Quick takes

Set topic
Frontpage
Global health
Animal welfare
Existential risk
Biosecurity & pandemics
10 more

GiveWell did their first "lookbacks" (reviews of past grants) to see if they've met initial expectations and what they could learn from them:

Lookbacks compare what we thought would happen before making a grant to what we think happened after at least some of the grant’s activities have been completed and we’ve conducted follow-up research. While we can’t know everything about a grant’s true impact, we can learn a lot by talking to grantees and external stakeholders, reviewing program data, and updating our research. We then create a new cost-effectiveness

... (read more)

I suspect there is a confusion of terminology here, and also perhaps some loss of institutional knowledge. Givewell did post-hoc analyses starting in 2011 of their 2009 and 2010 recommendations to donate to VillageReach, but these were not technically "grants", but rather "charity recommendations", so I guess wouldn't be considered a "grant lookback".

In recent years GiveWell shifted from a charity recommendation model to a more direct grantmaking model, so this could be the first reviews of grants under that new model.

4
Karthik Tadepalli
This should be very easy for you to buy! The opportunity cost of lookbacks is investigating new grants. It's not obvious that lookbacks are the right way to spend limited research capacity. Worth remembering that GW only has around 30 researchers and makes grants in a lot of areas. And while they are a leading EA grantmaker, it's only recently that their giving has scaled up to being a notable player in the total development ecosystem.
2
NickLaing
Yep I agree! I've done a quicky sanity check on the New Incentives numbers and it doesn't seem quite plausible, but my it was fast and I could be plain wrong. https://forum.effectivealtruism.org/posts/FxAtFMRnJZ2dbLBhA/sanity-check-givewell-s-new-incentives-estimate-seems I would also like to see OpenPhil look back at a bunch of their "hits based" grants. They've done a decent amount of them and I think we should be able to get some idea about whether the approach is working as planned. It wouldn't have to be too detailed. They could even do something a bit loose, like categorising them into maybe 4 buckets like ..... 1. Miss          2. Probable miss         3. Some benefit       4. Home Run hit successful! Or similar

I notice a pattern in my conversations where someone is making a career decision: the most helpful parts are often prompted by "what are your strengths and weaknesses?" and "what kinds of work have you historically enjoyed or not enjoyed?"

I can think of a couple cases (one where I was the recipient of career decision advice, another where I was the advice-giver) where we were kinda spinning our wheels, going over the same considerations, and then we brought up those topics >20 minutes into the conversation and immediately made more progress than the res... (read more)

An excerpt about the creation of PEPFAR, from "Days of Fire" by Peter Baker. I found this moving.

Another major initiative was shaping up around the same time. Since taking office, Bush had developed an interest in fighting AIDS in Africa. He had agreed to contribute to an international fund battling the disease and later started a program aimed at providing drugs to HIV-infected pregnant women to reduce the chances of transmitting the virus to their babies. But it had only whetted his appetite to do more. “When we did it, it revealed how unbelievably pathe

... (read more)

The book "Careless People" starts as a critique of Facebook — a key EA funding source — and unexpectedly lands on AI safety, x-risk, and global institutional failure.

I just finished Sarah Wynn-Williams' recently published book. I had planned to post earlier — mainly about EA’s funding sources — but after reading the surprising epilogue, I now think both the book and the author might deserve even broader attention within EA and longtermist circles.

1. The harms associated with the origins of our funding

The early chapters examine the psychology and incentives... (read more)

Showing 3 of 4 replies (Click to show all)

I read the book a while back and I enjoyed it. It was kind of fun to get some juicy details about bad things inside Facebook. My main takeaway was something along the lines of "a fish rots from the head." Leaders of an organization set priorities, direction, culture (to a great extent), and this books served as sort of a case study of leadership that has a fairly narrow focus. Poor social skills and poor common sense, entitlement, and the general idea that you get everything you want all stood out to me. The levels of sycophancy and self-interest were a bi... (read more)

8
Lorenzo Buonanno🔸
Bloomberg's valuation of Moskovitz's fortune recently dropped by ~60% (from $30B to $11B) as his level of ownership of Meta was not significant enough to show up in their filings. (But Forbes' estimate didn't change much at $19B)
2
Benevolent_Rain
Fixed! Thanks for pointing that out.

