All of Lucas Lewit-Mendes's Comments + Replies

Thank you, that makes sense! Keen to hear what comes from it. 

This looks super cool Ren and KvPelt! Curious - do you plan to advocate for electrical stunning after the six months of investigation? I'm wondering if the theory of change would be similar to FWI and SWP, or if it's still to be determined based on this initial investigation work? 

1
KvPelt
5mo
Hi Lucas, thanks for the kind reply! Our current project is limited to researching and gathering information. If the findings and insights we gather during the initial six months of investigation show that there are cost-effective ways for us to help out, we will consider continuing our work. What this potential future work will look like also fully depends on our findings.

Thanks Martin, looking forward to it! Just checking - should I eat dinner beforehand? (Is it mostly snacks?) Also, I’m allergic to nuts (except for almonds), but no problem if you’ve already bought food :)

1
Chris Popa
6mo
Hi Lucas, it is more of a dinner than snacks so it should be fine to not eat beforehand. Also, we will make sure there are no nuts in the food except for almonds. :)

Hi! I'm visiting from Australia and would love to meet some new people in EA Berlin. Is it ok for me to attend? :) 

1
MartinWicke
6mo
Hi Lucas, yes of course - cool that you‘re visiting Berlin! :-) Feel warmly invited to join any events of your interest we post here on the forum.

I very much agree with what Kyle said. 

Please consider focusing on improving your mental health as priority, because your wellbeing matters, and you deserve to live a flourishing and joyful life. 

I also agree that IQ is not particularly important in certain jobs, and there is something impactful for everyone if you enjoy it and your heart is in the right place. 

Good luck :) 

Hi team, thanks for your amazing work. Just out of curiosity, is there any particular reason climate charities are no longer recommended?

6
Luke Freeman
1y
Thanks Lucas! The Founders Pledge Climate Change Fund is recommended under "multiple cause areas":  https://www.givingwhatwecan.org/best-charities-to-donate-to-2023#toprated-funds-working-across-multiple-cause-areas CATF and TerraPraxis are also available on the donation platform (but not "top rated" based on the recommendations of our trusted evaluators): https://www.givingwhatwecan.org/donate/organizations 

Thanks very much Bob, appreciate the hot take! I'll get in touch if we'd like any more detail :) 

Wow fascinating, thanks for this post Vasco!

I'd be inclined to take a Bayesian approach to this kind of cost-effectiveness modelling, where the "prior evidence" is the estimated impact on lives saved. This is something we have strong reason to believe is good under many world views. Then the "additional evidence" would be the reduction in insect welfare caused by deforestation. I'm just so very uncertain about whether the second one is really a negative effect that I think it would be swamped by the impact on lives saved. This is because we have several st... (read more)

2
Vasco Grilo
1y
Hi Lucas, Thanks for engaging! I think the approach you are suggesting is very much in line with the one of section "Applying Bayesian adjustments to cost-effectiveness estimates for donations, actions, etc." of this post from Holden Karnofsky. I used to apply the above as (CE stands for cost-effectiveness, E for expected value, and V for variance): * E("CE") = "weight of modelled effects"*E("CE for modelled effects") + "weight of non-modelled effects"*E("CE for non-modelled effects"). * "Weight of modelled effects" = 1/V("CE for modelled effects")/(1/V("CE for modelled effects") + 1/V("CE for non-modelled effects")). This tends to 1 as the uncertainty of the non-modelled effects increases. * "Weight of non-modelled effects" = 1/V("CE for non-modelled effects")/(1/V("CE for modelled effects") + 1/V("CE for non-modelled effects")). This tends to 0 as the uncertainty of the non-modelled effects increases. If the modelled effects are lives saved in the near term, and the non-modelled effects are the impact on the welfare of terrestrial arthropods (which are not modelled by GW), V("CE for modelled effects") << V("CE for non-modelled effects"). So, based on the above, you are saying that we should give much more weight to the lives saved in the near term, and therefore these are the driver for the cost-effectiveness. I believe the formula of the 1st bullet is not correct. I will try to illustrate with a sort of reversed Pascal's mugging. Imagine there was one button which would destroy the whole universe with probability 50 % when pressed, and someone was considering whether to press it or not. For the sake of the argument, we can suppose the person would certainly (i.e. with probability of 100 %) be happy while pressing the button. Based on the formula of the 1st bullet, it looks like all weight would go to the pretty negligible effect on the person pressing the button, because it would be a certain effect. So the cost-effectiveness of pressing the button would

Hi Bob, thanks for this post, really interesting stuff. 

