I count mussels and oysters as meat-plants, but not clams. Clams are capable of motility! My reasoning for this as the relevant boundary here is the educated guess that 'pain is evolutionarily expensive'. If an animal can't move, it probably doesn't have a sophisticated pain response. If an animal *can* move then I'm inclined to give it the benefit of the doubt. Still, their neurons number in the thousands so this isn't exactly an issue of serious concern to me, there's only so much pain that can be going on in a system that simple.
Less than a mussel (surely if nothing else, bacteria would also live within and on mussels), unsure about trees and not very interested in the question. I think on the 'pain is evolutionarily costly' theory it probably would be very costly for something as small as a bacterium and it could probably accomplish something similar without pain? Like, motility is necessary but not sufficient for establishing that a creature can feel pain in my model
I think the caring about the individual vs. groups difference is a pretty fundamental one.
Many political causes are not saying "I care more about people in this category", but rather "I care about this category as its own object". This is why you get people so concerned about "the great replacement" and genocide, while much less concerned about individual murder of larger numbers of people in larger demographic groups.
And the same applies to non-humans. A typical environmentalist finds the death and suffering of millions of pigeons basically irrelevant, but the extinction of pigeons would be a big deal. Thus we get these laws about how companies must protect the existence of *species*, but are free to harm as many *animals* as they want.
A big part of effective altruists' "weirdness" comes from not caring about preserving groups at all, and they focus only on the largest numbers of sufferers. These naturally tend to come from the most common species, which normal people care the *least* about.
There's also the whole aspect that group-thinking is pretty normal (which doesn't mean it's right necessarily). We evolved to get into groups and fight wars with other groups (just look around the world), so the effective altruist perspective (which tends to be the extreme form of the universalist 'there is no outgroup') is really weird to most people.
It might be tricky to interpret the results of this study, because people could be conflating the welfare of the thing with the thing's impact on aggregate welfare.
For example, it's very strange that the welfare of trees and mountains is given relatively high standing, because trees and mountains do not have welfare. However, trees and mountains do affect aggregate welfare.
Similarly, we might say people gave negative consideration to the welfare of murderers and child molesters because they want them to suffer (plausible), or perhaps they thought that the existence of murderers and child molesters has a negative impact on aggregate welfare.
There's an adjacent point in the comments about how people could be conflating the intrinsic importance of a person's welfare with the obligation to care for the person's welfare.
I do want to recognize that people really do value the intrinsic welfare of different groups differently (and even strangely), but some of the results only make sense on this alternate framing.
I agree. I also think this survey, as well as the moral circle expansion concept in general, unhelpfully fail to distinguish between *absolute moral worth* and *relative moral obligation*. Most people believe that animals are *absolutely* less morally valuable than humans, but that they have lesser *relative* obligations to foreigners than to their fellow-citizens, in the same way that they have special obligations to their family and friends.
I think I qualify as a "normie" EA. I identify with the movement, donate to EA charities, read some blogs, but I'm not otherwise involved in the subculture at all.
Your first few EA posts seemed dead-on to me, but this one doesn't feel like it's describing a belief system that I ascribe to, or that I've gotten the impression is the strong consensus belief of the community. I think moral circle expansionism is certainly a core EA belief, and your final point, if a stranger is dying in horrible pain, we should do something to help them also rings very true.
You mentioned at the beginning that the goal was an anthropological description and not a manifesto. Would three specific moral circles, a complete rejection of moral desert and a complete rejection of consideration by species be a belief that the vast majority of EAs would endorse? I think certainly as a movement it holds to those concepts much less strongly than the general public, but I'm not sure most reject them completely or even hold a consistent philosophical opinion there, unlike, say consequentialism which seems integral to the movement.
Or it could just be that I've gotten the wrong impression, given it's not something front and center in the marketing.
