Hi!
Iâm currently (Aug 2023) a Software Developer at Giving What We Can, helping make giving significantly and effectively a social norm.
Iâm also a forum mod, which, shamelessly stealing from Edo, âmostly means that I care about this forum and about you! So let me know if thereâs anything I can do to help.â
Please have a very low bar for reaching out!
I won the 2022 donor lottery, happy to chat about that as well
Not sure who to alert to this
I think itâs probably best to alert whoever sent you the survey, I wouldnât rely on them noticing quick takes on the EA Forum
Did you mean to link to âImpact Maximisation through Supported Regranting: a Funding Strategy Hackâ?
Expanding acronyms for readers who might not know them:
EAA: Effective Animal Advocacy
ICAPs: Importance and Counterfactuals-Adjusted Placements, an Animal Advocacy Careers specific metric
Thanks for writing this! I would be curious to know what you think about this 4 years later, and now that interest rates are much higher.
I would find it valuable if you could share some public version of the spreadsheet, or if you quickly remember some specific examples. Hiring/âcontracting is very hard but almost always necessary.
Iâm probably less informed than you are, but depending on what you mean by âsources of fundersâ I disagree.
I think if you can demonstrate getting valuable results and want funding to scale, people will be happy to fund you. My impression is that several people influencing >=6 digit allocations are genuinely looking for projects to fund that can be even more effective than what theyâre currently funding.
Iâm fairly confident that if anyone hosted a conference or online program, got good results, had a clear theory of change with measurable metrics, and gradually asked for funding to scale, people will be happy to fund that.
I get exactly the same: itâs the upvote/âreactions notifications, not the replies
I agree with 3 of your points but I disagree with the first one:
EA atm doesnât seem very practical to take action on except for donating and/âor applying to a small set of jobs funded mostly by one source.
On jobs: 80k, Probably Good and Animal Advocacy Careers have job boards with lots of jobs (including all AIM jobs) and get regularly recommended to people seeking jobs. I met someone new to EA at EAGx Berlin a month ago, and 3 days ago they posted on LinkedIn that they started working at The Life You Can Save.
On donations: Iâm biased but I think donations can be a really valuable action, and EA promotes donations to a large number of causes (including AIM).
My guess is this is reducing the number of âoperatorâ types that get involved and selects for cerebral philosophising types.
Itâs really hard for me to tell if this is a good or bad thing, especially because I think itâs possible that things like animal welfare or GCR reduction can plausibly be significantly more effective than more obviously good âpractical testableâ work (and the reason to favour âR&Dâ mentioned previously)
I heard about 80k and CEA first but it was the practical testable AIM charities that sparked my interest THEN Iâve developed more of an interest in implications from AI and GCRs.
Not really a disagreement, but I think itâs great that thereâs cross-pollination, with people getting into AIM from 80k and CEA, and into 80k and CEA from AIM
I do think the comment would have been much better received if it was more concise and simple to read (regardless of how it was written), see The value of content density
I feel Iâm not informed enough to reply to this, and it feels weird to speculate about orgs I know very little about, but I worry that the people most informed wonât reply here for various reasons, so Iâm sharing some thoughts based on what very little information I have (almost entirely from reading posts on this forum, and of course speaking only for myself). This is all very low confidence.
âEffectiveâ Altruism implies a value judgement that requires strong evidence to back up
I think if you frame it as a question, something like âWe are trying to do altruism effectively, this is the best that weâre able to do so farâ, it doesnât require that much evidence (for better or worse)
âAmbitious Impactâ implies more speculative, less easy to measure activities in pursuit of even higher impact returns.
That is not clear to me, one can be very ambitious by working on things that are very easy to measure. For example, people going through AIMâs âfounding to giveâ program seem to have a goal thatâs easy to measure: money donated in 10 years[1], but I still think of them as clearly âambitiousâ if they try to donate millions. Google defines ambitious as âhaving or showing a strong desire and determination to succeedâ
My understanding is that Open Philanthropy split from GiveWell because of the realisation that there was more marginal funding required for âDo-gooding R&Dâ with a lower existing evidence base.
That is not my understanding, reading their public comms I thought OP split from GiveWell to better serve 7-figure donors âOur current product is a poor fit with the people who may represent our most potentially impactful audience.â (which I assumed implicitly meant that Moskovitz and Tuna could use more bespoke recommendations)
Why do we need âDo-Gooding R&Dâ?
I agree with this! I liked Finding before funding: Why EA should probably invest more in research, but I expect that the âR&Dâ work itself might be tricky to do in practice. Still, Iâm very excited about GiveWellâs RCT grants
IMO AIM has outcompeted CEA on a number of fronts (their training is better, their content (if not their marketing) is better, they are agile and improve over time). Probably 80% of the useful and practical things Iâve learned about how to do effective altruism, Iâve learned from them.
I agree that AIM is more impressive than CEA on many fronts, but I think they mostly have different scopes.[2] My impression is that CEA doesnât focus much on specific ways to implement specific approaches to âhow to do effective altruismâ, but on things like âwhy to do effective altruismâ and âhere are some things people are doing/âwriting about how to do good, go read/âtalk to them for detailsâ.
