hpm.domain.name.hidden wrote:
>"Jesse Mazer" <lasermazer.domain.name.hidden>:
> > [re: rock is a good implementation of any computation]
> > It depends what you mean by "good implementation." The context of my
> > comment above was, *if* you believe there is a single true set of
> > psychophysical laws, are the laws likely to be defined in terms of
> > "computation" or not? If you don't believe in these sorts of laws,
> > then the question of what's a "good" implementation is somewhat
> > academic--it's really just a question of what's the neatest way to
> > define our terms.
>
>Many of the people on this list (in common with a lot of western
>philosophy at least since Descartes) are hoping to construct their
>existence measures on the bedrock of the objectively decidable
>self-awareness. They've built very interesting structures, but you
>may notice there's been no progress at all on stabilizing the
>foundation. Instead we have on this list the same debates that
>endlessly, repetitively and inconclusively flood comp.ai.philosophy,
>never mind philosophy journals and books.
But my position isn't just "consciousness is real, therefore there must be a
mathematical theory to describe it so it'll have firm foundations." Rather,
I was drawn to the idea of a "theory of consciousness" based on
consideration of the anthropic principle, as well as "continuity of
consciousness" thought-experiments like the one where I'll be replicated
only if the coin falls heads. In both cases we're dealing with
probabilistic questions that seem to call out for a well-defined
mathematical theory...in the replicator experiment, do you think there's a
single true answer to the question of whether I should expect a 1/2 or 2/3
(or something else) chance that I will see the coin fall heads?
In any case, It might be useful to temporarily shift this debate away from
the question of "consciousness"--what I want to know is, do you believe in a
global "existence measure" at all? As I said, the main problem I have with
the TOE you've been advocating is that it seems as though it would be
impossible to have a single global measure of any kind. After all, you're
not just saying that any physical system can be interpreted as implementing
any computation--since you believe that the "physical" universe is itself
just a sort of Platonic computation, you're essentially saying that any
Platonic computation can be interpreted as any other Platonic computation,
which would seem to make it impossible to put a global measure on this set.
Without a global existence measure there's no way to talk about probability,
and no reason at all to say I'm any more likely to find myself in a universe
where the laws of physics will stay stable than one where they will suddenly
go haywire. So do you admit the possibility of *some* kind of global
existence measure on "everything?" If you do, then it's a separate argument
as to whether or not the measure should be on all possible conscious
observer-moments or on some other notion of "everything." Personally, I
think the measure needs to be on observer-moments to avoid having a
fundamentally dualistic theory in which the existence measure is completely
separate from anthropic considerations...but there wouldn't be much point in
discussing this if you didn't even believe in a global existence measure in
the first place.
>Gerard O'Neill, the late Princeton physicist best known for his space
>colony studies, once said that if you met a race that insisted that
>logical developments must be built step by step from a firm
>foundation, you could be pretty sure they were planet dwellers. Races
>that live in space realize that it's perfectly OK to build structures
>that have no foundation at all. They can be circular and unsupported,
>yet if you spin them they'll have gravity just like the ponderous
>planetary piles!
I don't think that's such a good analogy--space station builders have to
take into account the same "foundations" that planet-bound builders do,
namely the laws of physics. Anyway, anyone who believes in any kind of
global existence measure is making just as much use of foundations as I
am...it's just that I happen to think the measure should be on the set of
all possible observer-moments, while others have some different notion of
the appropriate set of "everything" to put a measure on.
>I think your insistence on the absolute underpinning of an objective
>consciousness is just planet-bound thinking. Bruno's, Juergen's,
>Russell's or Max Tegmark's analyses can just as well be built on
>arbitrary selections of what's conscious (Turing test passers?
>biological brains? red-haired people? teddy bears?). The teddy bear
>universes may have different probabilities than the biological brain
>universes or the Turing test universes, but so what? Each is as
>likely to be self-consistent as another.
True, once you have an existence measure you can then choose an arbitrary
definition of "observer" and do a weighted sum over all universes to see
where "observers" are most likely to be. The problem is that you will get
very different answers as to what's a "likely" universe depending on your
original definition. But the fact that I find myself in *this* particular
type of universe doesn't depend on my own "choice of reference class"--it's
thrust upon me by reality. If you want to explain that, I think you need
reality to have a single "true" definition of what it means to be an
observer.
With a weird definition of "observer" I could no doubt predict that I'm
exceedingly likely to see the laws of physics break down within the next
minute (perhaps by saying that those who don't observe such a thing don't
count as 'observers' at all). Is this definition as good as any other? If
so, why shouldn't I take this prediction as seriously as any other one?
Again, the problem is that whatever we think about the matter, reality seems
to have its own ideas, and a coherent TOE really should take this into
account.
>i.e. You don't have to give up the goals of this list just because you
>don't believe there is an objective fact of the matter to
>consciousness.
>
> > My question for you is, does your Platonic view of things rule out
> > the idea of a single global theory of consciousness?
>
>Don't need it, don't want it.
Is there any reason to predict that the laws-of-physics-going-crazy scenario
is less likely than anything else in your theory? If not I think you *do*
need to modify your theory in some way, whether or not that involves a
"theory of consciousness."
> > I believe that there is more to me than my outward actions.
>
>So does a lookup table. Just ask it, it will tell you all about
>its complex motives and internal mental life.
Well, I can build a program that just prints out "I have a lot of complex
feelings and thoughts that I'm not gonna tell you anything about, so nyeah
nyeah." And I suppose you'd say that under the right interpretation it does
have such thoughts--but even so, the fact that it *says* it has these kinds
of thoughts has zilch to do with anything, since according to you the same
would be true of a clock or any other system. That's basically the
situation we have with a lookup table as well...what the lookup table does
or doesn't say isn't relevant to your case here.
> > I'm making the assumption that there's *some* set of objective
> > psychophysical laws, and that these laws would tell you a lookup
> > table doesn't give rise to the same experiences as a detailed
> > simulation.
>
>Groundhog! I make no such assumption.
...Space cadet! Seriously, I think good theories depend on a balance
between "spacey" intuitive leaps and "grounded" critical standards...best
not to pit one against the other in the first place.
Jesse Mazer
_________________________________________________________________
Get your FREE download of MSN Explorer at
http://explorer.msn.com
Received on Thu Feb 15 2001 - 10:33:58 PST