I'm a researcher at the Shrimp Welfare Project. Do you have a sense of whether any of the broad conclusions are likely to apply to shrimps as well, or do you think it would require an entirely new research project? 

Thanks, 

Lucas 

6
Bob Fischer
1y
Great question, Lucas. My hunch is that all the broad conclusions probably apply, though I'd want to think through the details more carefully before standing behind that claim. I suppose one thing that really affects my thinking is whether the organism has to navigate its environment in search of resources. My impression is that the youngest shrimp aren't doing this; they're just being carried along like plankton. So, that lowers my estimation of their capacities to something more like grubs than juvenile crickets. But of course I haven't investigated this at all, so please don't put too much weight on that hot take! Happy to discuss if it would be helpful; feel free to DM me.

I love this! Happy holidays :) 

1
Angelina Li
1y
<3 happy holidays!

I'd like to see someone in the EA community do some work related to preventing bullying, which seems likely to be one the most intense forms of suffering for children. 

This is an awesome and beautifully written post, thanks James!

Thanks so much for your awesome work!! :) 

What is StrongMinds' room for more funding, and do you expect the cost-effectiveness of the marginal dollar (ie. additional funds) to be any worse than the average cost-effectiveness of StrongMinds?

1
Sean Mayberry
1y
StrongMinds seeks to treat 300,000 women and adolescents between 2022 through 2024. To do that, we need to raise $30M U.S. dollars. To date, we've raised about $13M, which leaves us at a shortfall of $17M. So there is lots of room for more funding! In 2022 we expect the cost to treat one patient will be $105 USD. By the end of 2024, we anticipate the cost per patient will have decreased to just $85. We will continue to reduce the cost of treating one woman even while our numbers increase. This is through effective scaling and continuing to evaluate where we can gain more cost savings. A donation to StrongMinds will be used as effectively and efficiently as possible. And when you think about what it costs for therapy in the United States, to spend just $105 and treat a woman for depression is a pretty incredible feat. 

Fantastic work - thank you! 

Re Jalil et. al (2022), it's interesting to see there was a decrease in poultry/fish consumption as a result of climate change messaging (in addition to red meat). My prior concern would've been that people might simply switch from red meat to poultry/fish. For those interested in the general topic, note also this meta-review on interventions that influence animal-product consumption. 

Ah yes, my apologies, I meant natural experiments (or in the case of Croke 2019, a natural experiment caused by an actual experiment). 

I suppose it is possible deworming would have a much smaller effect when children also receive these other interventions. However, I would've thought many children currently being treated for worms are also receiving such interventions, therefore making it decision-relevant for GiveWell-funded deworming programs? 

Thanks for this post, the critique of GiveDirectly seems particularly compelling and important. 

On the issue of effects on males vs females, were you able to look into whether they may have converged towards more homogenous effects over time? It seems most of the eradication campaigns studied in the papers listed happened in the 1950s - I would suspect labour market opportunities are significantly stronger for women today, though I haven't looked at the data or whether this is true for the low-income countries where GiveWell's malaria charities do the... (read more)

5
JoelMcGuire
1y
Yeah, I think increasing labor participation rates by women would mitigate this effect seen in the historical data. However, in some places like South Asia it's remained low (and even declined) since the 90s.  </div>

Note that  if observational (i.e. non-experimental) studies are being included, one would probably also want to consider Croke 2019, which shows null effects on literacy and numeracy. 

There is also Makamu et al. 2018, but I don't think the natural experiment is very plausible (they use variation in which regions had deworming campaigns, but this is likely to be correlated with other policies/economic factors). 