When is it acceptable to cite an LLM?

I couldn't find any discussion or consensus about this, so I'll ask. 

I am surprised to find seemingly well-thought-out articles openly citing LLMs as research. For contrast, no one (who knows how to use Wikipedia) would cite Wikipedia: you are supposed to go to its sources. That's why [citation needed] is a thing: you don't want to build on unsubstantiated BS. So I find it hard to believe that people would cite infamously hallucination/BS-prone LLMs instead, expecting some reasoning built on that to be accepted.

If ... (read more)

I think you should get the LLM to give you the citation and then cite that (ideally after checking it yourself).

2
MichaelDickens
I don't cite LLMs for objective facts. In casual situations I think it's basically okay to cite an LLM if you have a good sense of what sorts of facts LLMs are unlikely to hallucinate, namely, well-established facts that are easy to find online (because they appear a lot in the training data). But for those sorts of facts, you can turn on LLM web search and it will find a reliable source for you and then you can cite that source instead. I think it's okay to cite LLMs for things along the lines of "I asked Claude for a list of fun things to do in Toronto and here's what it came up with".
1
hmijail
If an Anthropic data scientist in a high-profile legal case can be hoodwinked by bad citations, I don't think that it is realistic at all to think that anyone can have a "good sense of what sorts of facts LLMs are unlikely to hallucinate". And I thought we all have heard about lists of fun things to do full of non-existent restaurants in the way to non-existent towns?

GiveWell's cost to save a life has gone from $4,500 to a range between $3,000 and $5,500:

https://www.givewell.org/how-much-does-it-cost-to-save-a-life

From at least as early as December 2023 (possibly as early as December 2021 when the page says it was first published) until February 2024, that page highlighted a $7.2 million 2020 grant to the Against Malaria Foundation at an estimated cost per life saved of $4,500.

The page now highlights a $6.4 million 2023 grant to the Malaria Consortium at an estimated cost per life saved of $3,000.

You can see all the es... (read more)

9
Mo Putera
I wonder how they select grants to showcase on that page. They've made grants that are both much larger and more cost-effective than that, e.g. this $71.5M grant in Jan '23 to HKI's vitamin A supplementation program that they estimate would save roughly 49,000 lives at ~$1,450 per life saved after all adjustments (or ~93,000 lives at $770 per life if only adjusting for internal and external validity, or nearly 280k lives at at $260 per life saved before any adjustments, i.e. the standard I usually see in most BOTECs claiming to "beat GW top charities"...). Only thing is, this wouldn't be obvious from their original CEA because they tend to input "donation (arbitrary size)" = $100k instead of the actual grant amounts; I had to manually input their grant budget breakdown into a copy of their CEA to get the numbers above (which also means I may have done it wrong, so caveat utilitor...)

I would guess that it's based on the marginal grant, but of course someone at GiveWell should be able to confirm.

Am i the only one who finds the X% disagree UX confusing? It's hard not to read it and intuitively think to myself that it's an alternative weighting/aggregation/expression of the Agree/Disagree votes. 

Note sure how to change the UX to be clearer, perhaps "X% disagree with question" would make it clearer to me.

2
Toby Tremlett🔹
I think it is still somewhat confusing (for example you might expect a neutral vote is 50%, but it's actually 0%) - when we implemented it, it was the least worst idea. Keen to hear alternatives!  I think the "X% disagree with question" would be a little unwieldy. A couple ideas: - just show a mini debate slider, with a line indicating where the user voted.  - a "see vote" link (this means one extra click, so that could be annoying). 

mini-slider seems reasonable, "X% disagree with question" also seems fine to me, at least on desktop. "See vote" link seems bad for reasons you mention

This is one of the first times I have seen lead poisoning make front page news! 

https://www.bbc.com/news/articles/cy4n7wn8l58o

Could this be an opportunity for the new lead alliance, or even just one of the lead orgs to help China do better on lead? Maybe there will be a window where they are more open to external help? It seems China has new "maximum" lead levels regulations in paint from 2020 but there are obviously still issues...