I'm not sure there's evidence for a difference between EA and non-EA people when it comes to marginalised groups here. (I get that you are unsure about this point, but I want to push back anyway). Note that the survey was asking participants to rank their "obligation" to show moral concern for each group, from 1 (absolutely no obligation) to 9 (very strong obligation).
When people say they feel a stronger obligation to care for a mentally challenged person than for a random citizen of their country, I assume they're reasoning is something like "mentally challenged people have difficulty taking care of themselves, therefore there is a higher obligation for other people to step in and compensate for this". I suppose some people might see mentally challenged people as more "morally pure" or whatever, but for most it seems to be more about treating people as equally worthy, and trying to lift them up to the same level.
Similarly when people are (very slightly) preferring homosexuals to random citizens, I assume this is also coming from a place of "these people are discriminated against, so we have a slightly higher obligation towards them in order to compensate for this".
EA's absolutely do feel more obligated towards people in the third world compared to the first: it's reflected in their donations.
Yes, the reason for the strong mutual hostility between EAs and "the Woke" (SJs thereafter) isn't due to differences in evaluating circles of concern (both are ostensibly universalist) but how this concern is expressed. EAs tend to view maximizing welfare as the sole moral concern, which SJs view as paternalistic or even neo-colonialist. SJs tend to view emancipation of marginalized people from structural oppression as the main concern, which EAs view as an inefficient use of resources based on qualitative sociology rather than objective neoclassical economics.
I think we're kind of rationalizing after the fact. (This is, after all, an offshoot of rationalism. ;) )
A lot of SJ rhetoric and ideology relies on *seeing your own group as the outgroup* (at least if white/male/etc) in order to correct for historic injustices. There's also a lot of using prior Christian framing about sin and punishment and repentance. EAs don't do that, from what I can tell.
How come caring more about people we know is "human and not doing so would be neither possible nor desirable", but caring less about outgroup members is not? Ingroup bias is a human universal just like bias for kinship is. I think reducing ingroup bias is more tractable, but I'm not convinced that we should categorize it differently.
Indeed, from what I can see around the globe, caring less about outgroup members is very human. Most tribes' name for themselves means something like 'the people' or 'the real people'.
I'm just going to go post-post and disagree with each and every one of them, am i?
i will be surprised if most EAs are actually behave that way. the way i model it, EA are people who have in their moral coalition element that value all people. but i expect them to care about various communities too. the SciFi community, their country, their hobby community.
there are a lot of people that believe you should FIRSt take care of the closer circles, and only after there are no more poor (or undeserving poor) in your country, help others. while EA have budget for everyone and different budget for local charity.
for example, Maximum Impact Israel is EA. it's seem ridiculus to me say that it's not: https://maximpact.org.il/
i sort of feel the same way about the six circle. there is a function-of-caring that assign every person positive wight, including Hitler. but i both have function of incentives and TDT, that gibe Hitler negative value. and sure, i sum them and the sum may get negative sometime, but the first function still exist.
and it very Value Differences As Differently Crystallized Metaphysical Heuristics thing, in my opinion. i just... don't see that important difference between having an impulse to punish and calculate you should? but, this is longer and more complicated issue, and this is already long comment. so...
in my model. EAs are people who have EA element in their moral coalition, when most people don't. they are not that element. they just actively have it.
I expect MOST EAs to care about communities that you declare "fake" (in a obviously wrong way - if community have mutual aid, it's look pretty real to me). i expect them to donate or to let sleep on the couch to people from their community, even of they are stranger. i'm pretty sure i already encounter descriptions of that in EA forums.
i find the claim they don't weird and surprising. like, it look obviously wrong to me, so you must have see it too. maybe i misunderstood? do you predict that most EAs will not help a member of their hobbyist community they don't know?
I agree. But I do think that an EA is going to have a very different idea of what communities should be treated differently, at least if you go by their revealed preferences. I predict that an EA would generally show more concern to a fellow EA, what but they show far less difference in how they treat a fellow American and say a Mexican. But I definitely think that there is a substantial difference here between how normal people behave and how most EAs behave.