2
JoelMcGuire
1y
To be clear, I primarily cite natural experiment which I'd argue are not evidentially equivalent to observational studies.  Aside: I only skimmed Croke (2019), but are the null effects surprising given that control children are already given a big boost? -- Plausibly including other GiveWell top charity interventions such as vitamin-A supplementation and vaccines? 
2
MHR
1y
Very nice find on Croke (2019), I wasn't aware of it! I think it's worth mentioning that though their results aren't statistically significant, their point estimates are positive for the literacy and numeracy effects. Since GiveWell is basically meta-analyzing the existing literature in trying to come up with its effect size estimates, there's a meaningful difference between a positive point estimate and a zero or negative point estimate in the case where a study doesn't find statistically significant results. 

Hi Holly, we’re not aware of how toxic ammonia is for other aquatic life. We believe it is always toxic, but some species may be more tolerant than others. Fish Welfare Initiative notes here that ammonia from mariculture farms may threaten aquatic life due to harmful algae blooms.

9
Holly_Elmore
1y
Thanks!

Hi MHR, thank you very much for your questions, these are important considerations! 

1. We certainly aim to consider the long-term effects on the total number of shrimps farmed when designing our interventions. Though we have not yet had an opportunity to precisely model the net effect, we expect a full analysis would need to account for: 

  • The reduction in mortality due to improved shrimp health
  • The opportunity for farmers to produce larger shrimps (and hence fewer individuals) due to improved health 
  • The long-term impacts of profitability on sh
... (read more)
6
MHR
1y
Thanks so much for your responses! Shrimp welfare is such a new field, so there are  bound to be questions we don't know the answers to yet. Thanks so much for the work you're doing, and I look forward to seeing how all the interventions you're working on turn out in the future!

Thanks for this thoughtful post Carolina! I would second Karthik's note here - I think there have also been a few other GE studies which show contradictory results, so it's not clear that the spillover effects would be positive once inflation and exchange rates effects are taken into account. Others have also raised concerns about possible negative pyschological spillovers, though from memory I think GiveDirectly typically provides cash to everyone in a village, which may mitigate this issue. 

Thanks for writing this up Joseph, these are really valuable questions to raise. I'd be particularly excited to see someone do a systematic review of spillovers on the control group after developmental interventions. 

Hi everyone, 

In this recent critique of EA, Erik Hoel claims that EA is sympathetic towards letting AGI develop because of the potential for billions of happy AIs (~35 mins) . He claims that this influences EA funding to go more towards alignment rather than trying to prevent/delay AGI (such as through regulation). 

Is this true, or is it a misrepresentation of why EA funding goes towards alignment? For example, perhaps it is because EAs think AGI is inevitable or it is too difficult to delay/prevent? 

Thanks very much! 

Lucas 

2
Lucas Lewit-Mendes
1y
Interesting, thanks both!
2
n99
1y
I can't speak for the donors, but only trying to prevent AGI doesn't seem like a good plan. We don't know what's required for AGI. It might be easy, so robustly preventing it would likely have a lot of collateral damage (to narrow AI and computing in general). Doing some alignment research is nowhere near as costly, and aligned AI could be useful.
2
Achim
1y
While I am also worried by Will MacAskill's view as cited by Erik Hoel in the podcast, I think that Erik Hoel does not really give evidence for his claim that "this influences EA funding to go more towards alignment rather than trying to prevent/delay AGI (such as through regulation)".

I don't know too much about this topic, but this might provide some useful resources? :) This is a really important topic, so hopefully someone in the community will be able to review the research at some point! 

Thanks for this really well-written post, I particularly like how you clarified the different connotations of longtermism and also the summary table of cost-effectiveness. 

I think one thing to note is that an X-risk event would not only wipe out humans, but also the billions of factory farmed animals. Taking into account  animal suffering would dramatically worsen the cost-effectiveness of X-risk from a neartermist point of view. I think this implies longtermism is necessary to justify working on X-risk (at least until factory farming is phased out). 

While I don't necessarily agree with Matty's view that total utilitarianism is wrong, I think this comment highlights a key distinction between a) improving the lives of future people and b) bringing lives into existance. 