If not for CEA, I think I probably wouldnât have heard of AIM (or GWWC, or 80k, or effective animal advocacy as a whole field). And if I had only interacted with AIM, Iâm not sure if I would have been exposed to as many different perspectives on things like animal welfare, longtermism, and spreading positive values[3]
The AIM folks Iâve spoken to are frustrated that their resultsâbased on exploiting cost-effective high-evidence base interventionsâare used to launder the reputation of OP funded low evidence base âDo-gooding R&D.â
I understand the frustration, especially given the brand concerns below and because I think many AIM folks think that a lot of the assumptions behind longtermism donât hold.[4] But I donât know if this âreputation launderingâ is actually happening that much:
My sense is that the (vague) relation to the Shrimp Welfare Project is not helping the reputation of some other EA-Adjacent projects
I think AIM is just really small compared to e.g. GiveWell, which I think is more often used to claim that EA is doing some good
When e.g. 80000hours interviews a LEEP cofounder, I think itâs because they believe that LEEP is really amazing (as everyone does), they want to promote it, and they want more people to do similar amazing things. I think the reason people talk about the best AIM projects is usually not to look better by association but to promote them as examples of things that are clearly great.
If we think about EA brand as a product, Iâd guess weâre in âThe Chasmâ below as the EA brand is too associated with the âweirdâ stuff that innovators are doing to be effectively sold to lower risk tolerance markets.
I personally believe that the EA brand is in a pretty bad place, and at the moment often associated with things like FTX, TESCREAL and OpenAI, and that is a bigger issue. I think EA is seen as a group non-altruistic people, not as a group of altruistic people who are too âweirdâ. (But I have even lower confidence on this than on the rest of this comment)
AIM should be the face of EA and should be feeding in A LOT more to general outreach efforts.
Related to the point above, itâs not clear to me why AIM should be the face of âEAâ instead of any other âdoing the most goodâ movement (e.g. Roots of Progress, School for Moral Ambition, Center for High Impact Philanthropy, âŚ). I think none of these would make a lot of sense, and donât see why âAIM being the face of AIMâ would be worse than AIM being the face of something else.
You can see in their 2023 annual review that they did deeply consider building a new community âbut ultimately feel that a more targeted approach focusing on certain careers with the most impact would be better for usâ.[5]
In general, I agree with your conclusions on wishing for more professionalization, and increasing the size of the pie (but it might be harder than one would think, and it might make sense to instead increase the number of separate pies)
I imagine positive externalities from the new organizations will also be a big part of their impact, but I expect the main measure will be amount donated.
And that this does not say that much about CEA as imho AIM is more impressive than the vast majority of other projects.
AIM obviously does a lot for animal welfare, but I donât think they focus on helping people reson about how to prioritize human vs non-human welfare/ârights/âpreferences.
I canât link to the quote, so Iâll copy-paste it here.
JOEY: Yeah, I basically think I donât find a really highly uncertain, but high-value expected value calculation as compelling. And they tend to be a lot more concretely focused on whatâs the specific outcome of this? Like, okay, how much are we banking on a very narrow sort of set of outcomes and how confident are we that weâre going to affect that, and whatâs the historical track record of people whoâve tried to affect the future and this sort of thing. Thereâs a million and a half weeds and assumptions that go in. And I think, most people on both sides of this issue in terms of near-term causes versus long-term causes just have not actually engaged that deeply with all the different arguments. Thereâs like a lot of assumptions made on either side of the spectrum. But I actually have gotten fairly deeply into this. I had this conversation a lot of times and thought about it quite thoroughly. And yeah, just a lot of the assumptions donât hold.
Linked from this post.
Full quote:
Building a new community (e.g. Impact Now):
An option many people have been asking us about in the wake of the struggles of the EA movement is if CE would consider building out a movement that brings back some of the strengths of EA 1.0. We considered this idea pretty deeply but ultimately feel that a more targeted approach focusing on certain careers with the most impact would be better for us. The logistical and time costs of running a movement are quite large and it seems as though often a huge % of the movementâs impact comes from a small number of actors and orgs. Although we like some things the EA movement has brought to the table when comparing it to more focused uses (e.g. GiveWell focuses more on GiveWellâs growth), we have ended up more pessimistic about the impact of new movements.
I donât think they linked to their 2024 annual report on the forum, so this might be different now.
Yes I think we agree, but I also think that itâs not a crux of the argument.
As Neel Nanda noted, whatever vaguely reasonable method you use to calculate impact will result in attributing a lot of impact to life-saving interventions.
I think there is a valuable concern about Triple counting impact in EA and I agree that there is a case for Shapley values being better than counterfactuals[1].
What I really donât agree with is that we should let someone choke and die, just because otherwise Henry Heimlich would get the credit anyway. The goal is not to get the most credit or Shapley values, but to help others, I donât see what prof. Wenar proposes as a better alternative to GiveWell.
I disagree that Shapley values are better than counterfactual in most cases, but I think itâs a reasonable stance to have.
