3 ms·

I think that this is a bait and switch here. I'm not aware of any significant and serious criticism of the "give lots of money to malaria net organizations" ap

I think that this is a bait and switch here. I'm not aware of any significant and serious criticism of the "give lots of money to malaria net organizations" approach to Effective Altruism. This also doesn't seem to be especially present in the article. I'm sure that you can find individual examples on twitter or whatever, but from my seat that seems to be a clear minority of the criticism. It is the version of Effective Altruism that has gone beyond that to "we need to focus on AI safety" and "we need to fly to mars" that people are calling bananas. Heck, the article even mentions that it is difficult to separate Effective Altruism from AI Doomerism and lists a bunch of vague accomplishments about AI Alignment as accomplishments alongside "Saved 200,000 lives." The argument that people should efficiently allocate their charitable giving is sound. Where things go off the rails is when people's model of efficiency involves wild future hypotheticals about robot uprisings.

Arnt 3y ago
AIUI the EA argument would be "focusing on EA safety is among the most effective usee of our resources" and "flying to Mars is among the most effective uses of our resources". I can see how the former might be viewed as a cost-effective way to improve the world, even if I personally haven't even tried to consider the matter. The latter, not so much. Can you point to an effective altruist who argues that that's likely effective? I don't mean someone who says effective altrusts think so (and are stupid to think so), I mean someone who is an effective altruist and says so.
- UncleMeat 3y ago
  Right that's the argument. I just think that people who say that are using a wildly incorrect function for altruistic efficiency.
  - yorwba 3y ago
    And who are those people who say that?
chaseadam17 3y ago
Exactly. Nonprofits have a principal agent problem — they are funded by donors, not the people they serve. As a result, they are incentivized to find ways to give a non financial ROI to donors in order to keep the money coming (eg galas, name a building, focus on the donor’s pet project, etc) instead of serving the interests of the people they are supposed to help. This dynamic — which is similar to politicians serving special interests rather than constituents — leads to lots of inefficient philanthropy. EA tries to solve this with - better spreadsheets! Surely we can just math our way out of this. This is a step in the right direction, but it doesn’t solve the root problem that donors, not beneficiaries, allocate capital and they are not in a great position to determine what is “effective” because it’s highly subjective. So even though I think EA largely means well, these same problems of donor motives creep back into giving, despite better spreadsheets and a bunch of phds. EA gets funding from tech VCs and all of a sudden Mars and AI are the central topics to giving. If hedge funds were the main donors, we’d probably be back focusing on microfinance. And in a sense, the false confidence and superiority that EA lends to donors, even though it does help on the margins creates more cover for the rare nefarious donors or nonprofit that use it to hide sinister ulterior motives. The whole thing is similar to communism not working and the people in charge think, “we need better spreadsheets to allocate resources!” Rather than, “maybe we should give some of our power back to the people we’re ostensibly trying to help so they can pick what best serves them.” The single thing missing from 99% of philanthropy conversations after all these years is STILL the voices of the people we’re all trying to help. We don’t need more phds or business tycoons deciding what’s best for the poorest people in the world. We just need to give them a little more money and power. But deep down, most of us don’t want to do that. We want to help in our way, on our terms, while controlling the purse strings and taking the credit.
- UncleMeat 3y ago
  Your last paragraph is so huge. There's a ton of "well we know the trick" stuff coming from donors and donor organizations that lead to just obviously broken approaches. One Child One Laptop is a great example. This is one reason why I like direct financial gifts. People receiving the money know what they need the most.
- henryaj 3y ago
henryaj 3y ago
> It is the version of Effective Altruism that has gone beyond that to "we need to focus on AI safety" and "we need to fly to mars" that people are calling bananas. There's been more of a move towards longtermism (and existential risk) in the last couple of years, but this is still not the majority of what the Effective Altruism community is about. I also don't think flying to Mars has ever been on their radar.

Arnt 3y ago

AIUI the EA argument would be "focusing on EA safety is among the most effective usee of our resources" and "flying to Mars is among the most effective uses of our resources". I can see how the former might be viewed as a cost-effective way to improve the world, even if I personally haven't even tried to consider the matter. The latter, not so much. Can you point to an effective altruist who argues that that's likely effective? I don't mean someone who says effective altrusts think so (and are stupid to think so), I mean someone who is an effective altruist and says so.

UncleMeat 3y ago
Right that's the argument. I just think that people who say that are using a wildly incorrect function for altruistic efficiency.
- yorwba 3y ago
  And who are those people who say that?

chaseadam17 3y ago

Exactly. Nonprofits have a principal agent problem — they are funded by donors, not the people they serve. As a result, they are incentivized to find ways to give a non financial ROI to donors in order to keep the money coming (eg galas, name a building, focus on the donor’s pet project, etc) instead of serving the interests of the people they are supposed to help. This dynamic — which is similar to politicians serving special interests rather than constituents — leads to lots of inefficient philanthropy. EA tries to solve this with - better spreadsheets! Surely we can just math our way out of this. This is a step in the right direction, but it doesn’t solve the root problem that donors, not beneficiaries, allocate capital and they are not in a great position to determine what is “effective” because it’s highly subjective. So even though I think EA largely means well, these same problems of donor motives creep back into giving, despite better spreadsheets and a bunch of phds. EA gets funding from tech VCs and all of a sudden Mars and AI are the central topics to giving. If hedge funds were the main donors, we’d probably be back focusing on microfinance. And in a sense, the false confidence and superiority that EA lends to donors, even though it does help on the margins creates more cover for the rare nefarious donors or nonprofit that use it to hide sinister ulterior motives. The whole thing is similar to communism not working and the people in charge think, “we need better spreadsheets to allocate resources!” Rather than, “maybe we should give some of our power back to the people we’re ostensibly trying to help so they can pick what best serves them.” The single thing missing from 99% of philanthropy conversations after all these years is STILL the voices of the people we’re all trying to help. We don’t need more phds or business tycoons deciding what’s best for the poorest people in the world. We just need to give them a little more money and power. But deep down, most of us don’t want to do that. We want to help in our way, on our terms, while controlling the purse strings and taking the credit.

UncleMeat 3y ago
Your last paragraph is so huge. There's a ton of "well we know the trick" stuff coming from donors and donor organizations that lead to just obviously broken approaches. One Child One Laptop is a great example. This is one reason why I like direct financial gifts. People receiving the money know what they need the most.
henryaj 3y ago

henryaj 3y ago

> It is the version of Effective Altruism that has gone beyond that to "we need to focus on AI safety" and "we need to fly to mars" that people are calling bananas. There's been more of a move towards longtermism (and existential risk) in the last couple of years, but this is still not the majority of what the Effective Altruism community is about. I also don't think flying to Mars has ever been on their radar.