The examples in this post are really useful to show that future people matter, but they don't show that we should bring people into existance. For example, if future people were going to live unhappy lives, it would still be good to do things that prevent their lives from being worse (e.g. improve education, prevent climate change, p... (read more)

Sorry I'm a bit late to the party on this, but thanks for the well-researched and well thought-out post. 

My two cents, as this line caught my eye: 

Notably, working on these issues can often improve the lives of people living today (e.g. working towards safe advanced AI includes addressing already present issues, like racial or gender bias in today’s systems). 

I think the line of reasoning concerns me. If working on racial/gender bias from AI is one of the most cost-effective ways to make people happier or save lives, then I would advocate&nb... (read more)

Such an amazing talk, well done!! :) 

Thanks for the response Samuel, would be interesting to hear GiveWell's rationale on using the log of average(earnings+consumption). 

Hi Joel, thanks for your response on this! 

I think my concern is that we can only "illustrate what would happen if GiveWell added decay to their model" if we have the right starting value. In the decay model's current form, I believe the model is not only adding decay, but also inadvertently changes the total earnings effect over the first 11 years of adulthood (yet we already have evidence on the total earnings effect for these years). 

However, as you noted, the main point certainly still holds either way. 

As a separate note, I'm not sure if it was intentional, but it appears HLI has calculated log effects slightly differently to GiveWell. 

  • GiveWell takes the average of earnings and consumption, and then calculates the log change.
  • HLI does the reverse, i.e. calculates the log of earnings and the log of consumption, and then takes the average. 
  • If we were to follow the GiveWell method, the effect at the second follow-up would be 0.239 instead of 0.185, i.e. there would be no decay between the first and second follow-up (but the size of the decay betwee
... (read more)
4
Samuel Dupret
2y
Thank you for your comment! Indeed, we did take the average of the logs instead of the log of the averages. This doesn’t change the end and start point, so it wouldn’t change the overall decay rate we estimate. We could do more complex modelling where effects between KLPS2 and KLPS3 see small growth and KLPS3 and KLPS4 see large decay. I think this shows that the overall results are sensitive to how we model effect across time. See Figure 4 of the appendix, which shows, whether in earnings or in consumption, that the relative gains, as shown by the log difference, decrease over time. We used the pooled data because it is what GiveWell does. In the appendix we note that the consumption and earnings data look different. So, perhaps a more principle way would be to look at the decay within earnings and within consumption. The decay within earnings (84%) and the decay within consumption (81%) are both stronger (i.e., would lead to smaller effects) than the 88% pooled decay.
6
Karthik Tadepalli
2y
Average of the log is more principled and I'm pretty surprised that givewell did it the reverse. These two quantities are always different (Jensen's inequality) and only one of them is what we care about. Log increase in consumption/income represents the % increase in that quantity. We want to find the average % increase across all people, so we should take the average of the log increase.

You can find their replicability adjustment here, and explanations are in this doc :)

3
Vaidehi Agarwalla
2y
Thank you! 

Full disclosure: I'm the primary author of a yet to be published SoGive report on deworming, however I'm commenting here in a personal capacity. 

Thanks for this thought provoking and well-written analysis! 

I have a query about whether the exponential decay model appropriately reflects the evidence: 

