On Tue, Jan 14, 2003 at 02:45:15PM -0800, Hal Finney wrote: > Another angle on this argument takes an even broader view. Let us > consider all observer-moments in the multiverse. By eliminating those > observer-moments which have a negative quality of life, we improve ^^^^^^^^ > the average quality of all observer-moments and therefore we make the ^^^^^^^ > multiverse a better place to live. This is the same basic argument, but > applied to all observers rather than just the individual who was injured.
Wouldn't the same argument apply if you replace "negative" with "below average"? But if you (recursively) eliminate all below average observer-moments, you'll be left with only one observer-moment. Surely that isn't the best possible state of the multiverse. Perhaps you meant "total" instead of "average" above. I think that makes a bit more sense, but it still conflicts with the intuition that running the same computation multiple times is pointless. Tim accuses us of trying to "perfect" the multiverse. Sure, that's something to look forward to :), but right now I'm still trying to figure out what makes one multiverse better than another.