Similarly, I agree that most people don’t actually have clear-cut values and beliefs instead of simply having certain patterns of behaviour they tend to follow. Certainly, I don’t think most people would consider it a meaningful difference. Whether you want to punish someone for revenge or because you want to make an example of them for the purpose of deterrence, and I expect that even if there was such a difference, both types of people would answer a survey similarly. Also for all that people claim and think that they would not seek retribution. If it had no benefits I do think tis would be less common if situations where these two values were in conflict weren’t so uncommon. There is also I think a smaller difference in Outlook than might appear on first glance because at the end of the day figures like Hitler haven’t directly affected most EA. It’s easy for your commitment to a moral philosophy or a subconscious desire to signal your adherence to the philosophy to override the mild preference that most people have in favour of Hitler suffering, but when it comes to situations where ordinary people have a strong preference for suffering, for example, people who have severely wronged them and their family I expect an EA would behave like anybody else. If somebody murdered an EA’s family I think they would want revenge and my probability of this is only very slightly lower than it would be for a normal person and honestly higher than it would be for many normal people in favour of compassion for criminals. But I do expect that having a well developed moral outlook that generally does not view retribution as inherently good does effect decision making in the real world, especially when there are no counteracting forces at play like in the Hitler situation.
if i comparing EAs to my Fantasy and SciFi community, i don't expect EAs to prefer locals less. the trend toward choice-based communities in certain circles is not EA-specific. so it's difference beween some Western-middle class-liberals and most people, not difference between EAs and most people.
"I expect an EA would behave like anybody else." - i actually don't. we all have impulses we disendorse, and self control and ability to not act on those impulses. i am, nevertheless, skeptical of the claim that EAs doesn't have those impulses.
why your probability on EAs revenge higher then some other people?
are you talking about ziz? i have no idea who else it could be, but it make no sense to bring her - she is pro revenge.
so when we talk about whether EAs believe in value X somewhat more then the general population, bringing the one person who is outlier the other direction look weirdly irrelevant, eve if you see ziz as relevantly EA - and she is not, under the first two definitions i thought about.
i have no problem with "gauche", whatever this even mean. i have problem with unsound arguments.
Stuff like Ziz is ironically one of the big reasons I never got deeper into EA. (The other is not having a lot of people into it around me.) Seems like I'd adopt a 'weird' ideology and not even be a better person by most metrics--EA and rationalism make cults just like every other belief system, it seems.
i find all this way of thinking really strange. what do you even mean by "got deeper"?
i believe things because to the best of my knowledge they are true. i do things because i want to, because to the best of my estimates i will get better results, by my values, by doing so.
i can't choose what true or not (and unwilling to compromise on believing things that are not true).
i can choose just not do things that i believe are net-negative. but i would not call it avoid to go deep.
so what even you mean? what someone say something like that i feel like we talking different languages.
You said you wouldn't expect EAs to behave like anybody else if they were personally affected, because you think EAs have better self-control and ability to not act on impulses. I think it's good to notice that empirically EAs¹ can totally use REA memes like game theory to rationalize those impulses anyway.
what? no, this is not what i said! i said that *people*, generally, have self control. so i expect that people who disendorse some shard of desire will act on it less then people who have no opinion on it, and those will act on it less then people who actively endorse this impulse.
so believe that revenge is bad -> do less revenge.
also, by Ozy's model (but not mine) believe revenge is bad -> EA.
I think that if someone knows that certain people share (some of) his important values, then he might not regard them as complete strangers, even if he never met them personally. I can even imagine him having certain positive feelings toward them that he does not have toward his family members who don’t share his important values.
You don't have control over what you care about. It's all automatic and unconscious. It's just a delusion to think you can reason yourself into changing what you care about.
Is this an aspirational essay? I ask because many of your statements aren't true.
