I am a generalist quantitative researcher. I am open to volunteering and paid work. I welcome suggestions for posts. You can give me feedback here (anonymously or not).
Hi Jim. I agree trying to come up with counterexamples is useful. Below are 3 potential counterexamples I have given. @Anthony DiGiovanni 🔸 does not consider them counterexamples (see Anthony's replies for details).
1st example, which Elliot already quoted in this thread.
Consider these 2 options for what I could do tomorrow:
Torturing my family, and friends, and then killing myself. I would never do this.
My understanding is that you think it is "irreduciblyindeterminate" which of the above is better to increase expected impartial welfare, whereas I believe the 2nd option is clearly better.
Hi Anthony. Do you think the expected welfare of 2 states of the world which only differ infinitesimally can be incomparable? I do not see how this could be possible. For example, it feels super counterintuitive to me that, given 2 identical states, moving an electron by 10^-100 m in one of the states would make their expected welfare incomparable. I guess one can get from any state of the universe to another in an astronomical number of infinitesimal steps, and I believe any 2 states which only differ infinitesimally are comparable. So I conclude any 2 states are comparable too, even if it is very hard to compare them, to the point that I do not know if electrically stunning shrimps increases or decreases welfare in expectation.
Here is a 4th example. Consider these 2 actions:
Killing the 100 people who are expected to decrease the most the uncertainty about how to compare the expected value of different actions. For example, Bob Fischer who has worked on decreasing uncertainty about comparing welfare across species.
Grating 10 M$ to the 100 people above (100 k$ per person, but the grant size could vary). The 10 M$ would otherwise be spent torturing people as much as possible.
I think we are justifyed in c-preferring the 2nd action. I believe Anthony disagrees
Hi Rocky and Tom. Great post. I share your concerns.
Our worry is not that welfare tech is a bad idea (some of it is likely excellent). Rather, the worry is that it has several traits that make it vulnerable to becoming a bubble. It has a clean theory of change, no obvious need for mass public persuasion, compatibility with short AI timelines, appeal to tech-oriented funders and builders, and the ability to be framed as a silver bullet: build the tool, deploy the tool, reduce suffering at scale.
I estimated SWP's stunning work is 383 and 139 times as cost-effective as broiler welfare and layer cage-free corporate campaigns (neglecting effectson non-target animals). However, I think the estimates depend way more on welfare comparisons across species than the probability of the stunning work benefiting and harming shrimps.
Say the cost-effectiveness of SWP's stunning work is proportional to the difference between the probability of benefiting and harming shrimps, and that shrimps are either benefited or harmed. Suppose the probability of benefiting shrimps decreases from 75 % to 60 %, which means the probability of harming shrimps increases from 25 % to 40 %. The cost-effectiveness would become 40 % (= (0.6 - 0.4)/(0.75 - 0.25)) as large.
My estimate for the cost-effectiveness of SWP's stunning work is proportional to the value of 1 shrimp-QALY. I assumed this to be 9.34 % (= 0.031/0.332) of the value of 1 chicken-QALY. Suppose the value of 1 anima-QALY is instead proportional to the individual number of neurons, which I consider plausible. I calculate shrimps have 0.0389 % as many neurons as chickens. So the cost-effectiveness would become 0.416 % (= 3.89*10^-4/0.0934) as large.
Thanks for the clarifications. They made sense to me. At the same time, I believe the most cost-effective ways of building capacity to help any given group of animals will generally be optimised with such animals in mind.
Hi Aidan. You estimated a giving multiplier of 7 for 2025 (for 15 k$ per new 10 % Pledge), 6 for 2023-2024 (for 15 k$ per new 10 % Pledge), and 30 for 2020-2022 (for 22 k$ per new 10 % Pledge). Your estimates for the value per new 10 % Pledge are similar. So the giving multiplier has decreased overwhelmingly because the cost per new 10 % Pledge has increased?
I would think about it as I do about grains of sand. The statement "a N ft man is tall" is not logically true or false (regardless of the value of N). However, a taller man is more likely to be described as tall. It is very unlikely a man A is described as tall if they have an height of 5 ft (1.5 m). A man B who is 10^-10 ft taller will be a super tiny bit more likely to be described as tall than man A. However, in practice, one can assume A and B to be as tall. I am not seeing which decisions would be influenced by rounding a height difference of 10^-10 ft to exactly 0.
In contrast, rounding to exactly 0 a super small moral weight could have significant implications. Moral weight might be roughly proportional to the individual number of neurons (or, more plausibly, proportional to "individual number of neurons"^"exponent"), and soil invertebrates have way fewer neurons per individual than humans, but more neurons in total.
Hi @Laura Duffy . Below is a print from the doc describing the cost-effectiveness estimates.
Animal Charity Evaluators (ACE) estimated The Humane League's (THL's) cage-free work in 2025 helped 11 (2 to 44) chickens per $. How did you get the above range of 3.6 to 22?
More importantly, ACE's estimate of 11 chickens helped per $ refer to 7.65 M$ spent by THL, but what matters to you is THL's marginal cost-effectiveness? I think this is closer to the cost-effectiveness of cage-free work funded through CCF than the cost-effectiveness of all cage-free work from THL. Relatedly, I woud like ACE to calculate the marginal cost-effectiveness of THL.
Nitpick. Would it be better to let people copy text from the doc? I shared a print because I am not able to copy the text.
claims we definitely want to accept like "0 grains is not a heap" or "a non-conscious non-living, mindless object isn't a moral patient"
I think 0 grains of sand could be a heap of sand under some definitions of being a heap of sand, and I also believe non-conscious non-living mindless objects could have moral value (and therefore be moral patients) under some definitions of consciousness, life, mind, and objects. For example, I think standard laptops could reasonably be described as non-conscious non-living mindless objects, but I would not assign them a moral value of exactly 0.
