I am open to work.
Vasco Grilo🔸
Karma: 8,776
Thanks for sharing!
The Shrimp Welfare Project (SWP) is one of my top recommendations too. Do you have any thoughts on my estimate that the cost-effectiveness of Fish Welfare Initiative’s (FWI’s) farm program from January to September 2024 was only 0.0111 % of SWP’s past cost-effectiveness? I guess FWI will become more cost-effective in the future, but I do not see how they would come close to SWP.
I am surprised you are recommending the Animal Legal Defense Fund (ALDF) without doing any cost-effectiveness analysis of their work. Here is a quick illustration of why I think they are much less cost-effective than SWP:
SWP has helped 1.50 k shrimp/year/$.
You say that, “since 2021, ALDF has had average annual expenditures of $16,189,282”. So they would have to have helped 24.3 billion animals per year (= 1.50*10^3*16.2*10^6) to have helped as many animals per year per $ as SWP.
9.91 billion land animals were slaughtered for meat in the United States (US) in 2023. So ALDF would have to have helped 2.45 (= 24.3*10^9/(9.91*10^9)) times as many animals per year as those slaughtered for meat in the US in 2023 to have helped as many animals per year per $ as SWP. This seems way too much to be feasible considering their interventions do not target aquatic animals, or land invertebrates.
I understand impact is not directly proportional to the number of animals helped, but I believe the impact per animal helped may well be larger for SWP than ALDF. I estimate SWP averts the equivalent of 0.0426 DALYs per shrimp helped, which I think is as good as preventing the existence of 0.463 broilers (= 0.0426/(0.754*0.122)) in a conventional scenario.
I believe that kids having a preference not to be slaughtered, is a very good reason not to have kids, when someone decides to have happy kids in order to slaughter and eat them.
Are you also against abortion? If not, why would it be fine to kill a human right before birth, but not right after it? I do not see a meaningful difference ignoring 2nd order effects, which I agree may be a valid reason for opposing the farming of animals with positive lives, but you seem to oppose this even ignoring such effects.
The welfare-discounted people cannot complain against their welfare being discounted, because in the chosen optimal situation, those people do not exist. This idea creates the procreation asymmetry: it is ok not to cause the existence of a happy person, but not ok to cause the existence of a miserable person (all else equal). See my theory of mild welfarism: https://stijnbruers.wordpress.com/2022/08/23/mild-welfarism-avoiding-the-demandingness-of-totalitarian-welfarism/ and the asymmetry: https://stijnbruers.wordpress.com/2024/06/17/the-asymmetry-in-population-ethics-and-deontological-ethics/
If you think future human-years of people who do not exist yet can be discounted, do you believe future human-years of people who already exist can be discounted too? If not, what is the relevant difference between future human-years of current and future people? One’s future self can be seen as a future person similar to one’s current self.
I agree with taking that perspective of the animals. However, in this context it’s complicated: if the animal does not exist, it has no perspective and also no preference for existence with a positive welfare.
You seem to be implying that increasing the population of farmed animals with positive lives is not good because they do not have a preference for existing before existing. I do not see how one can believe this without also thinking that decreasing the population of farmed animals with negative lives is not good because they do not have a preference for not existing before existing. An asymmetry does not make sense to me.
And if you bring an animal into existence, give it a positive live, use that animal for food and slaughter it, it is likely that from that animal’s perspective, that animal would prefer a longer, more positive live at an animal sanctuary.
Farmed animals with positive lives would prefer to live longer lives, but they would also prefer to live a shorter positive live over not having lived at all. The same applies to humans. Do you think people preferring to live longer (while they consider their lives worth living) is a reason not to have kids? I guess not because you see death from disease differently than death from slaughter, but none are voluntary, in the sense the human or animal would prefer to continue living for longer if their life was still positive, so I do not know what would be the relevant difference ignoring 2nd order effects. Are these the main reason for you thinking that farming animals with positive lives would be bad?
Arthropods tend to have higher reproduction rates (more offspring per adult) than large herbivores, and I believe high reproduction rates negatively correlate with welfare of the offspring. Also: as the reproduction rate of most farmed animals is higher than those of wild large herbivores, replacing farmed animals by wild large herbivores (when agricultural land is replaced by grasslands with wild horses) would increase animal welfare and decrease suffering.
I want to maximise the additional welfare per $ spent, which is the product between the additional welfare per animal-year improved, and animal-years improved per $ spent. I think the additional welfare per animal-year improved is higher for vertebrates than invertebrates given the larger welfare range of vertebrates. However, I believe one can get more animal-years improved per $ spent on helping invertebrates than on helping vertebrates. I expect this last effect to be stronger, as I estimate invetebrate welfare interventions are much more cost-effective than vertebrate welfare ones, and that invertebrates have a larger welfare range per calorie consumption than vertebrates. So I think it is still worth targeting invertebrates, which have a high reproduction rate, regardless of whether they have positive or negative lives.