I donât see how asking for higher standards for criticism makes EA defenseless against âbullshit.â
I actually would argue the opposite: if we keep encouraging and incentivizing any kind of criticism, and tolerate needlessly acrimonious personal attacks, we end up in an environment where nobody proposes anything besides the status quo, and the status quo becomes increasingly less transparent.
Three recent examples that come to mind:
User @Omega stopped writing their criticism of AI Safety organizations partly because of how the criticism of their criticism assumed bad faith
@Happier Lives Institute stopped writing their criticism of/âalternatives to GiveWell in part because some criticism they got was largely personal attacks
@Dustin Moskovitz claims that they[1] write less on this forum because of attacks from EAs. He also wrote the attitude in EA communities is âgive an inch, fight a mileâ. So Iâll choose to be less legible instead.
I think Holly_Elmore herself is another example: she used to write posts like âWe are in triage every second of every dayâ, which I think are very useful to make EA less âbullshitâ, but now mostly doesnât post on this forum, partly because of the bad quality costly criticism she receives.
I largely agree with the last section of this comment from Aaron Gertler written one year ago:
The Forum has a hard balance to strike:
I think the average comment is just a bit less argumentative /â critical than would be ideal.
I think the average critical comment is less kind than would be ideal.
I want criticism to be kind, but I also want it to exist, and pushing people to be kinder might also reduce the overall quantity of criticism. Iâm not sure what the best realistic outcome is.
I personally fear that the current discussion environment on this forum errs too much in the âunkind criticismâ direction, and I see at least two large downsides:
It encourages some form of âenlightened immobilismâ, where anyone proposing doing anything differently from the status quo gets instantly shut down.
It strongly discourages transparency from most projects, especially (but not exclusively) less established or more speculative ones.
I used to think that accepting callousness was required to have technical excellence, e.g. reading how people like famous software engineer Linus Torvalds used to communicate. After seeing many extremely competent people communicate criticism in a professional and constructive manner, I have completely changed my mind. Torvalds also apologised and changed communication style years ago.
I believe that a culture of more constructive and higher-quality criticism would encourage more discussion overall, not less, especially from experienced professionals who have different perspectives from mainline EA thinking.
See also this paragraph from the Charity Entrepreneurship handbook:
Writing as myself, not as a moderator
Iâm not sure if he meant Good Ventures, Open Philanthropy, or some other group
I would also be curious to hear more about why/âif you are >~95% confident that pigs are not entities that experience suffering, while most humans are.[1]
Is it about the ability to have second-order beliefs, the ability to have complex language and certain kinds of social structures, or something else entirely?
I think pigs are much more similar to humans than broiler chickens, so are a better species to examine the difference
Thanks! Hereâs what the full distribution looks like right now:
Edit: here is what it looks like at the end of the week
Reminds me of xkcd.com/â1162
Iâm really surprised by the distribution. I think many voters might be too optimistic about the actual impact of marginal funding to animal welfare projects.
If I remember correctly, we decided not to list them in both groups because people already need to scroll a lot (especially on mobile) to see all the 15 programs, if we added the 6 recommended ones it would become 21
I agree that not seeing the top programs in the various categories is also confusing though, especially if you want to link to them directly
What kinds of open questions do you have in mind (perhaps some examples would help)?
Random example: I just wanted to ask today if anyone knew of a good review of âThe Good It Promises, the Harm It Doesâ written by a non-male, given that I think one of the key criticisms of EA in the feminist-vegetarian community is that its leaders are mostly white males, but I didnât know where to ask.
I havenât looked at this model, but in GWWCâs 2020â2022 Impact evaluation, you can change the annual discount rate here or here (and other key parameters in cells nearby)
If I understand correctly, a ~10% yearly discount rate ~halves the expected value of a pledge and changes the best guess non-marginal multiplier from 30x to 23x
I think the equivalent in this model is here and a ~10% discount rate changes the marginal multiplier from 14x to 8x
Thank you for raising this!
After your email last week, we agreed to edit that section and copy EVâs terms on Feedback. Iâve just changed the text on the website.
We only removed the part about âall Feedback we request from you will be collected on an anonymous basisâ, as we might want to collect non-anonymous feedback in the future.
If anyone else has any feedback, make sure to also send us an email (like Eevee did) as we might miss things on the EA Forum.
Having read your reflections, Iâm still curious as to why you donât think non-OpenPhil donors should give to farmed animal welfare, if you feel comfortable sharing it publicly. I guessed four options, ordered from most to least likely, but I might have misunderstood the post
We should donate to wild animal welfare instead, as itâs more cost-effective
There are no donation opportunities that counterfactually help a significant amount of farmed animals
There is no strong moral obligation to improve future lives, and donations to farmed animal welfare necessarily improve future lives, as farmed animal lives are very short
Tomasik-style arguments on the impact of animal farming on the amount of wild animal suffering
Is it a combination of these? As a concrete example, Iâm curious if you believe that the Shrimp Welfare Project shouldnât be funded, should be funded by ânon-EAâ donors, or will be funded anyway and donors shouldnât worry about it.
By the way, thank you for nudging towards sharing evaluations with the evaluated organization before posting, I think itâs a really valuable norm.