Director of Operations at GovAI. I have a blog about nonprofit ops and strategy.
I previously co-founded and served as Executive Director at Wild Animal Initiative, was the COO of Rethink Priorities from 2020 to 2024, and ran an operations consultancy, Good Structures, from 2024-2025.
There is a (currently) non-public organization working on making this happen. If anyone is interested in learning more, feel free to reach out (we’re hiring)!
Yeah, I agree with all those being challenges here—I think I was mainly responding to what I perceive to be a push (maybe explicitly in this case) to reduce the options presented to new funders to a few funds with fairly similar views, which I think is possibly a strategic mistake, even if the alternative isn’t ideal either.
I think I feel less convinced than you that scaling most these are going to end up resulting in meaningful positive impact for more animals — the exception is welfare technology, which I’m quite excited about, but my impression is that the good opportunities here are pretty fundable right now.
To be clear, I also don’t think loads of money should go into neglected animals either (though on the margin I’m more excited about things here than FAW) — I think there is a lot more potential for helping animals in wild animal and invertebrate welfare, but there aren’t ways to absorb tons (e.g. tens of millions of dollars) of funding there either (at least not yet).
I generally in both cases am excited by a smaller, more highly coordinated and strategic movement (or set of movements) than a larger one, but I think more funding right now would be used primarily to try to make a bigger one. I’m guessing this is a lot of the crux between us. But, I also know that I’m a bit on my own island with these views at times, and am genuinely pro-pluralism in the space. So I appreciate you pushing on it so hard!
I think AI safety mostly can’t absorb new funding that effectively, except for political things (which maybe are complicated due to various backfire risks), but it also has a better track record so far than FAW which suggests it can use the money it has more effectively. But I’m not a partisan here really — at heart I’m an animal welfare person who mainly feels sad that it might be pretty hard to help more animals than we already are effectively.
For sure—I am quite negative on these marginal opportunities in FAW being good. To be clear, if there weren’t any funds primarily (or even partially) focused on farmed animal welfare, I’d make a bid for those to exist too rather than just funds for neglected animals / whatever it is I prefer (I’m not sure I even know). I think there shouldn’t just be one view / dominant position on how to help animals, multiple perspectives should get a seat at the table / we should make multiple types of bets, even those I disagree with because I’m sure I’m wrong about many things. I also agree that there would probably be a lot of wasted money under this model, but just think the higher upside offsets it.
I’d be happy to chat about views on FAW work specifically, but probably won’t here in detail (just because they are complicated and messy and hard to get into without talking about specific groups). At a high-level, I think that diet change work is basically intractable (or where it is tractable it just risks increasing chicken farming), and the vast majority of the value from welfare work is coming from the first $20M-$40M spent on it, in a pretty predictable way, such that marginal dollars are hard to spend effectively, and beyond those there isn’t anything super scalable it seems like. I think the climate/global heath analogies aren’t quite right, because the majority (maybe even large majority) of that money is spent in pretty ineffective ways — I probably wouldn’t be excited about marginal money to a random global development charity, vs a GiveWell top charity, which have much more limited room for more funding.
I disagree voted, because I don’t think it is a terrible policy / think it is a hard problem and they’ve solved it in probably the most reasonable way.
I think that it probably isn’t perfect and has a lot of issues, but pledged donations are counterfactual (no one would donate otherwise), while doing a direct work role is not as clearly counterfactual (the organization would usually probably hire someone else, but maybe they’d be less good than you, etc). I think that feels messy to litigate properly—in some cases doing direct work is way better than the counterfactual, but in others it might not be obvious, etc.
It just doesn’t seem clear cut how to resolve, and I think the explicit line of “you give away money that would otherwise be yours at this particular moment” seems like a fine way to slice it. If the pledge was “do the most good” pledge, I might agree, but then lots of other things besides taking a lower paying job or giving away money might count.
I’m pretty surprised by how high people are and that there are relatively few people in the middle (and I say this as a probably maximally invertebrate welfare friendly person who thinks we should spend lots of money on shrimp welfare but also only rated this question at 20%)! Interested in what’s leading to these views being so different and polarized. I guess my view for insects and other invertebrates is that the evidence we do have is very promising, but we really don’t have that much evidence.
Err towards applying for funding from an appropriate fund.
