The rules of effective altruism. Rule #1: don’t die

post by William_MacAskill · 2013-06-24T04:00:29.000Z · score: 5 (4 votes) · EA · GW · Legacy · None comments

Effective altruists (EAs) should have a very different attitude to risks of death than egoists (whether or not the egoist is effective or ineffective).

Consider the decision to go hang gliding. Going hang gliding has 8 micromorts per trip — that is, if you go hang gliding, there’s a 1 in 125,000 chance that you’ll die. And, if you expect to live 50 more years, taking 1 in 125,000 means losing 210 minutes of expected life. So if you’re an egoist you might reasonably think: “Well, it sucks to lose 210 minutes of expected life, but it would be really fun to go hang gliding, I’d have great memories, and so on, so I’m willing to make the trade off” and then jump off a cliff. (Attached to a hang glider, of course.)

However, if you’re an EA, then it’s not just your life on the line. If you’re doing well, you should expect to save tens, hundreds, and maybe even thousands of lives over the course of your lifetime--or to do an equivalent amount of good. (I actually think that the amount of good you can do is far more than the value of saving hundreds of lives, but that requires argument and clarification, so I’ll leave it for another time). So if you die young, all those lives are lost.

Let’s suppose, then, that you expect to save 100 lives over the course of your future life. Given that, hang gliding doesn’t look so attractive. The amount of fun from gliding will stay the same, let’s suppose. But the cost is much greater. In fact, it’s more than a hundred times greater: by hang gliding you create a 1 in 125,000 chance of dying, and you create this chance for the 100 people you would save as well. This brings the total expected loss of life to 21,200 minutes, or two weeks of life. If thinking just in terms of self-interest, would you take a roughly one in 1,000 chance of death, losing two expected weeks of life, for the fun of hang gliding? Probably not. And if not, then you shouldn't take that benefit to yourself when it's other people's lives on the line either. So, to a very first approximation, as an EA you shouldn’t go hang gliding. In general, EAs should be much more concerned about their own survival than they might intuitively be inclined to be.

(NB: there’s much more to say on this issue, including: weighing risks of death that save you time; risks that make you more productive; and looking into which particular activities are too risky to be worth it, and which are ok. I’ll elaborate in further posts.)

None comments

Comments sorted by top scores.