Bruno Marchal wrote:
I agree with George, but note that I arrive at an equivalent
assertion without using that "lower levels have lower complexity
and therefore higher measure". That is possible, but
the problem is that it is a priori hard to estimate the "dumbness"
of the universal dovetailer which is quite capable to entangle high
complexity programs with low complexity programs, so that
the "multiplication" related to low-complexity can be inherited to
high-complexity (due to dovetailing). But you may be right, I have not
proved that "a" UD could be that dumb!
Gosh, Bruno, I don't understand what you are saying. Maybe I am too naive! Or maybe our background conceptions are too different so even if the language is the same it does not make sense. For one I don't see how a first person experience needs to depend on a UD. My view is that the "observer-experience" simply consists in the (virtual) transitions from one "observer-moment" to another where the transition is filtered by having to be consistent with the "observer-state." Note how the observer bootstraps himself into consciousness out of the plenitude. So maybe my UD is the "nul UD" : it is the maximally dumb UD.
From a suggestion of Jacques Bailhache (an old everythinger) I have try to build an explicit UD which makes the measure on computations arbitrary, but I have not succeed, in the limit (on which bears the first points of view), the "right measure" seems to self-correct by itself. It is that self-measure I study with provability logic. Another problem with the idea of "low" level, or of "simple program" is that even a program with 2^2^2^2^2^2^2^2^2^2^2^2^2^2^2^2^2^64 as minimal bit-length is quite little in comparison of almost all number in Plato Heaven.
At 15:56 05/05/04 -0700, George Levy wrote:
This has been an interesting thread. Unfortunately I was too busy to contribute much. However, here is a thought regarding simulation versus first and third person points of view.
It does make sense to talk about a 3rd person point of view about simulation of a conscious entity on a computer. However, I don't think it applies to a first person point of view.
In the plenitude we'll have an infinite number of levels of simulation as well as an infinite number of simulations per level (2^aleph_0 as suggested by Bruno in a previous post, or higher)
From a first person point of view any observer moment in any simulation and at any level can transit to another observer moment in a different simulation at a different level provided the transition is consistent with the observer. Therefore from the first person point of view there is no such a thing as living in a simulator. As first persons we live in all simulators and at all levels.
In addition, since lower levels have lower complexity and therefore higher measure, the number of simulations is higher at lower levels.
Therefore we are more likely to occupy ensembles of simulations located at the lower levels. Is there a lowest level in the level hierarchy, that is a level below which there is no simulation, just the plenitude? Possibly. If so, we are most likely to exist "most of the time" at that base level, but we cannot exclude that "some of the time" we may be in a higher level.
hmmmm. This argument points to the fact that "most of the time" we do not live in a simulator!