  • If I understand the model correctly, this cell seems to imply that the annual consumption effect of deworming in the first year of adulthood is 0.006 logs.
  • As HLI is aware, this is based on GiveWell's estimated annual earnings effect - GiveWell get
... (read more)
6
JoelMcGuire
2y
Thank you for your comment Lucas! Looking forward to seeing your forthcoming report.  Firstly, to clarify, we are doing a comparison between GiveWell’s model without decay and with decay. So to make the closest comparison possible we use the starting value and the time values that GiveWell uses. Rows 17, 18, and 19 of their CEA show the values they use for these. They consider the effects of starting 8 years after the deworming ends (~when participants start joining the labour force, see here) and continuing for 40 years with 0.006 each year. We get the same (similar because of our discretisation) total effects as GiveWell of 0.115 (0.113) for their model and show that if we use the exponential decay, we get a ~60% smaller total effect of 0.047. While it’s plausible there’s a better value to start with; we’re trying to illustrate what would happen if GiveWell added decay to their model. It’s unclear if they would also change the starting value too, but seems like a plausible choice.    The advantage of exponential decay is that it is based on % and so we can extract it from the study and use it on any start value and period, as long as we use the same as GW on these, we can get a proportional decrease in the effect.  We also considered linear decay. When we used linear decay, we found that the reduction in benefits is more dramatic: an 88% reduction. With linear decay, we had to change the start value, but we did this both for the constant effect model and the decay models so we could compare the proportional change. Of course, a more complex analysis, which neither ourselves nor GiveWell present, would be to model this with the whole individual data. The main point here is that the effect is very sensitive to the choice of modelling over time and thereby should be explicitly mentioned in GiveWell’s analysis and reporting. I think this point holds.   
5
Lucas Lewit-Mendes
2y
As a separate note, I'm not sure if it was intentional, but it appears HLI has calculated log effects slightly differently to GiveWell.  * GiveWell takes the average of earnings and consumption, and then calculates the log change. * HLI does the reverse, i.e. calculates the log of earnings and the log of consumption, and then takes the average.  * If we were to follow the GiveWell method, the effect at the second follow-up would be 0.239 instead of 0.185, i.e. there would be no decay between the first and second follow-up (but the size of the decay between the first and third follow-up would be unaffected).  * If the decay theory relies only on a single data point, does this place the theory on slightly shakier ground?  I don't have a good intuition on which of these approaches is better. Was there any rationale for applying the second approach for this calculation?

Thank you for writing this Mitra, it's always valuable to hear critiques of current approaches in the EA community. As Peter noted above, your experiences and views would be greatly valued by the community. 

I will attempt to respond to some of these questions, but note that my responses may not reflect the views of everyone in the community, and I may miss some crucial points. 

  • Are you effective enough to notice that you could be 10x more effective if instead of selling wells to villages, you focused resources of finding and supporting local entre
... (read more)
3
Mitra
2y
Thanks Lucas I should have numbered the points, would make it easier to reference ! I agree with the critique in that post you linked - RCTs will rarely show you what intervention will be catalytic and change things in the long run, it will systematically under-prioritize high-risk, and ignore approaches that require refining to get to scale. This is why I don't pay much attention to GiveWell, though it might catch the first kind of saving (a cheaper well) but GiveWell's methodology - like those of most other charity evaluators - is designed to ignore most of the other questions that determine effectiveness. For example their RCT won't be back in 10 years to see if the wells are still working. The rest of that post you refer to is about economic arguments (direct versus indirect poverty elimination) - from what I've seen it is probably only partially correct (and I haven't had time to read the whole article) , GDP is lousy measure of "happiness" and the GDP/capita measure also ignores HOW that wealth is spread, inequality not only means that a rich country can have a lot of very poor (e.g. in the US) but inequality is itself a significant cause of unhappiness. Reasonable people could hold reasoned positions on different sides of that argument so I don't want to dive too deep. In terms of impact measurement, I argue that it mostly collects meaningless numbers based on experience in the field, the measurement is typically designed to gather the numbers the donors want, not the numbers that actually reflect an increase in effectiveness. More importantly collecting and reporting data is a significant cost, often as much as 5-10% of the total budget, and that means 5-10% of the budget is diverted from creating impact to measuring it. Any effective organization is gathering data, but is gathering data that help it determine whether what its doing is effective, so for example it might just take a small time-bound sample - enough to know it should change something, and

Thanks for writing this Caroline, really interesting post! I think it's probably true that having talented people doing important things work really hard is higher impact than having people donate a little bit more money. 

However, I am concerned about the idea that one should prioritize their impact over relationships with family, friends, romantic partners, or children, for two reasons: 

  1. I think it's important to note that, personally, donating 10-20% of my income to effective charities literally makes zero difference to my life enjoyment.* But n
... (read more)

Thanks for writing this up Rumtin and Krystal! 

Does the scope of the project allow for engagment with academics as well as policy-makers/public servants? While there obvious risks with expanding the scope too broadly, I wonder whether collaboration with academia could be valuable for research efforts. There is also the possibility that some academic work (e.g. gain-of function research) could undermine policy efforts, so perhaps coordination between EA-aligned policy-makers/public servants and academics could reduce this risk? 