Hi Abraham,
The Insect Institute has 3 peer-reviewed publications listed on their website.
Hi Yarrow,
Assuming there were only 4 experts predicting full automation by 2048 with 10 %, 20 %, 30 %, and 40 % chance, the middle half of experts would be predicting full automation by 2048 with 20 % to 30 % chance. If there were 1 k experts each predicting the probability of full automation by multiple dates, one could infer 1 k predictions for the probability of full automation for any given date, order such preductions by ascending order, and then report the 25th and 75th percentile predictions as the lower and upper bound of the middle half of predictions.
Thanks, Richard! That makes sense to me.
Thanks for sharing, Richard! Would you also support factory-farming animals with sufficiently good lives, ignoring effects on wild animals? I would.
Hi Angelina,
I think the credits could be set up such that farmers would profit from electrically stunning more shrimp, but not from increasing shrimp production. For example, if conventional and electrically stunned shrimp marginally costed 10 and 10.1 $/kg, the credits could cost just slighty more than 0.1 $/kg (= 10.1 − 10). If conventional shrimp marginally costed 10 $/kg, it would not make sense to increase production to sell credits at slightly more than 0.1 $/kg if the additional electrically stunned shrimp were not bought by consumers, because this would result in a marginal loss of slightly less than 9.90 $/kg (= 0.1 − 10).
Hi Angelina and Austin,
I am not sure this helps you, but I think the idea is as follows. Some retailers would like to be resposible for electrically stunning more shrimp, but have suppliers which are not willing to do so. Those retailers could buy stunning credits as a way of certifying they have been responsible for electrically stunning more shrimp by paying other suppliers to do so. These suppliers may initially be uncertain about whether those retailers will buy their stunning credits, so SWP commiting to buying their credits would make them more willing to electrically stun more shrimp.
Thanks for the post, Aaron!
But then we have the issue of an audited supply. For GFP, this was solved through certification schemes, but there is currently no certification scheme for stunned shrimps. This is where we’re hoping Precision Aquaculture [“on-farm continuous data monitoring”], or Precision Welfare, can come into play.
I am not a fan of the term precision welfare. It makes it sound like data is being monitored with the purpose of increasing welfare, whereas I assume the goal is very often increasing profit regardless of whether welfare is made better or worse.
Certification schemes are slow to adapt to changes and often have an unusual incentive structure where their customers are the producers themselves, so trying to be too innovative or progressive can cost them business.
The organisations ensuring the shrimp were electrically stunned, via precision aquaculture or other methods, would in practice be certifiers. I understand certifiers cannot be too innovative, but this applies to both current certifiers, and prospective ones that would rely on precision aquaculture.
We think Precision Welfare could be another way of verifying that the shrimps have been stunned. Installing devices onto shrimp stunners, for example, could automate the tracking of stunned shrimps. Once a shrimp has been stunned, we can verify that it has been stunned and generate a unique credit. Another reason I’m excited about this is that it incentivises the installation of tracking devices on farms and automates the transfer of that data to a verifiable 3rd party database.
Producers will have an incentive to put shrimp which is not fit for consumption (for example, which died before slaughter time) in the electrical stunners to earn more stunning credits. Ideally, one would only assign credits to shrimp fit for consumption. I assume this can be done by ensuring the shrimps were moving before they are electrically stunned. In addition, no shrimps moving afterwards is a necessary condition of effective stunning, but I guess it is not sufficient.
As this is already the prevalant method of trading shrimps in the industry, we assume that credits will be assigned per kg, rather than per shrimp.
I wonder whether assigning credits per shrimp would be worse than per kg. Assigning credits per shrimp would incentivise farmers to slaughter more shrimp, which can be achieved by slaughtering shrimp at a younger age when they are smaller. It is unclear to me whether this would be good or bad, as I do not know whether the last stages of shrimps’ lives before slaughter are better or worse than a random moment of their lives before slaughter. On the other hand, I guess the growth rate of shrimps slows down as they age, such that slaughtering younger shrimp would imply an increase in shrimp-years/shrimp-kg, and therefore a larger population of shrimps. I believe this would be bad because I estimate shrimps have negative lives, even with electrical stunning.
Thanks for the comprehensive update, Aaron! The Shrimp Welfare Project is one of my favourite charities.
We’ve already had some promising conversations in this space, such as with We’ve already had some promising conversations in this space, including with companies in Singapore, Germany, and the US.
Nitpick. You repeated the 1st part of the sentence twice.
Thanks for all your work on the Forum, JP!
Thanks, Stijn.
There are several (strong, according to many people) moral objections to have happy animal farming, where farmed animals with a net-positive welfare are slaughtered and eaten.
The most important for me is the perspective of the animals, and they would want to be farmed if they had positive lives. I am wary of going against the will of the most affected beings. The will of the animals has to be inferred by humans, but there are cases where one can confidently say animals would prefer to have existed over not having been born. For example, if their conditions were equivalent to those of well treated pets, or healthy animals in sanctuaries.
