Hacker News Reader

You've used a lot of value-laden words here (dodgy, magical) and made a lot of meta-claims about how bad this kind of thinking is, but I don't think you've shown any actual fallacy in the person's thinking.

These folks are saying they see an opportunity to save a life at very low risk to themselves. I don't see the specific problem with this reasoning. Yes, the donation is not guaranteed to save a life and everyone still dies, but these are not compelling objections from a starting point of 1 in 4,000.

> These folks are saying they see an opportunity to save a life at very low risk to themselves.

No, that's not what they 're saying. One says that "not making the donation would have meant he valued his life at 4,000 times that of a stranger" and the other is saying that "1/3,000 risk of death in surgery is like sacrificing yourself to save 3,000 people".

"They see an opportunity to save a life at very low risk to themselves" is your interpretation of what they are actually saying. It might be true, but I'm discussing what was actually said. Do you want to discuss what was actually said, or your interpretation of it?

These claims are based on what would happen if many people donated kidneys and the statistics held true. If 4,000 people donated kidneys to people who needed them to survive, on average, one person would die and 4,000 lives would be saved.

From an expected value perspective, one person donating a kidney is like 1/4000th of this.

If you base moral decisions on expected value, as effective altruists try to do, this calculus is straightforward.

The calculation is "not making the donation would have meant he valued his life at 4,000 times that of a stranger". That doesn't follow from the "expected value perspective".

The "4000" is swapped, slight-of-hand like, between unequal, orthogonal values. One is the number of people who donate a kidney, the other is the value of a human life. It's counting 4000 apples and then saying "I got 4000 oranges". It's the old switcheroo, and it means nothing.

Btw, you say "From an expected value perspective, one person donating a kidney is like 1/4000th of this."

"Of this" what? What is "one person donating a kidney" 1/4000'th of? I am asking because I honestly have no idea what you mean. Sorry if my tone sounds combative, but in this I'm just confused.

I think my post is pretty clear. There are four sentences. There is only one sentence the "this" can refer to, and it's the one that comes right before. Could you read it again?

The statements sound clear, and seems to pretty much match our intuition about how it should work. But perhaps incomplete in the math? Since you are not alone in saving 4000 people, but that 4000 people are saving 3999 lives. Or that you are saving 3999/4000 of a person, or roughly one person per donation. So the expected value of the profit of a donation is approximately 1 person saved. But the expected value of the cost of that is that 1 of 4000 people is personally sacrificed. Then, taking the ratio of those, we get that the expected value for social-value over the personal-cost is about 4000-to-1 (approximate, since this assumes that the life lost is of equal value to the average value of the lives saved). Or that society values the donation being about 4000x the value of your life.

(your own personal utility function might disagree however)

A more precise model still might be to try to factor in the expected loss in healthy years for the donor, and the average lifespan gain of a kidney recipient.

If you pull a child out of the way of a speeding truck, you've saved their life without losing any quality of life for yourself or for them. That's a "perfect" life save. Organ transplant is more subtle.

If you already plan to donate your kidneys at death, donating one of them when younger seems less obviously good. It'd seem worthwhile to build a more detailed model before going through with it.

Edit: I need to learn to summarisy my thoughts better. What I'm saying below and above is that the expected value of live kidney donation to society is misapplied to calculate the value of live kidney donation to the donor. We tacitly accept as a society that some people will die so that others may get a new kidney. That has nothing to do with the value that one person donating a kidney places over another person's health (and not life, because you're not "saving" lives).

Although I understand your comment much better than the OP -4000 to 1, not 1/4000- those maths don't make any sense at all to me. First of all, nobody saves anyone by donating a kidney. You certainly don't "save" an entire person that way. You add some quality-adjusted years of life to their life expectancy, but that's all. And they're not that many, average is about 10. So at best you can say that each person donating a kidney is "helping" or "supporting" one other person.

So it makes no sense to say that 3999/4000th of a person is "saved" or "helped" or whatever by one's _personal_ sacrifice. If I was the person dying while donating my kidney, my death would not equal 4000 people living, or living longer. It would mean my life ending for one person's life lasting a little while longer.

One person! That's how many people are directly affected by one person's donation of a kidney. To claim that it's 4000 instead, is just sophistry with fractions. Or rather, it's an oversimplification because understanding the trade-off between one entire life and 10-ish quality adjusted years of life is too hard and does not easily support self-aggrandising statements.

