I’m new to the idea of EA, and interested in pursuing it further. I’m finding that I have a hard time buying the argument of using expected value to make a decision, when I don’t have a large number of decisions I get to make. Expected value works because it “averages out” over the long term. I’m especially concerned about the recommendations by GiveWell on deworming. I am not a rich person, and I want to be able to make an impact if I can. Nothing in life is a certainty. That said, putting a large portion of the $ that I can contribute to something that has a high probability of not succeeding means that there is a high probability that my contribution will make no difference whatsoever. If I put money into a large number of causes, each of which has that profile, and if they are independent of each other, I accept that expected value kicks in. But in the case where I only get to make a contribution to a very limited number of causes, I don’t see that expected value is the right philosophy. Can anyone help clarify?
Hi. Firstly, I want to say that many people within the movement differ in how risk averse they are and make decisions with that taken into account. For example, this flowchart on deciding which cause to work on http://globalprioritiesproject.org/2015/09/flowhart/ has a question “would you rather do something that has a 1% chance of saving 1,000 lives than save one life for sure?” I know some smart people who would answer that question with a “no” and behave accordingly.
However, many EAs think that the 1% chance option is better and many EAs spend entire life’s effort on causes like AI-safety even though it’ll almost surely have no impact because for them it’s not that important to make sure you have at least some impact, for them the small possibility of having a huge impact is just as motivating. I do share your feeling that having at least some impact is better, but personally I try to somewhat ignore it as a bias when making important decisions. To me, in some abstract sense, 1% chance of saving 1,000 lives is better than 100% chance of saving 1 life. In the same way that helping 10 people africa is in some abstract sense better than helping 1 person who lives in my country. And in the same way that helping 10 people who will live in million years is better than helping 1 person who is living now. Part of my brain disagrees but I choose to call that a bias rather than a part of my moral compass. Which IMO is a totally subjective choice.
Even if all of us were risk averse, it might still make sense for all of us to cooperate and put money into different risky causes, because then there’s a high probability that all of us combined will have a big positive impact. Instead of making sure that you yourself make a significant difference, you could think how EA community as whole (or humanity as a whole) could make a big positive difference. EA already supports many charities and maybe the risky charity that you donate to personally won’t have an impact, but if many people support different risky charities like you will, all of us combined will have a bigger impact with a high probability.
I’m new to the idea of EA, and interested in pursuing it further. I’m finding that I have a hard time buying the argument of using expected value to make a decision, when I don’t have a large number of decisions I get to make. Expected value works because it “averages out” over the long term. I’m especially concerned about the recommendations by GiveWell on deworming. I am not a rich person, and I want to be able to make an impact if I can. Nothing in life is a certainty. That said, putting a large portion of the $ that I can contribute to something that has a high probability of not succeeding means that there is a high probability that my contribution will make no difference whatsoever. If I put money into a large number of causes, each of which has that profile, and if they are independent of each other, I accept that expected value kicks in. But in the case where I only get to make a contribution to a very limited number of causes, I don’t see that expected value is the right philosophy. Can anyone help clarify?
Hi. Firstly, I want to say that many people within the movement differ in how risk averse they are and make decisions with that taken into account. For example, this flowchart on deciding which cause to work on http://globalprioritiesproject.org/2015/09/flowhart/ has a question “would you rather do something that has a 1% chance of saving 1,000 lives than save one life for sure?” I know some smart people who would answer that question with a “no” and behave accordingly.
However, many EAs think that the 1% chance option is better and many EAs spend entire life’s effort on causes like AI-safety even though it’ll almost surely have no impact because for them it’s not that important to make sure you have at least some impact, for them the small possibility of having a huge impact is just as motivating. I do share your feeling that having at least some impact is better, but personally I try to somewhat ignore it as a bias when making important decisions. To me, in some abstract sense, 1% chance of saving 1,000 lives is better than 100% chance of saving 1 life. In the same way that helping 10 people africa is in some abstract sense better than helping 1 person who lives in my country. And in the same way that helping 10 people who will live in million years is better than helping 1 person who is living now. Part of my brain disagrees but I choose to call that a bias rather than a part of my moral compass. Which IMO is a totally subjective choice.
Even if all of us were risk averse, it might still make sense for all of us to cooperate and put money into different risky causes, because then there’s a high probability that all of us combined will have a big positive impact. Instead of making sure that you yourself make a significant difference, you could think how EA community as whole (or humanity as a whole) could make a big positive difference. EA already supports many charities and maybe the risky charity that you donate to personally won’t have an impact, but if many people support different risky charities like you will, all of us combined will have a bigger impact with a high probability.
All that said, some EAs donate some money to causes that make sure that they have at least some impact and some money to risky causes with high expected value. More on this in https://www.lesswrong.com/posts/3p3CYauiX8oLjmwRF/purchase-fuzzies-and-utilons-separately