If this was the case, I would expect places with less harsh penalties, or with lower probabilities of being convicted, to have a significantly higher homicide rate (all other things being equal). Does anyone have statistics about that?
I don’t get it… He wouldn’t gain anything by killing you, so why would you be afraid he would do that? (Also, I’m not sure the assumptions of traditional game theory apply to humans.)
Have you seen The Dark Knight? This is exactly the situation with the two boats. (Not going into spoiler-y detail.) Causal decision theory demands that you kill the other person as quickly and safely (to you) as possible, just as it demands that you always defect on the one-shot (or known-iteration) Prisoner’s Dilemma.
Anyway, I think you shouldn’t end up murdering each other even in that case, and if everyone were timeless decision theorists (and this was mutual knowledge) they wouldn’t. But among humans? Plenty of them would.
I haven’t read it yet, but I think this is basically the thesis of Steven Pinker’s The Better Angels of our Nature.
Have you seen The Dark Knight? This is exactly the situation with the two boats. (Not going into spoiler-y detail.) Causal decision theory demands that you kill the other person as quickly and safely (to you) as possible, just as it demands that you always defect on the one-shot (or known-iteration) Prisoner’s Dilemma.
Anyway, I think you shouldn’t end up murdering each other even in that case, and if everyone were timeless decision theorists (and this was mutual knowledge) they wouldn’t. But among humans? Plenty of them would.