Look at it the other way: you won't let 4000 people die by refusing to donate a kidney. But this is more or less what is claimed in the two opinions above. Everydobdy who does not donate a kidney is an asshole who values their own life more than that of 4000 strangers.

No, your comment is not clear at all. I read it again. Here's what it reads like if I replace "this" with the sentence that comes before it:

From an expected value perspective, one person donating a kidney is like 1/4000th of [if 4,000 people donated kidneys to people who needed them to survive, on average, one person would die and 4,000 lives would be saved].

Is there any chance you may want to try and help me become uncofused?

Btw, I'm arguing that the EAs' maths is wrong and you, and others here, are arguing that it's sound.

So I have to ask you, and the rest who think the EAs' maths is sound: have you, or are you planning to donate your kidney to a stranger?

Because either the EAs' maths is sound, or you donate a kidney to a stranger, or otherwise you think your life is worth more than 4000 or 3000 or however many other people.

And sorry to be so lawyery about this, but I detect an undercurrent in this thread of "shut up, the maths have spoken" (see for example cjbprime's comment just below: I'm confused and the maths is sound). Well, if the maths is sound then go do the right thing, or else quit arguing just for the sake of argument.

Because the two people up there, who justified their decision to live-donate a kidney, at least they followed their principles, however misguided, and nobody can fault them for that. And I think their thinking is wrong and I only have one kidney anyway. But who are you, and how do you live up to your morals?

There are several options here:

One could think the math is correct, but conclude that they don't value other people that much. That doesn't mean the math is wrong. It means the assumptions behind the math ("I don't value myself more than 3000 other people") aren't held by them right now.

Or one could think the math is correct, and be generally inclined towards kidney donation as a result, but think it would be better at a different time, such as when they don't have very young kids at home to care for, or when they have a job which is less demanding and would tolerate them taking a month away from work at once.

Or one could believe in the general statistical inference, but think that the specific 3000 number is too high and insufficiently studied, and the actual risk to self is much higher. I don't believe this myself, but it sounds like you might. This is not a criticism of the method, just of the specific number being plugged in to it.

You can argue against these philosophical positions and assumptions without having to retreat to questioning what numbers mean.

No, that doesn't work as you say and it's not a matter of accepting philosophical positions.

That's because the EAs are not making a philosophical argument but a mathematical argument. And if you think their maths are right then you must agree with their agument, otherwise you must think the maths are wrong.

What the EAs' maths are trying to do is to define an objective measure of morality. That's the point of attempting to quantify the value of a human life and that's the purpose of using maths in general: because maths is objectively true or false, while morality is otherwise not. So if they get the maths right, their conclusions must apply to everyone and anyone, regardless of other assumptions.

That is the appeal that EA has to quantitatively-trained and generally mathematically-minded types. Let's do away with the subjectivity of moral philosophy and calculate the truth about morality. We face a question of morality? Calculemus!

So if you think their maths are right you must accept their conclusions, and you must donate one of your kidneys or accept that you are acting immorally. It doesn't matter when you choose to do it or how moral you think it is, what matters is that you accept it is the moral thing to do.

You can't have your cake and eat it: either the maths are wrong, or refusing to be a live kidney donor is wrong.

I disagree with everything you wrote, and I doubt you can find a prominent EA who doesn't.

The mathematics are there to benefit people who share moral assumptions like "the extreme suffering of other humans is bad" or "I don't value myself more than 3000 other people". There is no objective morality. But there are many people who share moral assumptions like the above ones, and the mathematical calculations are for their benefit.

EAs are actually incredibly thorough about writing down all of their subjective moral weights and comparing them -- GiveWell's staff has done this and published the result for many years, for example. They've created spreadsheets where you can plug in your own moral weights to see how it affects their giving suggestions. The fact that morality is necessarily subjective and individual is an extremely normal part of the EA conversation.

That's just splitting hairs. The weights don't matter. What matters is that the EAs claim that their maths measure the morality of actions.

It doesn't matter if they disagree over the parameters, what matters is that they agree their maths quantify morality. That is the objectivity that they claim.

And they even have spreadsheets to do it, huh? Wow. But, what are these spreadsheets calculating then? I mean, how can you calculate something subjective? If a quantity is subjective, then why can't I calculate it any way I like? If I can calculate a quantity any way I like, then does that quantity really measure anything? Can I use E = mc² to calculate the morality of my actions? If not, why not?

That stuff just doesn't make any sense, sorry.

> I disagree with everything you wrote, and I doubt you can find a prominent EA who doesn't.