EA's show preference for one group of strangers over another all the time by the sheer physical fact that they are more familiar with one over another. For example everyone knows about the category 'starving Africans' but 'starving Mongolians' is somewhat obscure. Certainly one group gets more donations than the other.
Secondly:
"For example, punishing bad behavior can keep people from behaving badly in the future, and sometimes you need to lock people away where they can’t hurt anyone"
That's reason enough to suppress a population.
Nonhuman example: You suppress a population of mosquitoes because they spread disease in the future
Human example: You suppress a population of bandits because they predate on traveling caravans in the future
The flaw woven throughout the points of the essay is that you don't know enough about the people in outer circles compared to the outsized effect your aid would or would not have on them. Draining a pond to kill mosquitoes has both positive and negative effects that are easy to foresee if it's your personal pond but impossible if it's a pond on an excel sheet you got from a third circle category. Likewise giving aid to Sunni vs Shiite in Iraq is going to have outsized consequences to other people which people in, say America, are ill equipped to know or understand.
Mongolia is a middle-income country with a fairly robust welfare state. The reason you don't hear about "starving Mongolians" is because there aren't that many starving Mongolians.
If you look at a map of the world coded by GDP (PPP) per capita, you'll see that quite a lot of the world is doing okay. There are only a few countries in desperate straits, and most of them are in sub-Saharan Africa.
I think this analysis misses the concept of reciprocity, which lies at the foundation of ethics. I owe more to the children who are tortured to make carpets for export than to the children who are tortured to sweep chimneys in their home countries because we indirectly interact in a way that benefits us both. Positioning yourself as a disinterested distributor of largess makes ethics seem less like an imperative and more like a whimsical hobby to be replaced as fashion changes.
I count mussels and oysters as meat-plants, but not clams. Clams are capable of motility! My reasoning for this as the relevant boundary here is the educated guess that 'pain is evolutionarily expensive'. If an animal can't move, it probably doesn't have a sophisticated pain response. If an animal *can* move then I'm inclined to give it the benefit of the doubt. Still, their neurons number in the thousands so this isn't exactly an issue of serious concern to me, there's only so much pain that can be going on in a system that simple.
Bacteria are capable of motility; would you assign them more weight than a mussel or a tree?
Less than a mussel (surely if nothing else, bacteria would also live within and on mussels), unsure about trees and not very interested in the question. I think on the 'pain is evolutionarily costly' theory it probably would be very costly for something as small as a bacterium and it could probably accomplish something similar without pain? Like, motility is necessary but not sufficient for establishing that a creature can feel pain in my model
I think the caring about the individual vs. groups difference is a pretty fundamental one.
Many political causes are not saying "I care more about people in this category", but rather "I care about this category as its own object". This is why you get people so concerned about "the great replacement" and genocide, while much less concerned about individual murder of larger numbers of people in larger demographic groups.
And the same applies to non-humans. A typical environmentalist finds the death and suffering of millions of pigeons basically irrelevant, but the extinction of pigeons would be a big deal. Thus we get these laws about how companies must protect the existence of *species*, but are free to harm as many *animals* as they want.
A big part of effective altruists' "weirdness" comes from not caring about preserving groups at all, and they focus only on the largest numbers of sufferers. These naturally tend to come from the most common species, which normal people care the *least* about.
There's also the whole aspect that group-thinking is pretty normal (which doesn't mean it's right necessarily). We evolved to get into groups and fight wars with other groups (just look around the world), so the effective altruist perspective (which tends to be the extreme form of the universalist 'there is no outgroup') is really weird to most people.
It might be tricky to interpret the results of this study, because people could be conflating the welfare of the thing with the thing's impact on aggregate welfare.
For example, it's very strange that the welfare of trees and mountains is given relatively high standing, because trees and mountains do not have welfare. However, trees and mountains do affect aggregate welfare.
Similarly, we might say people gave negative consideration to the welfare of murderers and child molesters because they want them to suffer (plausible), or perhaps they thought that the existence of murderers and child molesters has a negative impact on aggregate welfare.
