Eliezer_Yudkowsky comments on Many Worlds, One Best Guess - Less Wrong

12 Post author: Eliezer_Yudkowsky 11 May 2008 08:32AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (75)

Sort By: Old

You are viewing a single comment's thread.

Comment author: Eliezer_Yudkowsky 12 May 2008 07:56:13AM 2 points [-]

What flows is not time, but causality. As you guessed, I shall expand on that later. I think Barbour's time capsules reflect his lack of cog-sci-phil background - a static disk drive should never contain any observers; something has to be processed. You cannot identify observer-moments with individual configurations, which seems to be what Barbour is trying to do.

From the perspective outside time, nothing changes, but things are nonetheless determined by their causal ancestors. This is what makes the notion of "local causality" or Markov neighborhoods meaningful. This flow of determination is what supports computation, which is what supports the existence of observers. This means that no observer is ever embedded in a single configuration; only a determination of future configurations' amplitude by past configurations' amplitude, can support computation and consciousness.

Which I consider as common sense. Timelessness, also, adds up to normality; there's still a future, there's still a past, and there's still a causal relation between throwing a rock and breaking a window. None of that goes away when you take a standpoint outside time.