This still feels like a “we need fifty Stalins” critique.
For me the biggest problems with the effective altruism movement are:
1: Most people aren’t utilitarians.
2: Maximizing QALY’s isn’t even the correct course of action under utilitarianism—its short sighted and silly. Which is worse under utilitarianism: Louis Pasteur dying in his childhood or 100,000 children in a third world country dying? I would argue that the death of Louis Pasteur is a far greater tragedy since his contributions to human knowledge have saved a lot more than 100,000 lives and have advanced society in other ways. But a QALY approach does not capture this. That’s extreme obviously, but my issue is that all lives are not equal. People in developed countries matter way more than people in developing countries in terms of advancing technology and society in general.
How fungible is Louis Pasteur? If he had died as a child, someone else would have done the same work, just perhaps a little later. How many lives would have been lost as a result of this delay? I don’t have a hard answer to this, but I have trouble putting the estimate as high as 100k.
How predictable is Louis Pasteur? Looking at his Wikipedia article, if we look at him as a child, we don’t predict he makes the contributions he does. Let’s say there’s a 0.1% chance of that happening. On the other hand, suppose there’s a child dying in the third world who we could bring to the first world for the cost of Louis Pasteur not dying in his childhood who has a 1% chance of making the same contributions. Clearly, losing the latter child is, on average, a greater tragedy than Louis Pasteur.
It’s reasonable to invest heavily in fewer people who can therefore make Pasteur-like contributions, rather than lightly in more people who won’t. Unless I’m mistaken, this is essentially what CFAR is doing. However, bell curves tell us that there’s more extraordinary people in the developing countries who could matter way more than people in developed countries, but only if we get them into developed countries where we can tap their potential. For every child in America who, given a standard education, has a 1% chance of making Pasteur-like contributions, there’s three in India, and, if we can identify them cheaply, it’s much more cost effective to move their chances of success from epsilon to 0.01 than the developed child’s chances from 0.01 to 0.02.
Truly rare talent is not fungible. Without Grigori Perelman, mathematicians would have struggled for a very long time to crack the Poincare Conjecture, even though all of the required tools were already there. The same is true in physics and possibly chemistry (Linus Pauling comes to mind). Maybe biology but I’m not sure. Its possible that Pasteur was fungible, but there is another issue I didn’t mention: the effects of losing great minds isn’t linear. Losing 100 scientists is worse than 100 times as bad as losing one (more on this later).
“However, bell curves tell us that there’s more extraordinary people in the developing countries who could matter way more than people in developed countries, but only if we get them into developed countries where we can tap their potential”
First of all countries differ in their average IQ so the math does not work that way. Also extraordinary students are already able to become scientists if they want in developed countries—universities take students from all over the world. Finally this isn’t what the effective altruist movement is focused on. A QALY based approach would not have us identify the brightest children in developing countries and bring them to developed countries. A full scholarship costs what $100,000 minimum? Clearly a QALY based approach would demand that we instead use that money to save several hundred lives.
Edit: I suspect that I may have come across as suggesting that we divert the effort from EA into saving potential Louis Pasteurs. That was not my intention: I was using an extreme example to show why QALYs (or pretty much anything that amounts to save as many lives as possible) are a poor metric—when you save the lives of a group of people: you have to consider what those people are going to do and how they are going to change society.
I don’t know what the most worthwhile thing to do is: I’m not that arrogant. But I don’t think that public health interventions in very poor countries are the most worthwhile things.
Maximizing DALY’s isn’t even the correct course of action under utilitarianism
That’s an understatement! DALYs are defined as intrinsically bad: one DALY is the loss of one year of healthy life relative to a reference lifespan, or equivalent morbidity. QALYs are the good ones that you want to increase.
I thought it was clear from the “Over-reliance on a small set of tools” section that I am strongly against relying on DALYs or similar metrics. Although I disagree with the framing of the solution being to weight different people differently. I’d prefer to move beyond the “maximize weighted sum of happiness” framing entirely (although still retain it as one of many reasoning tools).