There's an adjacent point in the comments about how people could be conflating the intrinsic importance of a person's welfare with the obligation to care for the person's welfare.
I do want to recognize that people really do value the intrinsic welfare of different groups differently (and even strangely), but some of the results only make sense on this alternate framing.
I agree. I also think this survey, as well as the moral circle expansion concept in general, unhelpfully fail to distinguish between *absolute moral worth* and *relative moral obligation*. Most people believe that animals are *absolutely* less morally valuable than humans, but that they have lesser *relative* obligations to foreigners than to their fellow-citizens, in the same way that they have special obligations to their family and friends.
I think I qualify as a "normie" EA. I identify with the movement, donate to EA charities, read some blogs, but I'm not otherwise involved in the subculture at all.
Your first few EA posts seemed dead-on to me, but this one doesn't feel like it's describing a belief system that I ascribe to, or that I've gotten the impression is the strong consensus belief of the community. I think moral circle expansionism is certainly a core EA belief, and your final point, if a stranger is dying in horrible pain, we should do something to help them also rings very true.
You mentioned at the beginning that the goal was an anthropological description and not a manifesto. Would three specific moral circles, a complete rejection of moral desert and a complete rejection of consideration by species be a belief that the vast majority of EAs would endorse? I think certainly as a movement it holds to those concepts much less strongly than the general public, but I'm not sure most reject them completely or even hold a consistent philosophical opinion there, unlike, say consequentialism which seems integral to the movement.
Or it could just be that I've gotten the wrong impression, given it's not something front and center in the marketing.
I'm not sure there's evidence for a difference between EA and non-EA people when it comes to marginalised groups here. (I get that you are unsure about this point, but I want to push back anyway). Note that the survey was asking participants to rank their "obligation" to show moral concern for each group, from 1 (absolutely no obligation) to 9 (very strong obligation).
When people say they feel a stronger obligation to care for a mentally challenged person than for a random citizen of their country, I assume they're reasoning is something like "mentally challenged people have difficulty taking care of themselves, therefore there is a higher obligation for other people to step in and compensate for this". I suppose some people might see mentally challenged people as more "morally pure" or whatever, but for most it seems to be more about treating people as equally worthy, and trying to lift them up to the same level.
Similarly when people are (very slightly) preferring homosexuals to random citizens, I assume this is also coming from a place of "these people are discriminated against, so we have a slightly higher obligation towards them in order to compensate for this".
EA's absolutely do feel more obligated towards people in the third world compared to the first: it's reflected in their donations.
Yes, the reason for the strong mutual hostility between EAs and "the Woke" (SJs thereafter) isn't due to differences in evaluating circles of concern (both are ostensibly universalist) but how this concern is expressed. EAs tend to view maximizing welfare as the sole moral concern, which SJs view as paternalistic or even neo-colonialist. SJs tend to view emancipation of marginalized people from structural oppression as the main concern, which EAs view as an inefficient use of resources based on qualitative sociology rather than objective neoclassical economics.
I think we're kind of rationalizing after the fact. (This is, after all, an offshoot of rationalism. ;) )
A lot of SJ rhetoric and ideology relies on *seeing your own group as the outgroup* (at least if white/male/etc) in order to correct for historic injustices. There's also a lot of using prior Christian framing about sin and punishment and repentance. EAs don't do that, from what I can tell.
This has the bizarre presupposition that social justice activists aren't generally themselves women, black, queer, etc.
How come caring more about people we know is "human and not doing so would be neither possible nor desirable", but caring less about outgroup members is not? Ingroup bias is a human universal just like bias for kinship is. I think reducing ingroup bias is more tractable, but I'm not convinced that we should categorize it differently.
Indeed, from what I can see around the globe, caring less about outgroup members is very human. Most tribes' name for themselves means something like 'the people' or 'the real people'.
