions that consume money “directly” like this. Since EA labor is the scarcer resource in many contexts, these types of interventions can make sense for grantmakers to prioritize.  

I think you're saying that your scholarships seem good to you, and that this has something to do with the value of your time versus the value of 80k staff time, but I'm not quite sure how you're connecting these variables, and exactly whose time you're saving with the scholarships (I imagine it takes you a lot of time to make the scholarship decisions, but maybe not).

Replies from: ClaireZabel, Linch
comment by ClaireZabel · 2022-03-10T23:30:43.099Z · EA(p) · GW(p)

Hm yeah, I can see how this was confusing, sorry!

I actually wasn't trying to stake out a position about the relative value of 80k vs. our time. I was saying that with 80k advising, the basic inputs per career shift are a moderate amount of funding from us and a little bit of our time and a lot of 80k advisor time, while with scholarships, the inputs per career shift are a lot of funding and a moderate amount of our time, and no 80k time. So the scholarship model is, according to me, more expensive in dollars per career shift, but less time-consuming of dedicated longtermist time per career shift. 

I think the scholarships are more time-consuming for us per dollar disbursed than giving grants to 80k, but less time-consuming in aggregate because there's effectively no grantee "middle man" also spending time. 

Of course, some of the scholarships directly fund people to do object-level valuable things, this argument just concerns their role in making certain career paths more attractive and accessible. 

Does that make more sense? 

Replies from: Ben Pace
comment by Ben Pace · 2022-03-11T22:14:57.669Z · EA(p) · GW(p)

Thanks! The core thing I'm hearing you say is that the scholarships are the sort of thing you wouldn't fund on a cost-effectiveness metric and 80k is, but that on a time-effectiveness metric that changes it so that the scholarships are now competitive.

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-11T22:45:39.481Z · EA(p) · GW(p)

No, that's not what I'd say (and again, sorry that I'm finding it hard to communicate about this clearly). This isn't necessarily making a clear material difference in what we're willing to fund in many cases (though it could in some), it's more about what metrics we hold ourselves to and how that leads us to prioritize.  

I think we'd fund at least many of the scholarships from a pure cost-effectiveness perspective. We think they meet the bar of beating the last dollar, despite being on average less cost-effective than 80k advising, because 80k advising doesn't have enough room for funding. If 80k advising could absorb a bunch more orders of magnitude of funding with no diminishing returns, then I could imagine us not wanting to fund these scholarships from a cost-effectiveness perspective but wanting to fund them from a time-effectiveness perspective.

A place where it could make a material difference is if I imagine a hypothetical generalist EA asking what they should work on. I can imagine them noting that a given intervention (e.g. mentoring a few promising people while taking a low salary) is more cost-effective (and I think cost-effectiveness is often the default frame EAs think in), and me encouraging them to investigation whether a different intervention allows them to accomplish more with their time while being less cost-effective (e.g. setting up a ton of digital advertising of a given piece of written work), and saying that right now, the second intervention might be better. 

comment by Linch · 2022-03-10T23:15:46.638Z · EA(p) · GW(p)

My personal reading of the post is that they think the scholarship decisions don't take up a lot of time, relative to 80k advisory stuff.

comment by Luise · 2022-03-24T12:13:40.424Z · EA(p) · GW(p)

Hi Claire,

what are your thoughts on "going one meta-level up" and trying to build the meta space? Specifically creating opportunities like UGAP, the GCP internships, or running organisers' summits to get more and better community builders? I'm unsure but I thought this might be at odds with some of the points you raised, e.g., that we might neglect object-level work and its community-building effect. I'd love to hear your thoughts!

Replies from: ClaireZabel
comment by ClaireZabel · 2022-03-25T07:07:05.185Z · EA(p) · GW(p)

I'm interested in and supportive of people running different experiments with meta-meta efforts, and I think they can be powerful levers for doing good. I'm pretty unsure right now if we're erring too far in the meta and meta-meta direction (potentially because people neglect the meta effects of object-level work) or should go farther, but hope to get more clarity on that down the road.