Russell Standish wrote:
>
> >
> > Alastair,
> >
> > I will try to make my post more explicit.
> >
> > > Under AUH, unfortunately it is neither *necessary* for a SAS to perceive
> > > totally law-like behaviour, nor try to fit all observed behaviour to laws.
> > > Firstly, there will be a relatively few unlucky SAS's who *do* perceive
> > > dragon/WR events in any AUH (some will have more than can be explained away
> > > as hallucinations etc); secondly, it is the case that the majority of SAS's
> > > on this planet would ascribe at least some paranormal events that they see
> > > or think they see (like miracles, angels, or nde's) to divine, rather than
> > > law-based, explanations - this doesn't disqualify them from being SAS's.
I don't like the description of this problem as the "dragon" problem
or the "flying rabbit" problem. As someone (Fred, I think) posted
recently, a dragon or a flying rabbit would not indicate lawless
behavior, but rather, would indicate strange, complicated, different
laws at work.
I think that really what is being discussed (by all means speak up
if you disagree) is the question of why we experience any physical
laws at all. It seems to me that either we should expect the
universe to be lawful, or we should expect our senses to provide
pure white-noise static from this instant onward. I still can't
get past the nagging conviction that I got a few weeks ago that
none of the discussion I've read so far does anything to justify
the expectation of law over chaos.
Schmidhuber says that it has to do with the measure of the program
to compute this universe and nothing else. Well, that doesn't work,
IMO, because he doesn't define "this universe". Does he mean the
universal wave function of QM? If so, that obviously has an infinite
number of histories of me, so why isn't that multiple universes? If
not, then I would think that this universe (i.e. this history) has
actually a rather high information content. The info. content of
QM only plummets when you take the whole wave function into account,
nicht wahr?
> > > I agree that a minimal number of dragon events could in practice be
> > > dismissed as a mass-hallucination or something like that, but one cannot use
> > > this fact to *explain* the lack of established dragon events, not least
> > > because the main counter to the AUH is that there should be *maximal*
> > > law-less behaviour, consistent with the existence of a SAS (so perhaps
> > > events on Earth would be straightforward, but elsewhere (except for the
> > > life-supporting Sun) they would be chaotic). To refute this we need more
> > > careful analysis, such as that in Russell's paper, or my web site. (For MWI
> > > instead of AUH it's a whole different ball-game.)
> > >
> >
> > What I am driving at is the SAS will do its best to minimize its perception of
> > lawlessness in its universe, by doing the curve-fitting, so to speak. With
> > complete or too much lawlessness, it will cease to be aware of itself, or
> > practically of anything, because there is no stable environment and no
> > principles it can use to learn and predict. An awareness of self requires a
> > stable enough, predictable enough background provided by laws. Knowledge of
> > these laws is another matter.
>
> It is unclear that a SAS embedded into such an extremely complex - ie
> chaotic, or random universe would cease to be aware of itself. You
> will need to justify that statement. What is true, is that the SAS
> itself would have high information content relative to being embedded
> in a universe where the physical laws give a good likelihood of a SAS
> arising, and thus having much smaller measure than the lawlike
> universe.
>
> Now the White Rabbit aka Dragon universe problem is a shade of grey in
> between completely lawlike and lawless behaviour. Since these
> universes at least have some law abiding behaviour, they have rather
> large measure compared with the completely lawlike universe. The
> result of Alistair and my argument is to show that the vast bulk of
> these dragon universes are in fact indistinguishable from the
> completely law-like universe (according to the SAS - an omniscient
> being, if such a thing exists is a competely different kettle of
> fish), and the ones that are distinguishable will have relatively
> small measure compared to the indistinguishable ones.
>
> >
> > Rather than taking the dragon event to be an example of lawlessness, it is
> > probably more helpful to treat it as a very complex event, requiring complex
> > laws or complex corresponding algorithms. So, your justification that dragons
> > are very improbable is the same as justifying that (overly) complex universes
> > are very improbable.
> >
> > >
> > >
> > > I don't agree that the range of fine-tuning is subjective, given a
> > > reasonable physics framework. We may have problems assessing it, or there
> > > may be other types of SAS that we are not aware of, but that is a different
> > > matter.
> > >
> >
> > What I mean here is best posed as a question: what is the difference between the
> > number of possible values for the fine structure constant, alpha, in the
> > following interval:
> >
> > (1/137.0359895-10^-25,1/137.0359895+10^-25)
> >
> > assuming this interval is anthropically allowed, i.e., it allows humans to
> > evolve. (Note: I am using the central value 1/137.0359895 from Max Tegmark's
> > paper,"Is 'the theory of everything' merely the ultimate ensemble theory?"
> > published in Annals of Physics, vol. 270, pp.1-51 (1998).) The point is, you
> > already have an infinite number of universes corresponding to each real number
> > contained in this anthropically allowed interval. If you make this interval
> > larger or smaller while preserving the SAP, how much larger or smaller an
> > infinite set would be enough to say there is or is no fine-tuning? There may be
> > a question of precision in determining alpha, but then the question is really
> > where do you draw the line, why, and does it matter if you already have an
> > infinite number of universes? To me, it seems kind of subjective.
> >
>
> No fine tuning means that any value of \alpha is allowed (at least
> physically consistent values). So the above range you quote is
> extremely fine-tuned. In reality, the level of fine-tuning is likely
> to considerably less (I'm not sure what Tegmark quoted, but I thought
> the allowable range was a few percent of \alpha - which is still fine-tuned).
>
> > Fred
> >
> >
>
> ----------------------------------------------------------------------------
> Dr. Russell Standish Director
> High Performance Computing Support Unit,
> University of NSW Phone 9385 6967
> Sydney 2052 Fax 9385 6965
> Australia R.Standish.domain.name.hidden
> Room 2075, Red Centre http://parallel.hpc.unsw.edu.au/rks
> ----------------------------------------------------------------------------
--
Chris Maloney
http://www.chrismaloney.com
"Donuts are so sweet and tasty."
-- Homer Simpson
Received on Wed Dec 15 1999 - 19:58:29 PST