The big, undetermined, crucial considerations are: as I learn more, new, previously unconsidered effects are likely to outweigh the positive impacts I was evaluating; what trade-offs to make between pain intensity and duration; and how much to weigh different animals. The first consideration, I try to “cover” by updating over time if a new effect becomes sufficiently[21] evidenced. For the other two considerations, I try to cover them with a portfolio, but when that is not possible, I consider multiple options worthwhile under my procedure and tiebreak on other factors.[22]
The 3 considerations you highlighted above are my major uncertainties too, especially uncertainty about welfare comparisons across species. I think this can make it very unclear whether any intervention increases or decreases welfare/suffering/happiness (even in expectation). Have you considered prioritising interventions aiming to decrease the uncertainties instead or portfolios of interventios? Under your own views, any portfolio of interventions can easily increase or decrease animal suffering?
Thanks for the discussion which prompted me to look into it.
Hi Jim. I agree trying to come up with counterexamples is useful. Below are 3 potential counterexamples I have given. @Anthony DiGiovanni 🔸 does not consider them counterexamples (see Anthony's replies for details).
1st example, which Elliot already quoted in this thread.
2nd example.
3rd example.
Here is a 4th example. Consider these 2 actions:
I think we are justifyed in c-preferring the 2nd action. I believe Anthony disagrees
Hi Rocky and Tom. Great post. I share your concerns.
I very much agree.
I estimated SWP's stunning work is 383 and 139 times as cost-effective as broiler welfare and layer cage-free corporate campaigns (neglecting effects on non-target animals). However, I think the estimates depend way more on welfare comparisons across species than the probability of the stunning work benefiting and harming shrimps.
Say the cost-effectiveness of SWP's stunning work is proportional to the difference between the probability of benefiting and harming shrimps, and that shrimps are either benefited or harmed. Suppose the probability of benefiting shrimps decreases from 75 % to 60 %, which means the probability of harming shrimps increases from 25 % to 40 %. The cost-effectiveness would become 40 % (= (0.6 - 0.4)/(0.75 - 0.25)) as large.
My estimate for the cost-effectiveness of SWP's stunning work is proportional to the value of 1 shrimp-QALY. I assumed this to be 9.34 % (= 0.031/0.332) of the value of 1 chicken-QALY. Suppose the value of 1 anima-QALY is instead proportional to the individual number of neurons, which I consider plausible. I calculate shrimps have 0.0389 % as many neurons as chickens. So the cost-effectiveness would become 0.416 % (= 3.89*10^-4/0.0934) as large.
Thanks for the clarifications. They made sense to me. At the same time, I believe the most cost-effective ways of building capacity to help any given group of animals will generally be optimised with such animals in mind.
Hi Aidan. You estimated a giving multiplier of 7 for 2025 (for 15 k$ per new 10 % Pledge), 6 for 2023-2024 (for 15 k$ per new 10 % Pledge), and 30 for 2020-2022 (for 22 k$ per new 10 % Pledge). Your estimates for the value per new 10 % Pledge are similar. So the giving multiplier has decreased overwhelmingly because the cost per new 10 % Pledge has increased?
I would think about it as I do about grains of sand. The statement "a N ft man is tall" is not logically true or false (regardless of the value of N). However, a taller man is more likely to be described as tall. It is very unlikely a man A is described as tall if they have an height of 5 ft (1.5 m). A man B who is 10^-10 ft taller will be a super tiny bit more likely to be described as tall than man A. However, in practice, one can assume A and B to be as tall. I am not seeing which decisions would be influenced by rounding a height difference of 10^-10 ft to exactly 0.
In contrast, rounding to exactly 0 a super small moral weight could have significant implications. Moral weight might be roughly proportional to the individual number of neurons (or, more plausibly, proportional to "individual number of neurons"^"exponent"), and soil invertebrates have way fewer neurons per individual than humans, but more neurons in total.
Hi @Laura Duffy . Below is a print from the doc describing the cost-effectiveness estimates.
Animal Charity Evaluators (ACE) estimated The Humane League's (THL's) cage-free work in 2025 helped 11 (2 to 44) chickens per $. How did you get the above range of 3.6 to 22?
More importantly, ACE's estimate of 11 chickens helped per $ refer to 7.65 M$ spent by THL, but what matters to you is THL's marginal cost-effectiveness? I think this is closer to the cost-effectiveness of cage-free work funded through CCF than the cost-effectiveness of all cage-free work from THL. Relatedly, I woud like ACE to calculate the marginal cost-effectiveness of THL.
Nitpick. Would it be better to let people copy text from the doc? I shared a print because I am not able to copy the text.
I think 0 grains of sand could be a heap of sand under some definitions of being a heap of sand, and I also believe non-conscious non-living mindless objects could have moral value (and therefore be moral patients) under some definitions of consciousness, life, mind, and objects. For example, I think standard laptops could reasonably be described as non-conscious non-living mindless objects, but I would not assign them a moral value of exactly 0.
Hi Jo. Great post.
The 3 considerations you highlighted above are my major uncertainties too, especially uncertainty about welfare comparisons across species. I think this can make it very unclear whether any intervention increases or decreases welfare/suffering/happiness (even in expectation). Have you considered prioritising interventions aiming to decrease the uncertainties instead or portfolios of interventios? Under your own views, any portfolio of interventions can easily increase or decrease animal suffering?
Hi Richard. Nice post.