FWIW, I think this approach limits downside risk, as you outline, but really caps upside potential for EA. The existing funds are highly correlated with each other in values / perspectives on certain issues (e.g. I think it is quite bad that there isn’t an animal welfare-focused fund that is skeptical of the value of marginal dollars spent on farmed animal welfare, which I think is a reasonable position, and all of the AI funds seem to coordinate quite heavily). There just aren’t that many funds, they aren’t that diverse in viewpoint, and donations primarily going via funds concentrates power in EA in the hands of a few dozen people.
I think that I’d feel more excited about this approach if there were tons and tons of good funds with independent theses, but there aren’t. Maybe that means it is time for lots of people to start new funds. But by default, I expect everything going through funds to mean way less grift, yes, but also way less experimentation, less risk taking, and less divergence from consensus views.
FTX Future Fund, for all its issues / impacts on community dynamics, spurred a massive change in what people thought was on the table. The regrantor program in particular seemed like a genuinely massive improvement in democratizing EA, which, in my view, makes EA better and lets people do high risk/reward bets. There are downsides to that, including grift. But the push for funds, especially in animal welfare, seems like a fairly large mistake to me, and I think the end result will be a bunch of money wasted on marginal farmed animal interventions that Coefficient could have otherwise funded, or were obviously not worth funding.
I agree with the problems you outline in not going with the fund approach, but I don’t think the solutions being applied, especially on the fund side, are the right ones for doing the most good.
Very pro groups approaching fundraising with honesty and integrity though!
RE Organizations want this to exist:
- I think that something like 20ish organizations reported that they would use a common app system, at least for operations roles (I think they were much less likely to use it for other kinds of roles, but it was dependent on seniority, etc).
RE it not creating savings:
- I asked organizations about various ways that this would save them time. In total, my estimate was a common application + pre-vetting would save organizations 500-1350 hours per year (based on their reports on how they’d use it and how much time they spend on hiring).
- A common app alone might be half that? So 250-675 hours per year?
- My estimate is that it would have cost more hours than this to run well.
I think the primary reasons for this are:
- Organizations won’t only rely on the common app—they’d like easy ways to get candidates, but also want to recruit on their own platforms. For many non-ops roles, they didn’t really want to use it at all.
- The common app will get a lot more candidates than organizations get — it both makes it easier to apply to jobs, so will increase applications, and makes is more generic, so more people will feel qualified to apply.
Note that I looked at this from the perspective of “if we do this will we spend more time running it than the time savings for organizations” and I think the answer was yes.
RE credibility:
- A lot of organizations were worried about centralizing application processing / decision making because it creates a single point of failure.
- If you are also vetting applications, the above is worse + they have to trust you in the first place to do the vetting.
- The organizations who would have trusted us to do the vetting tended to be groups who had worked with us before on hiring and had a good experience.
Happy to have a call to talk about learnings from this, since as far as I know, my project was the closest the ecosystem has gotten to having a common app! Overall, I agree with the sense of there being lots of inefficiency in the hiring ecosystem — the complicated thing to me feels like candidates often want to solve for the problem of the candidate experience being bad, while the organizations want to solve for the problem of the organization experience being bad, and the causes of those problems are somewhat different.
Thanks so much for sharing!
Last year, I explored building a common application for EA/AI organizations, in collaboration with a funder in the space.
Specifically, we explored a version that might work like:
Applicants submit one application, and indicate organizations they’d be happy for application materials to be shared with.
Applicants, based on role type or skills, might get screening interviews / work tests from the common app system.
When an organization is ready to hire, they can quickly pull from pre-vetted candidates, skipping initial screening (since they have materials from our assessments).
Applicants save time by going through a process once, orgs save time by getting to skip advertising and initial screening windows.
I surveyed several dozen organizations about this idea, and talked to a few organizations directly about it. Here’s what I found:
Organizations wanted this to exist.
Organizations would be happy to recruit candidates out of a shared hiring pool.
Organizations wouldn’t rely heavily on this for applicants.
Organizations were generally unlikely to want this to be the only source of candidates. This means that they’d still open their own applications anyway.
I think this is primarily due to wanting diverse candidate pools / seeing value in doing their own advertising.
Organizations also generally wanted candidates to go through their own application process separate from the common app—basically, organizations perceive themselves as having heterogenous application processes.