This still feels like a “we need fifty Stalins” critique.
For me the biggest problems with the effective altruism movement are:
1: Most people aren’t utilitarians.
2: Maximizing QALY’s isn’t even the correct course of action under utilitarianism—its short sighted and silly. Which is worse under utilitarianism: Louis Pasteur dying in his childhood or 100,000 children in a third world country dying? I would argue that the death of Louis Pasteur is a far greater tragedy since his contributions to human knowledge have saved a lot more than 100,000 lives and have advanced society in other ways. But a QALY approach does not capture this. That’s extreme obviously, but my issue is that all lives are not equal. People in developed countries matter way more than people in developing countries in terms of advancing technology and society in general.
How fungible is Louis Pasteur? If he had died as a child, someone else would have done the same work, just perhaps a little later. How many lives would have been lost as a result of this delay? I don’t have a hard answer to this, but I have trouble putting the estimate as high as 100k.
How predictable is Louis Pasteur? Looking at his Wikipedia article, if we look at him as a child, we don’t predict he makes the contributions he does. Let’s say there’s a 0.1% chance of that happening. On the other hand, suppose there’s a child dying in the third world who we could bring to the first world for the cost of Louis Pasteur not dying in his childhood who has a 1% chance of making the same contributions. Clearly, losing the latter child is, on average, a greater tragedy than Louis Pasteur.
It’s reasonable to invest heavily in fewer people who can therefore make Pasteur-like contributions, rather than lightly in more people who won’t. Unless I’m mistaken, this is essentially what CFAR is doing. However, bell curves tell us that there’s more extraordinary people in the developing countries who could matter way more than people in developed countries, but only if we get them into developed countries where we can tap their potential. For every child in America who, given a standard education, has a 1% chance of making Pasteur-like contributions, there’s three in India, and, if we can identify them cheaply, it’s much more cost effective to move their chances of success from epsilon to 0.01 than the developed child’s chances from 0.01 to 0.02.
Truly rare talent is not fungible. Without Grigori Perelman, mathematicians would have struggled for a very long time to crack the Poincare Conjecture, even though all of the required tools were already there. The same is true in physics and possibly chemistry (Linus Pauling comes to mind). Maybe biology but I’m not sure. Its possible that Pasteur was fungible, but there is another issue I didn’t mention: the effects of losing great minds isn’t linear. Losing 100 scientists is worse than 100 times as bad as losing one (more on this later).
“However, bell curves tell us that there’s more extraordinary people in the developing countries who could matter way more than people in developed countries, but only if we get them into developed countries where we can tap their potential”
First of all countries differ in their average IQ so the math does not work that way. Also extraordinary students are already able to become scientists if they want in developed countries—universities take students from all over the world. Finally this isn’t what the effective altruist movement is focused on. A QALY based approach would not have us identify the brightest children in developing countries and bring them to developed countries. A full scholarship costs what $100,000 minimum? Clearly a QALY based approach would demand that we instead use that money to save several hundred lives.
Edit: I suspect that I may have come across as suggesting that we divert the effort from EA into saving potential Louis Pasteurs. That was not my intention: I was using an extreme example to show why QALYs (or pretty much anything that amounts to save as many lives as possible) are a poor metric—when you save the lives of a group of people: you have to consider what those people are going to do and how they are going to change society.
I don’t know what the most worthwhile thing to do is: I’m not that arrogant. But I don’t think that public health interventions in very poor countries are the most worthwhile things.
That’s an understatement! DALYs are defined as intrinsically bad: one DALY is the loss of one year of healthy life relative to a reference lifespan, or equivalent morbidity. QALYs are the good ones that you want to increase.
Edited
I thought it was clear from the “Over-reliance on a small set of tools” section that I am strongly against relying on DALYs or similar metrics. Although I disagree with the framing of the solution being to weight different people differently. I’d prefer to move beyond the “maximize weighted sum of happiness” framing entirely (although still retain it as one of many reasoning tools).
Good point sorry, what criteria would you use?