2
krystal_h
2y
Thanks for the comment Lucas! Apologies for the delayed response.  We are definitely hoping to coordinate with academics, policymakers and public servants. That will come at a later stage though, after we've completed the prioritisation and deep-dives of the issues we found. 

Thanks for writing this up!

This post does resonate with me, as when I was first introduced to EA, I was sceptical about the idea of "discussing the best ways to do good". This was because I wanted to volunteer rather than just talk about doing good (this was before I realised how much more impact I could have with my career/donations) and I think I would’ve been even more deterred if I’d heard that donated funds were being spent on my dinners.

However, it sounds like my attitude might have been quite different to others, reading the comments here. Also, I suspect I would’ve ended up becoming involved in EA either way as long as I heard about the core ideas.

Thanks Nathan, that would make a lot of sense, and motivates the conversation about whether CEA can realisticly attract as many people through advertising as Goldman etc. 

I guess the question is then whether: 

a) Goldman's activities are actually effective at attracting students; and

b) This is a relevant baseline prior for the types of activities that local EA groups undertake with CEA's funding (e.g. dinners for EA scholars students)

Hi Jessica, 

Thanks for outlining your reasoning here, and I'm really excited about the progress EA groups are making around the world. 

I could easily be missing something here, but why are we comparing the value of CEA's community building grants to the value of Mckinsey etc? 

Isn't the relevant comparison CEA's community building grants vs other EA spending, for example GiveWell's marginally funded programs (around 5x the cost-effectiveness of cash transfers)? 

If CEA is getting funding from non-EA sources, however, this query would be i... (read more)

I'm obviously not speaking for Jessica here, but I think the reason the comparison is relevant is that the high spend by Goldman ect suggests that spending a lot on recruitment at unis is effective. 

If this is the case, which I think is also supported by the success of well funded groups with full or part time organisers, and that EA is in an adversarial relationship to  with these large firms, which I think is large true, then it makes sense for EA to spend similar amounts of money trying to attract students. 

The relvent comparison is then comparing the value of the marginal student recurited with malaria nets ect. 

Really interesting and well-written post about the Australian political context! Do you think EA grant makers should consider funding political campaigns by minor parties, or would you prefer to see EA-aligned volunteers/staff leverage other sources of funds?

Thanks, Lucas

2
Ren Ryba
2y
Thanks Lucas - good question. If Farrer's theory is on the right track (which I think it is), then traditional orgs and political parties are substitutable strategies to achieve any particular policy goal. The most effective strategy would depend on the goal and the context. Given this, it would make perfect sense for EA grant makers to also consider funding political campaigns by minor parties. I've seen that grant makers often explicitly exclude political parties, which I gather is an understandable concession to optics. An argument against this is neglectedness - at least in my experience with the AJP, the party can readily generate its own funds through fundraising around election time. The government also provides funding for parties that achieve a particular threshold (not sure if this happens in other countries). Since minor political parties have access to these two sources of funding, this is a good reason why grant makers might choose not to fund political parties. To me, it would make sense for EA grant makers to consider funding campaigns (subject to optics considerations of course), but it would also make sense for grant makers to require a strong argument why the political party in question can't get its funding from sources like these.

Thank you for raising some interesting concerns JP.

I just wanted to note that the value of a market for bednets may be small relative to the value of philanthropic funding for several reasons: 

  • Having gone down the philanthropy path, ceasing to provide bednets philanthropically now would be unlikely to lead to a flourishing bednet market. See more on this here under "People may not purchase ITNs because they are unavailable in local markets or because they expect to be given them for free"
  • There are many reasons people may buy fewer bednets in a market
... (read more)
2
JPHoughton
2y
Thanks Lucas.  Agree that these may be reasons for 100% coverage to be a reasonable philanthropic target which would be unachievable through commercial means.   Your first point includes the idea that some people may not purchase ITNs because they expect to be given them for free.  This reinforces the idea in the essay that there is an extra cost to such distributions as nobody can make a living from selling nets in areas where people think the price should be zero.

Thanks, these are really interesting and useful thoughts!

Thanks very much Saulius, that all makes sense! 

Happy new year! 

Load more