It has been claimed on this site that the fundamental question of rationality is "What do you believe, and why do you believe it?".
A good question it is, but I claim there is another of equal importance. I ask you, Less Wrong...
What are you doing?
And why are you doing it?
That's very interesting. It sounds like you start digging into the problems with level 4 multiverse ethics I plan to write a series of sci-fi novels about. What I have already written (not too much) can be read with Google Wave if you add "our-ascent-noofactory@googlegroups.com" to you contacts and then display the group waves. There are a couple of underlying concepts and questions which come together in my fictional work:
Why do I care about the hypothetical consequences of modal realism? Because I think it's the best ontology for modelling the world as correctly as possible, and I'm pretty convinced that it's true (for reasons similar to those in the map that is the territory, but more founded in mathematical logic and philosophy of mathematics). Trying to apply "pure" utilitarianist reasoning to a modal realism multiverse leads to serious problems, for example:
A) The amount of joy and suffering are actually infinite, which destroys the point of summing it up or integrating it. You could fix the problem by doing "local" computations, but then how do you define the "locations" in the right way, if worlds are nested or infinite in space or time. All of this is a huge headache for a convinced hedonistic utilitarianist, which I used to be before. (I find it hard to tell what I am, in an ethical sense, now. Possibly, "confused" might be the best shortest description.)
B) Every configuration of sentience actually is realized somewhere in the modal realism omnicosmos, which I call Apeiron. From a purely abstract point of view the mental states with positive and negative valence should be in one-to-one correspondence, which means that from a purely apeironal (maximally holistic) point of view, there is a perfect balance of good and bad feelings in every respect. Interestingly, this observation seconds that theories like hedonic utilitarianism are only meaningful if applied "locally".
C) Above of each (computable) world lies an infinite (!) chain of worlds in which the first one is simulated directly or indirectly (unless that is impossible, which I think is not the case). If you haven't considered the problems of simulation ethics yet, this is a good reason for starting to do so.
D) Trying to define a probability measure over anything on a whole level 4 multiverse is rather hopeless. Maybe it's possible to define some fancy something measures, but ultimately you have to face the probem of unbounded infinite cardinality.
Oh, I don't know how so solve these problems in the most convenient way. All those questions and thoughts have left me with some kind of meta-ethical nihilism. However, I tried to invent some new meta-ethical concepts like "(meta-)ethical synergism" (quantify ethical systems and use ensembles of those systems for making moral judgements) and "thelemanomics" (extract the underlying economic, social, political and ethical systems from the volitions of all people; roughly comparable to CEV), which could fix some meta-ethical problems.
I think we should stay in contact.
I should mention that Metafire and I have spent lots of time discussing these concepts, and I am familiar with the ideas he is presenting in this comment. The metaphysics that Metafire is trying to describe here is pretty much identical to what ata described in the post on The mathematical universe: the map that is the territory.
Oh, and please excuse the confusing grammar, English isn't Metafire's native language. Metafire lives in Germany. Metafire, please check today's post in my waveblog for some suggestions on how to make the grammar of that comment... (read more)