"Prominent", huh? Interesting hedging there. Why should I care that someone is "prominent"? Don't peoples' opinions count if they're not "prominent"? And what's "prominent" anyway? Like, X followers on Instagram?

You know, the more I'm having this conversation, the more it sounds to me like some weird kind of Silicon Valley roleplaying that's just out of touch with reality.

In case it helps to hear it from someone else: you are confused and the statistics are sound and not a trick.

Good heavens, of course I wouldn't presume to question the wisdom of statistics!

I'm just saying that the way the statistics are used doesn't make sense.

To play EA's advocate: if you assume that donating the kidney will "save" a life, then if you had 4,000 kidneys, you could, on average, sacrifice yourself to "save" 4,000 lives. Of course you don't have 4,000 kidneys, but the intuition is that the decision is the same whether you're sacrificing 1/4,000th of your life to save one person, or your whole life to save 4,000.

However, I agree with your point that this isn't really "saving" a life as much as "extending" or "delaying the inevitable". In terms of QALYs (Quality-Adjusted Life Years), the benefit may be much smaller while the risk remains the same. And I especially agree with your conclusion in the last sentence.

Actually the assumption is that you have an endless quantity of kidneys and the expected number of donated kidneys until you die is 4000.

That alone should help to see why it's all a completely absurd idea.

> To play EA's advocate: if you assume that donating the kidney will "save" a life, then if you had 4,000 kidneys, you could, on average, sacrifice yourself to "save" 4,000 lives.

No, see, this is what I mean when I say that maths are being misapplied. We know how many kidneys you can donate before you die: two. Or three if you have three. Some people even have four. So what does the "4000" have to do with that? The "4000" is the number of different people who each donated one kidney, so it doesn't tell us anything about one person donating 4000 kidneys.

And anyway, if you need to imagine a person with 4000 kidneys before you can figure out the value of live kidney donation, then you don't have a good way to figure out the value of live kidney donation. Which is my point.

>These two numbers, "1" and "4000" have no obvious relation to the value one places on one's life compared to the lives of others.

If you think donating a kidney has a 1:4000 chance of killing you, and a 1:1 chance of saving someone else's life, then I do think deciding not to do it is saying that you value your continued existence 4000x someone else's. We can argue about the specific numbers, and about whether this is the best way to apply your life to helping others, but I don't think the interpretation is wrong.

> For example, "4000" is not the number of others lives saved by donating one kidney.

I don't think see anyone claiming that?

> By donating one kidney one can "save" one other person's life at most

Often you're able to save more than one: it is reasonably common for altruistic donors to be able to facilitate donation chains that would not otherwise happen. The situation is that there are many people who are willing to give a kidney to help a specific person they know, but who are not compatible donors. One way that can work out is that I give my kidney to your friend, and you give your kidney to my friend: no altruistic donor needed. But often this doesn't quite work out in terms of compatibility, and having an altruistic donor can make multiple transplants happen that wouldn't otherwise.

>it's not really "saving" as it is delaying the inevitable

That's right. Especially considering that even post transplant the recipient is going to be on immunosuppressive drugs and is not likely to have the same duration or quality of life as the typical person. Often when talking about this, people use the concept of a QALY: how many quality-adjusted life years is it?

> Equally dodgy is the calculation of "1/3,000 risk of death in surgery is like sacrificing yourself to save 3,000 people" earlier in the article.

Why is this calculation dodgy?

(While I strongly admire people who have donated kidneys, I don't think it is a very good trade-off of cost to yourself versus benefit to others, and personally I still have both kidneys)

> "4000" is not the number of others lives saved by donating one kidney.

Why did you leave out the "For example" preceding that sentence in my comment? Leaving "For example" out completely changes the meaning of my comment.

It doesn't look like you're interested in a honest conversation.

> (While I strongly admire people who have donated kidneys, I don't think it is a very good trade-off of cost to yourself versus benefit to others, and personally I still have both kidneys)

I don't, and I suggest you keep both of yours because you are going to need them.

> Why did you leave out the "For example" preceding that sentence in my comment? Leaving "For example" out completely changes the meaning of my comment.

I interpreted your use of "for example" as indicating that what followed was a specific claim that backed up the more general claim in the previous sentence? I didn't include it in my quote, because I was responding to your two claims separately. While it's still not clear to me how you see it changing the meaning of your comment, I've edited mine now to include it.

I think it comes from the statistical expectation. Like, if you were able to donate your kidney any amount of times, you would on average expect to lose your life on the 3,000th donation. So the question is, do you want to be the kind of person who would do that at the cost of your own life?