Ethics

Lee Daniel Crocker (lee@piclab.com)
Wed, 15 Jul 1998 13:33:50 -0700

DF> <sigh> As I've already said many times before, this is true
> when the game is iterated, which is most of the time. I'd say
> utilitarianism and egoism coincide at least 80% of the time,
> if not more. However, egoism leaves us worse off in that
> remaining 20% region, representing games where you don't know
> who your opponent is or in which one or more players will not
> play again.

  1. The "don't know who your opponent is" condition is irrelevant; consistent identity is /not/ required for egoism to produce good results in an iterated situation. It is not even strictly necessary for your opponent to know who you are, either, so long as each prospective opponent has /some/ information about you or a group to which you belong that might influence him; something like a reputation for honesty among men in a certain profession, for example, is enough to give an egoist positive value in maintaining that reputation in dealing with anonymous others.
  2. That "20%" speculation of yours is absurd in the real world. In my 35 years on the planet, I have yet to encounter a single situation where my interaction with someone else would not have consequences later in life. For me, then, the percentage of non-iterated prisoner's dilemma situations in real life is zero. I'd like to hear a convincing example of such a situation, but even if one or two could be contrived--just as I'm sure I could contrive an example or two of utilitarian atrocities--one should live by principles that reflect how life is in general, not by contrived examples. 
--
Lee Daniel Crocker <lee@piclab.com>