On organizational self-reporting, no money would be saved.
While this process seems like it might produce savings, based on the time savings organizations reported this would generate for them, my estimate was that the cost-effectiveness of a funder paying for this service to exist was pretty low.
Basically, the issue is that without targeting a specific job, we end up vetting and screening a lot of people who might not be a good fit for roles in the ecosystem, and who might be quickly passed over by organizations.
My estimate of the time/cost it would take us to run the program, vs the self-reported time savings from organizations, was that it wasn’t cost-effective / wouldn’t really save the ecosystem money.
That being said, the program I explored was more comprehensive than just a common app. The issues I see with a pure common app are:
The organization running it would need to have sufficient credibility for the organizations using it to want to forego their own application processes. I think a random person starting it would have very low credibility. My company, which had run several dozen hiring rounds for many organizations had maybe 50% the credibility necessary. This seems like a hard bar.
Candidates have to trust the centralization — e.g. if the common app service also does vetting (which it doesn’t have to do, but which has the most value for the organizations using it), then they have to do a good job, as the stakes are high!
That being said, @Nina Friedrich🔸 and High Impact Professionals is doing tons of amazing work here, including some partial implementations of some of these ideas — their talent database, with candidate consent, lists organizations that candidates were finalists with, which is really useful for hiring.
RE sharing candidate information: this practice is really widespread in the ecosystem. I get probably 3-5 emails a month asking for referrals for candidates for roles, and typically share silver medalists from our similar hiring rounds who consented to sharing.
I think part of the disconnect is that organizations aren’t really optimizing on candidate time — they are optimizing on their own time and needs (whether or not this is a mistake).
Thanks again for writing this up! I think there are huge gains to be made here, and hope my notes on my exploration of it are useful for anyone thinking about it!
Oh yeah! Sorry, missed that. But to be clear, I definitely agree that this was an important point to put out there and am glad you did! :) Thanks for writing it.
Thanks for writing this up!
One note is that I haven’t really seen anyone doing welfare biology / wild animal welfare science themselves make these arguments—they’ve mostly been discussed in philosophy/econ papers on the topic. Interestingly, the scientists in the space seems to put much less weight on things like the Evening Out argument you describe, and generally seem to view welfare as just much more contingent on species/specific life-history factors that are hard to make abstracted claims about.
I feel less confident in this specific approach — I think downstream effects are just often very contingent. It’s easy to imagine scenarios like: eliminate very bad disease 1, population goes up, now very bad disease 2 emerges or is transmitted from some other population because contact occurs, etc.
I suppose this could be less of an issue for the very worst diseases/issues, etc., but I suspect those by their nature are less common (e.g. could be self-limiting, etc), and it seems like the method would only work insofar as changing population levels, etc. doesn’t increase the risk of novel worse things, which seems only possible for the very worst things.
That being said I am very sympathetic to just trying more things, at least for some animals, especially vaccinations.
Yep, I agree that the case is complicated by total welfare potentially being dominated by invertebrates. That being said, I think many people in the community who might not be motivated by helping insects or nematodes or mites might still care about shrimp, and humans still kill 25 trillion wild shrimp (!) annually.
Yep, I think farmed animal advocates sometimes miss that even if you only care about human-impacted animals (and not naturogenic suffering), the vast majority are wild animals, not farmed animals. The classic ACE graph could be replicated again with wild animals as the large boxes, and farmed animals as the small ones, and that’s putting aside climate change (which impacts way more wild animals).
FWIW:
Cooperation: We let Veganuary know about our intention to launch this campaign at the very start of our planning process and have kept them informed throughout. Our campaign provides them with another opportunity to put forward the benefits of diet change. We are all on good terms and there is absolutely no infighting.
I think it would be more useful to clarify if Veganuary supported you doing this campaign. If the answer is yes, that seems great! If the answer is no, this seems explicitly not cooperative, and in that case, it would be misleading to frame this as a cooperative effort (independent of if this was good or bad to do). I don’t think whether or not Veganuary was informed was what folks were looking for, but if they endorsed the idea or did not endorse it / anti-endorsed it.
I think this just seems like a clarification worth making here given how negative the reaction has been to the campaign (from within the movement—hopefully it had a positive reaction externally!)