Could well still not be worth it though if there are other "lower hanging fruit" on the lead front.

I think suicide prevention might be an underrated cause (need to firmly fact check before my confidence in this is high)
(1) if you delay someone from commiting suicide for just 30 minutes they will almost always change their mind
(2) suicidal people usually spend years inbetween attempts
(3) after someone "fails" a suicide attempt via changing their mind they usually feel a lot better emotionally (excluding failed attempts, only failure via changing your mind)
 

a charity in the UK places 1 hour of phone time is £44, if we assume 10% of people who call th... (read more)

Hey! You might be interested in this forum post on suicide hotlines from a while back. https://forum.effectivealtruism.org/posts/Xh5oyGkngGjCmNhcP/how-impactful-is-being-a-suicide-hotline-volunteer

Unfortunately, I think the comments in the post (including one from a former volunteer) make a really compelling case that it takes a lot more than 5 hours to prevent a suicide. I think it takes at least hundreds and probably thousands of hours of calls.

3
geoffrey
Not sure what the right numbers are but I really like the back-of-the-envelope approach you're taking here. It's simple and concrete enough that it's probably going to bounce around in my head for a while
5
huw
You might be surprised to learn that CEAs of mental health interventions in the EA space (example) don’t count the value of preventing suicide and self-harm. But on a DALY basis, self-harm and suicide have roughly the same burden in total as depression as a whole, precisely because they’re so much worse (this is to say nothing of effects on income). I think that mental health interventions, and especially direct suicide counselling, might be really underrated, simply because the research hasn’t been done in a lot of depth (that I’m aware of). Part of this is because it’s quite hard to accurately quantify the effects of interventions on long-term suicide rates (you need many years, and access to death records). However, there are a few good studies that I think might point to ways this could at least be estimated, and then we can agree on an appropriate evidence discount. The theory of change gets quite convoluted. For example, people may report feeling suicidal or having suicidal thoughts, but this may have no real correlation with their attempts. Or we may find that people who report feeling less suicidal commit suicide less often, but it may not be true that reducing their feelings of suicidality actually has any effect on their long-term attempt rate. Here’s a good paper looking at 84,000 people’s PHQ-9 scores and follow-ups with suicidality. Unfortunately, the evidence isn’t causal, but this study looks at suicide prevention RCTs and finds that, at least, depression interventions reduce ideation, but not suicidality. I’ll be doing a bit more research in this area because I think it’s quite promising!

POLL: Is it OK to eat honey[1]?

I've appreciated the Honey wars. We've seen the kind of earnest inquiry that makes EA pretty great. 

I'm interested to see where the community stands here. I have so much uncertainty that I'm close to the neutral point, but I've been updated towards it maybe not being OK - I previously slurped the honey without a thought. What do you think[2]?
 

It's OK to eat honey
BS
A
GR
AD
AS
JXL
BC
F
MK
S
NM
N
O
GP
K
L
M
BM
DB
I
RG
ZJ
CD
C
C
AK
NN
A
IT
L
NK
Z
Z
disagree
agree
  1. ^

    This is a non-specific question. "OK" could mean a number of things (you choose). It could mean you think eating net honey is "net positive" (My pleasure/health > sma

... (read more)
Showing 3 of 9 replies (Click to show all)
akash 🔸
3
0
0
10% disagree

It's OK to eat honey

I am quite uncertain because I am unsure to what extend a consumption boycott affects production; however, I lean slightly on the disagree side because boycotting animal-based foods is important for:

  1. Establishing pro-animal cultural norms
  2. Incentivizing plant-based products  (like Honee) that already face an uphill climb towards mass adoption
3
Jeroen Willems🔸
I try to avoid it, but it's hard for me to believe it's as bad or worse than most animal products. Especially in the quantities it's usually consumed. Who eats a kg of honey per year? I do think the treatment of bees is very unclear. But I've also heard that some non-animal products involve a lot of insects, like avocados, so I'm curious how it compares.
9
NickLaing
Ha I love this! I'm interested that your cluelessness places you heavily on the "It's OK side". I think if I felt "clueless" here my vote might be slightly on the "Not OK" side, because I would would worry that eating honey could still be really bad. 

