The Grad Student and the Dropout

[Jeff Bone]

On Tuesday, Apr 29, 2003, at 15:10 US/Central, Russell Turpin wrote:

> Alas, that's [demon-haunted world] sort of the implication of EPW.

I adamantly disagree with this conclusion.  You can take off and dwell 
on that possibility, if you like, and make it the centerpiece of the 
whole thing, a conveniently wacky strawman to take swings at.  But I 
think that's missing the point.  Even in a world where everything 
possible happens an arbitrarily large number of times, some things are 
more probable than others.  And that says nothing at all about *why* 
those things happen, how those state declarations came about, etc.

We're going to start talking about Cantor Numbers any minute now, I can 
just feel it. ;-)

> Think of all the possible worlds where our
> observable universe is embedded as a part,
> i.e., all the possible worlds where this
> discussion is occurring on FoRK. There is
> exactly one where the ultimate physics of
> the universe is precisely what our physicists
> determine.

No, that would be true for all the possible histories where our physics 
determines outcome, existing simultaneously and each comprising a 
different instance of the class "our kind of universe," among an 
arbitrarily large number of other classes of universes, each itself an 
instance of the metaclass "universe."

> In all the infinitely myriad
> others, this world is a simulation (or
> equivalently, a created reality) with some
> other, more complex, unimaginable ultimate
> physics behind it, and this simulation is set
> in motion by some variety of gods, demons,
> norns, transhuman descendants, alien races
> beyond strange, or superintelligent computers.

There's absolutely no implication of this whatsoever.  This is merely 
wild extrapolation from the unfortunate idea that, for "simulation" 
implies some simulator.  If you'll notice, I've suggested nothing in 
the way of hypotheses for how the bits got there "in the first place."  
The notion of "the first place" may be entirely broken anyway.

It's just bits all the way down.

I sense a corner looming in your path...

> So in some sense, the hard-headed empiricist
> who shrugs off the possibility of any realm
> beyond the observable

Look, I already admitted that I don't have any particular experiments 
in mind for this.  MWQM-probing experiments are a good start...  and I 
go back to the idea that there's no essential difference between MWQM 
and EPW, except for an arbitrary and a priori trimming you're doing of 
the MWQM phase space to make it fit the model of what you think QM is 
saying.  At worst, MWQM is EPW with certain constraints, but we're 
going to disagree on whether or not those constraints are what you 
think they are for various practictioners' interpretations of QM.

Elaboration:

I have three "slices" of phase space I want to consider.  Each of these 
slices represents my current context at some instant.  We'll call one 
of these S0, one S1, and one S2.  S1 and S2 represent possible futures 
of S0, paths through the phase space denoted S0-->S1 and S0-->S2 
respectively.  S0 and S1 differ (locally) by only a few quanta;  these 
deviations follow what might be expected;  S0-->S1 occurs as a subset 
of transitions between arbitrary T0 and Tx for some large number of 
paths T0-->Tx.  S2 deviates from S0 by a very large number of 
Plank-scale state differences;  in S2 Bugs Bunny appears beside me and 
will at some slice Sx, x>2 flatten me into a pancake with a cast-iron 
skillet w/o killing me.  S2 and its implied entailed states are not 
prohibited by the rules of phase space state transitions nor are they 
*necessarily* excluded by the laws of what makes a valid "slice" of 
phase space.  S0-->S2 is less "real" than S0-->S1 because it occurs 
less frequently when considering arbitrary paths T0-->Tx.

> So which will it be, Jeff? The demon-haunted
> world spawned by the singularity? Or the much
> more meagre world of the hard-headed
> empiricist?

Look, you've left the playing field here, Russell.  Does a Singularity 
itself spawn a demon-haunted world, to the extent that all such 
demon-haunted worlds can be simulated?  Maybe so.  But does that rule 
out local empiricism?  Not at all.  Empiricism lets you discover the 
nature of your probability-clustered neighborhood in the phase space, 
and --- maybe --- even learn things about the macro nature of the phase 
space at large.  To suggest that the simulation hypothesis and EPW 
devalue empiricism in any way, that it's an either / or proposition, is 
rhetorically irresponsible.  It's a false dichotomy and a convenient 
straw man wrapped up in one.

jb