As for the effects on wild animals, I guess many of the agricultural land that is now used for animal farming (including feed crops), can best be turned into grasslands with large herbivores (wild horses,...). Those types of ecosystems may have lower animal suffering densities (amount of suffering per square mile) than animal farming.
A lower population density would decrease total suffering, but it would also tend to decrease total happiness. Ideally, one would expand ecosystems with lots of positive lives (for instance, happy arthropods). In any case, it could still be the case that the decrease in cropland resulting from a lower consumption of animal-based products leads to an increase in the number of wild animals with negative lives which makes alternative proteins harmful at least nearterm.
Thanks, Marc!
Considering there are 10^10 (= 10^(20 − 10)) times as many marine arthropods as humans, and my guess that the absolute value of the welfare of a random arthropod-year is 5*10^-5 of the welfare of a random human-year[1], I estimate the absolute welfare of marine arthropods is 500 k (= 10^10*5*10^-5) times the welfare of humans. Since there is lots of uncertainty about whether marine arthropods have positive or negative lives now, and whether humans or our descendents will improve them in the future, I do not know whether the future welfare of life on Earth is positive or negative. So I would focus on improving welfare instead of decreasing the risk of human extinction. In any case, I think reducing the nearterm risk of human extinction would not be astronomically cost-effective even if the future welfare of life on Earth was robustly positive. I am also sceptical of the possibility of meaningful longterm impacts.
Thanks for this sequence, Neel!
Great work, Michael!
No amount of controlling for confounders fixes this problem.
I liked this post! I did not know Dynomight.
Thanks, Stuart. I think Ege’s point is that there is less variation in activities that humans have done for a long time, which does not apply to all sporty activities. I would say most of olympic gymnastics and professional basketball do not qualify, whereas walking and running do. The fastest marathon was run in around 2 h, whereas a random person with age 30 can maybe complete one in around 9 h moving at a walking pace of 4.67 km/h[1] (= 42⁄9), which is 4.5 (= 9⁄5) times as much time. In contrast, a random person would not beat the best chess players even if they had 4.5 times as much time (unless they had very little time like 4.5 s against 1 s for the top players).
- ^
Healthy young people can walk all day if needed even if they do not exercise regularly.
- ^
Great post, Tom! I strongly upvoted it.
I think the cost-effectiveness of advocating for cultivated meat is also lower due to the possibility of it decreasing the number of farmed animals with positive lives in the future. I guess significant replacement of animal-based products by cultivated meat would only happen in a few decades, and that farmed chickens in some regions will have positive lives then. I estimate broiler welfare and cage-free campaigns increase welfare per living time by 92.9 % and 80.4 %, which are not far from the increase of 100 % that would be obtained for improved conditions respecting neutral lives. For Ambitious Impact’s pain intensities, and my guess that hurtful pain is as intense as fully healthy life, I estimate broilers in a reformed scenario, and hens in cage-free aviaries have slightly positive lives.
In addition, I believe decreasing the consumption of animal-based products may be harmful due to effects on wild animals. Brian Tomasik thinks “it’s pretty unclear whether promoting vegetarianism reduces or increases total animal suffering, both when considering short-run effects on wild animals on Earth and when considering long-run effects on society’s values”.
I recommend donating to the Shrimp Welfare Project (SWP) not only because of super high cost-effectiveness ignoring non-target animals[1], but also minor effects on wild animals. Its main intervention only adds electrical stunning before slaughter, which does not change the amount of feed required, which I think would be the main way of impacting wild animals.
Interesting point, Stuart! I think when people say a given ability is easy to achieve they are referring to the chance of achieving it conditional on putting some effort into it. Sensorimotor tasks may be hard in the sense of requiring lots of effort (even if we forgot about it), but the vast majority of people achieve a great level conditional on the training effort of a typical upbringing. I guess there would be much greater variance in chess and maths ability conditional on a given amount of training.
Thanks for the clarifications.
You estimated the animals helped per $ for the Fish Welfare Initiative (FWI), and SWP. I think your analyses would be more valuable if you got estimates for the increase in welfare per $. I know one needs to make contentious assumptions to compute these, but I still think producing them is valuable such that people can see which organisations increase welfare the most per $ under their preferred assumptions. Making calibrated adjustments is harder without an underlying model. Animal Charity Evaluators (ACE) estimated the cost-effectiveness of some charities in terms of suffering averted per $, which I believe is quite similar to welfare increased per $[1].
I encourage you to clarify which are your criteria for recommending a charity. I think impact-focussed evaluators avoid large differences in cost-effectiveness among their recommendations[2]. So people may infer you do not think there are large differences in cost-effectiveness among your recommendations.
I think the impact of the interventions ACE assessed mostly comes from decreasing suffering, not increasing happiness.
Vince Mak, ACE’s charity evaluations manager, said “We do not recommend charities if there is a large enough gap between their expected marginal cost-effectiveness and that of our other charities”. GiveWell’s top charities save a life for 3.5 k to 5.5 k$.