There seems to be a pattern where I get excited about some potential projects and ideas during an EA Global, fill EA Global survey suggesting that the conference was extremely useful for me, but then those projects never materialise for various reasons. If others relate, I worry that EA conferences are not as useful as feedback surveys suggest.

If you haven't already, you should consider posting your project ideas!

  1. You can get feedback on which ideas seem most promising, so maybe you end up getting those ones done
  2. Other people might pick up the ideas, or it could inspire related ideas
21
OllieBase
Yep, I think this is right, but we don't totally rely on these kinds of surveys! We also conduct follow-up surveys to check what actually happens a few months after each event and unsurprisingly, you do see intentions and projects dissipate (as well as many materialising). A problem we face is that these surveys have much lower response rates. Other more reliable evidence about the impact of EAG comes from surveys which ask people how they found impactful work (e.g., the EA Survey, Open Phil's surveys), and EAG is cited a lot. We'll usually turn to this kind of evidence to think about our impact, though end-of-event feedback surveys are useful for feedback about content, venue, catering, attendee interactions etc. and you can also do things like discounting reported impact in end-of-event surveys using follow-up survey data.

Marginal returns to work (probably) go up with funding cuts, not down.

It can be demoralizing when a field you’re working in gets funding cuts. Job security goes down, less stuff is happening in your area, and people may pay you less attention since they believe others are doing more important work. But assuming you have job security and mostly make career decisions on inside views (meaning you’re not updating too heavily on funders de-prioritizing your cause area), then your skills are more valuable than they were previously.

Lots of caveats apply of course... (read more)

7
Joel Tan🔸
It's true that there are diminishing marginal returns, and with less funding and fewer projects/people around, there is now a bunch of opportunities for impact which you can exploit (where previously someone else would have done it anyway). However, there's also a countervailing reduction in marginal value of labour due to reduced availability of capital and non-labour input, especially since cuts aren't necessarily well targeted (e.g. keeping staff while capital investment is slashed). Loss of infrastructure field-wide is also a critical problem (e.g. all those interventions and programmes piggybacking on AIDS clinics)

Good point. In a toy model, it'd depend on relative cuts to labor versus non-labor inputs. Now that I think about it, it probably points towards exiting being better in mission-driven fields. People are more attached to their careers so the non-labor resources get cut deeply while all the staff try to hold onto their jobs.

Maybe I'd amend it to... if you're willing to switch jobs, then you can benefit from increasing marginal returns in some sub-cause areas. Because maybe there's a sub-cause area where lots of staff are quitting (out of fear the cause area ... (read more)

I’m in a WeChat group initiated by Plant Future and Good Food Fund. It is meant to connect young Chinese students who are interested in promoting vegetarianism. We have a weekly discussion (in English) every Wednesday morning 7:30AM. If you’re interested in joining, please send me a message.

Note that you do need a WeChat account.

Recently I got curious about the situation of animal farming in China. So I asked the popular AI tools (ChatGPT, Gemini, Perplexity) to do some research on this topic. I have put the result into a NotebookLM note here: https://notebooklm.google.com/notebook/071bb8ac-1745-4965-904a-d0afb9437682

If you have resources that you think I should include, please let me know.

The original reports can be found here: https://u.pcloud.link/publink/show?code=kZiW1f5Zf9YhpJUeHqfPVbuv9Afhozu1XSgy

I have also written a short summary.

Load more