I'm just going to go post-post and disagree with each and every one of them, am i?
i will be surprised if most EAs are actually behave that way. the way i model it, EA are people who have in their moral coalition element that value all people. but i expect them to care about various communities too. the SciFi community, their country, their hobby community.
there are a lot of people that believe you should FIRSt take care of the closer circles, and only after there are no more poor (or undeserving poor) in your country, help others. while EA have budget for everyone and different budget for local charity.
for example, Maximum Impact Israel is EA. it's seem ridiculus to me say that it's not: https://maximpact.org.il/
it's the place when you can donate to GiveWell in israel and get it count for taxes. it's also have both worldwide recommended Charities: https://maximpact.org.il/effective-global-orgs/
and local ones.
i sort of feel the same way about the six circle. there is a function-of-caring that assign every person positive wight, including Hitler. but i both have function of incentives and TDT, that gibe Hitler negative value. and sure, i sum them and the sum may get negative sometime, but the first function still exist.
https://slatestarcodex.com/2018/07/24/value-differences-as-differently-crystallized-metaphysical-heuristics/
and it very Value Differences As Differently Crystallized Metaphysical Heuristics thing, in my opinion. i just... don't see that important difference between having an impulse to punish and calculate you should? but, this is longer and more complicated issue, and this is already long comment. so...
in my model. EAs are people who have EA element in their moral coalition, when most people don't. they are not that element. they just actively have it.
I expect MOST EAs to care about communities that you declare "fake" (in a obviously wrong way - if community have mutual aid, it's look pretty real to me). i expect them to donate or to let sleep on the couch to people from their community, even of they are stranger. i'm pretty sure i already encounter descriptions of that in EA forums.
i find the claim they don't weird and surprising. like, it look obviously wrong to me, so you must have see it too. maybe i misunderstood? do you predict that most EAs will not help a member of their hobbyist community they don't know?
I agree. But I do think that an EA is going to have a very different idea of what communities should be treated differently, at least if you go by their revealed preferences. I predict that an EA would generally show more concern to a fellow EA, what but they show far less difference in how they treat a fellow American and say a Mexican. But I definitely think that there is a substantial difference here between how normal people behave and how most EAs behave.
Similarly, I agree that most people don’t actually have clear-cut values and beliefs instead of simply having certain patterns of behaviour they tend to follow. Certainly, I don’t think most people would consider it a meaningful difference. Whether you want to punish someone for revenge or because you want to make an example of them for the purpose of deterrence, and I expect that even if there was such a difference, both types of people would answer a survey similarly. Also for all that people claim and think that they would not seek retribution. If it had no benefits I do think tis would be less common if situations where these two values were in conflict weren’t so uncommon. There is also I think a smaller difference in Outlook than might appear on first glance because at the end of the day figures like Hitler haven’t directly affected most EA. It’s easy for your commitment to a moral philosophy or a subconscious desire to signal your adherence to the philosophy to override the mild preference that most people have in favour of Hitler suffering, but when it comes to situations where ordinary people have a strong preference for suffering, for example, people who have severely wronged them and their family I expect an EA would behave like anybody else. If somebody murdered an EA’s family I think they would want revenge and my probability of this is only very slightly lower than it would be for a normal person and honestly higher than it would be for many normal people in favour of compassion for criminals. But I do expect that having a well developed moral outlook that generally does not view retribution as inherently good does effect decision making in the real world, especially when there are no counteracting forces at play like in the Hitler situation.
if i comparing EAs to my Fantasy and SciFi community, i don't expect EAs to prefer locals less. the trend toward choice-based communities in certain circles is not EA-specific. so it's difference beween some Western-middle class-liberals and most people, not difference between EAs and most people.
"I expect an EA would behave like anybody else." - i actually don't. we all have impulses we disendorse, and self control and ability to not act on those impulses. i am, nevertheless, skeptical of the claim that EAs doesn't have those impulses.
why your probability on EAs revenge higher then some other people?