I think that the good opportunities in the farmed (vertebrate) animal welfare space are:
Smaller in scale than Coefficient’s budget (I think that EAs have been overestimating the cost-effectiveness of corporate campaigns for at least a few years, and the good opportunities are actually pretty limited).
Pretty likely to be funded by non-EAs / people who will give to farmed animal welfare no matter what.
I think that there are likely a couple exceptions to this (shrimp welfare, insect farming, and some other things that Coefficient cannot fund currently), but they are fairly small in scale, and have decent routes to funding.
I think the opportunity for impact for wild animal welfare is way bigger, and it’s much more “normal” (e.g. it seems like there are more viable interventions that are acceptable in the mainstream, don’t require significant lifestyle changes of people, etc, WAI has gotten some traction within conservation), and generally is more neglected.
I noticed the checkbox for confirming awareness of the CC-BY license disappeared on new posts. I’m curious why the decision was made to drop it? It was useful for me personally to remind me when publishing things.
For what it’s worth (as someone who helped found Arthropoda but is no longer involved), I’d very much like there to be more convincing arguments against taking insects and other arthropods seriously. I feel pretty heavily incentivized to believe arguments against it as doing the animal welfare work I care more about emotionally (wild animal welfare) would be far easier. Working on animal welfare (and any other issue, if you care about second order effects) is vastly harder if you care about effects on insects, and I’d prefer the simpler world of only caring about vertebrates.
I think it’s pretty typical for the people who work on a cause area to be convinced that cause area matters. This is of course a source of bias, but, for example, asking global health charities to hire at least some people skeptical that we should improve the lives of people in developing countries seems like…. a hard request to fulfill at a minimum?
And, I believe that I and probably other people who have worked in this space are skeptics—just not extreme ones. I personally would not bet on any insects having morally relevant experiences, and put the odds at probably <30%. Relative to many this is less skeptical, but in absolute terms it still is skepticism—it sounds like you’re just advocating for there to be extreme skeptics—e.g. people who put the odds at, say, <1%. To analogize to global health again, it already feels odd to say “global health organizations should have folks who think there is a >70% chance this isn’t good thing to do”, let alone asking them to have staff who think there is a >99% chance.
A non-exhaustive list of things that seem like plausible candidates from a scale perspective, but are at varying points in the quality of research (and many are probably not near the certainty level we would need on the overall sign, but could be fairly easily, at least for target effects), and a rough guess at the scale of the number of animals that could be impacted by target effects:
Adapting more humane insecticides (hundreds of trillions?)
Indoor cats outside the US (where it’s mostly successfully been done) (low billions)
Eradicating rabies (mostly done successfully in Europe, very much not done in the US and other parts of the world) (tens of millions)
Rodent fertility control (hundreds of millions)
Other fertility control treatments for “pest” animals (pigeons, etc) (tens of millions)
Bird safe glass (already required by law in many jurisdictions (e.g. New York City for new construction)) (hundreds of millions)
More effective and humane island predator removal (millions)
Screwworm eradication (tens of billions)
All of these seem feasible in the nearer future, but still are minor compared to the scale of the bigger problems in the space, which I think academic field building is fundamental to address. If I could choose only one, I’d choose doing further academic field building over implementing any of these (though luckily we don’t have to choose between them).
(also, to be clear, WAI’s views might be very different than my own—just trying to give a flavor of what kind of timelines I was thinking about when setting up WAI).
Math nitpicks are helpful, thanks! Both were right—just doing math too quickly :).
RE welfare comparisons: I could imagine a difference between us being relative confidence that empirical research will improve our understanding? I think I might be less bullish on this sort of work because I don’t feel confident we’ll meaningfully reduce our uncertainty about welfare ranges. But, I’m not confident in this. Would you expect the most useful work for reducing your own uncertainty to be philosophical or empirical?
RE nematodes: I agree that this isn’t clear cut in some sense, but I feel fairly confident that they should be bracketed out unless we significantly advance in our understanding of animal consciousness (and see above—maybe my own lack of confidence in our ability to make empirical progress on this is part of the reason I’m more confident in casting them aside).
RE cage-free: yes — I think the meaningful counterfactual is that money spent on cage free otherwise not being spent on animal welfare at all, or being spent in mostly useless ways, and I’d endorse cage-free campaigns over that most likely, despite agreeing with you on non-target uncertainty being high, but I haven’t thought about it much.