Thanks for the comment, Alene! I will focus on nematodes because I calculated the increase in the welfare of soil nematodes from increasing cropland to range from 90.5 % to 94.3 % of the increase in the welfare of soil nematodes, mites, and springtails, depending on the original biome.
I think decreasing the uncertainty about the probability of sentience of nematodes is less important than decreasing the uncertainty about whether they have positive or negative lives. I believe it would have to be unreasonably low for the effects on farmed animals to dominate. For example, I estimated eating chicken meat benefits soil animals 4.81 k times as much as it harms directly affected animals, so the effects on soil animals would have to become less than 0.0208 % (= 1/ā(4.81*10^3)) as large for the effects on directly affected animals to be larger. This could be achieved, for instance, if both the their probability of sentience, and welfare range given sentience became 1.44 % (= (2.08*10^-4)^0.5) as large. RP estimated a probability of sentience for nematodes of 6.8 %, 6.8 % that of humans of 100 %. As a result, for the effects of eating chicken meat on directly affected animals to be larger than those on soil animals, the probability of sentience of nematodes would have to drop by a factor 4.72 (= 0.068/ā0.0144) times as large as the factor linked to decrease in RPās probability of sentience from humans to nematodes.
I assume there is a probability of 93.2 % (= 1 ā 0.068) of soil nematodes having practically negligible welfare based on RPās probability of sentience for nematodes of 6.8 %. However, whether they have positive or negative lives in expectation is determined by what happens given sentience.
RPās probability of sentience for nematodes of 6.8 % may seem low, but people take care to decrease way smaller risks. For example, travelling 370 km by car only increases oneās risk of death by around 10^-6, but people still wear seat belts.
Got it. I think I understand what youāre saying. Iām not as good with math so Iām not sure if I followed the calculations. But to try to put what youāre saying in less mathy terms, I think youāre basically saying:
1) There are WAY WAY WAY more nematodes than farmed animals.
2) Nematodes are significantly less likely to be sentient than farmed animals.
3) But the fact that there are WAY WAY WAY more nematodes than farmed animals still means that, from an expected value perspective, one would still expect the effect of farming on nematodes to be much bigger than the effect of farming on farmed animals.
Is that right?
Like, if you could enter a deal where a person is guaranteed to pay you $1 up front, but in exchange you accept a 6% chance that the person will later take $4,810,000 from you, itād be a bad deal to make, even though the most likely outcome is you simply gain a dollar and donāt pay anything. Is that a good analogy?
Yes, that is basically right[1]. For example, I estimate nematodes are 7.76 % (= 0.068/ā0.876) as likely to be sentient as chickens (āsignificantly lessā), but that there are 16.7 billion (= 4.89*10^20/ā(29.2*10^9)) times as many soil nematodes as farmed chickens (āWAY WAY WAY moreā).
Yes, I think that analogy illustrates why I think broadly advocating for decreasing the consumption of animal-based foods tends to be harmful to animals.
With the caveat that what matters besides population is the welfare per animal-year, which is the product between the probability of sentience, welfare range given sentience, and welfare per animal-year as a fraction of the welfare range.
Thanks for the comment, Alene! I will focus on nematodes because I calculated the increase in the welfare of soil nematodes from increasing cropland to range from 90.5 % to 94.3 % of the increase in the welfare of soil nematodes, mites, and springtails, depending on the original biome.
I think decreasing the uncertainty about the probability of sentience of nematodes is less important than decreasing the uncertainty about whether they have positive or negative lives. I believe it would have to be unreasonably low for the effects on farmed animals to dominate. For example, I estimated eating chicken meat benefits soil animals 4.81 k times as much as it harms directly affected animals, so the effects on soil animals would have to become less than 0.0208 % (= 1/ā(4.81*10^3)) as large for the effects on directly affected animals to be larger. This could be achieved, for instance, if both the their probability of sentience, and welfare range given sentience became 1.44 % (= (2.08*10^-4)^0.5) as large. RP estimated a probability of sentience for nematodes of 6.8 %, 6.8 % that of humans of 100 %. As a result, for the effects of eating chicken meat on directly affected animals to be larger than those on soil animals, the probability of sentience of nematodes would have to drop by a factor 4.72 (= 0.068/ā0.0144) times as large as the factor linked to decrease in RPās probability of sentience from humans to nematodes.
I assume there is a probability of 93.2 % (= 1 ā 0.068) of soil nematodes having practically negligible welfare based on RPās probability of sentience for nematodes of 6.8 %. However, whether they have positive or negative lives in expectation is determined by what happens given sentience.
RPās probability of sentience for nematodes of 6.8 % may seem low, but people take care to decrease way smaller risks. For example, travelling 370 km by car only increases oneās risk of death by around 10^-6, but people still wear seat belts.
Got it. I think I understand what youāre saying. Iām not as good with math so Iām not sure if I followed the calculations. But to try to put what youāre saying in less mathy terms, I think youāre basically saying:
1) There are WAY WAY WAY more nematodes than farmed animals.
2) Nematodes are significantly less likely to be sentient than farmed animals.
3) But the fact that there are WAY WAY WAY more nematodes than farmed animals still means that, from an expected value perspective, one would still expect the effect of farming on nematodes to be much bigger than the effect of farming on farmed animals.
Is that right?
Like, if you could enter a deal where a person is guaranteed to pay you $1 up front, but in exchange you accept a 6% chance that the person will later take $4,810,000 from you, itād be a bad deal to make, even though the most likely outcome is you simply gain a dollar and donāt pay anything. Is that a good analogy?
Yes, that is basically right[1]. For example, I estimate nematodes are 7.76 % (= 0.068/ā0.876) as likely to be sentient as chickens (āsignificantly lessā), but that there are 16.7 billion (= 4.89*10^20/ā(29.2*10^9)) times as many soil nematodes as farmed chickens (āWAY WAY WAY moreā).
Yes, I think that analogy illustrates why I think broadly advocating for decreasing the consumption of animal-based foods tends to be harmful to animals.
With the caveat that what matters besides population is the welfare per animal-year, which is the product between the probability of sentience, welfare range given sentience, and welfare per animal-year as a fraction of the welfare range.
Got it. Thank you so much for explaining so patiently!