(It might be a little gauche to bring this up here, but a bunch of EAs have been arrested this year for membership in a revenge-obsessed phyg.)
are you talking about ziz? i have no idea who else it could be, but it make no sense to bring her - she is pro revenge.
so when we talk about whether EAs believe in value X somewhat more then the general population, bringing the one person who is outlier the other direction look weirdly irrelevant, eve if you see ziz as relevantly EA - and she is not, under the first two definitions i thought about.
i have no problem with "gauche", whatever this even mean. i have problem with unsound arguments.
Stuff like Ziz is ironically one of the big reasons I never got deeper into EA. (The other is not having a lot of people into it around me.) Seems like I'd adopt a 'weird' ideology and not even be a better person by most metrics--EA and rationalism make cults just like every other belief system, it seems.
i find all this way of thinking really strange. what do you even mean by "got deeper"?
i believe things because to the best of my knowledge they are true. i do things because i want to, because to the best of my estimates i will get better results, by my values, by doing so.
i can't choose what true or not (and unwilling to compromise on believing things that are not true).
i can choose just not do things that i believe are net-negative. but i would not call it avoid to go deep.
so what even you mean? what someone say something like that i feel like we talking different languages.
You said you wouldn't expect EAs to behave like anybody else if they were personally affected, because you think EAs have better self-control and ability to not act on impulses. I think it's good to notice that empirically EAs¹ can totally use REA memes like game theory to rationalize those impulses anyway.
¹: if you want to nitpick over whether they were EAs: https://archive.is/WQZiC#selection-1490.0-1490.1, https://archive.is/h0WAQ
what? no, this is not what i said! i said that *people*, generally, have self control. so i expect that people who disendorse some shard of desire will act on it less then people who have no opinion on it, and those will act on it less then people who actively endorse this impulse.
so believe that revenge is bad -> do less revenge.
also, by Ozy's model (but not mine) believe revenge is bad -> EA.
I think that if someone knows that certain people share (some of) his important values, then he might not regard them as complete strangers, even if he never met them personally. I can even imagine him having certain positive feelings toward them that he does not have toward his family members who don’t share his important values.
You don't have control over what you care about. It's all automatic and unconscious. It's just a delusion to think you can reason yourself into changing what you care about.
Is this an aspirational essay? I ask because many of your statements aren't true.
EA's show preference for one group of strangers over another all the time by the sheer physical fact that they are more familiar with one over another. For example everyone knows about the category 'starving Africans' but 'starving Mongolians' is somewhat obscure. Certainly one group gets more donations than the other.
Secondly:
"For example, punishing bad behavior can keep people from behaving badly in the future, and sometimes you need to lock people away where they can’t hurt anyone"
That's reason enough to suppress a population.
Nonhuman example: You suppress a population of mosquitoes because they spread disease in the future
Human example: You suppress a population of bandits because they predate on traveling caravans in the future
The flaw woven throughout the points of the essay is that you don't know enough about the people in outer circles compared to the outsized effect your aid would or would not have on them. Draining a pond to kill mosquitoes has both positive and negative effects that are easy to foresee if it's your personal pond but impossible if it's a pond on an excel sheet you got from a third circle category. Likewise giving aid to Sunni vs Shiite in Iraq is going to have outsized consequences to other people which people in, say America, are ill equipped to know or understand.
Mongolia is a middle-income country with a fairly robust welfare state. The reason you don't hear about "starving Mongolians" is because there aren't that many starving Mongolians.
If you look at a map of the world coded by GDP (PPP) per capita, you'll see that quite a lot of the world is doing okay. There are only a few countries in desperate straits, and most of them are in sub-Saharan Africa.
I think this analysis misses the concept of reciprocity, which lies at the foundation of ethics. I owe more to the children who are tortured to make carpets for export than to the children who are tortured to sweep chimneys in their home countries because we indirectly interact in a way that benefits us both. Positioning yourself as a disinterested distributor of largess makes ethics seem less like an imperative and more like a whimsical hobby to be replaced as fashion changes.