Quentin Anciaux wrote:
2010/1/11 Brent Meeker <[email protected]
<mailto:[email protected]>>
Stathis Papaioannou wrote:
2010/1/11 Brent Meeker <[email protected]
<mailto:[email protected]>>:
But aren't you assuming that consciousness is produced by
the abstract
Platonic computation - rather than by the actual physical
process (which is
not the same) - in other words assuming the thing being
argued?
No, I'm at this point assuming only that consciousness is
produced by
the physical process. We can assume for simplicity that the two
machines M1 and M2 have similar architecture and similar operating
systems. Once the program is loaded into M2 from the disc, S2
proceeds
exactly the same as it would have had the computation been
allowed to
continue running on M1. Therefore, at least after the first few
milliseconds, the subjective content of S2 must be the same as it
would have been on the one machine. Could the subjective
content be
different at the transition between S1 and S2 if the
computation is
split up? If there is a subjective difference it won't be
something
the subject can notice because, later in the course of S2, he
can have
no memory of it.
But if you're only assuming that consciousness is produced by the
physical process then the process of downloading and uploading the
microstates and shifting the data into registers in the CPU and
memory could produce a difference in consciousness. These are all
computations too, done by the operating system. And why can't
there be memory of it in the sense that it effects some later
conscious state? There are traces of the transfer process left on
the original computer, the disc, and the second computer. Some
subsequent program could retrieve these traces, as is done in
forensic cases. If physical processes instantiate consciousness,
why shouldn't these make a difference.
Because those states are not part of the "computation" you sliced on
the two computers.
They are not part of the abstract Platonic computation - but they are
part of the physical computation. So the question is, on which does
consciousness depend? My point is not to argue against Bruno's theory,
but only to point out that saying "yes" to the doctor may not be the
same as betting that consciousness=(Platonic) computation. If the
doctor proposed to replace your brain with an abstract computation you'd
probably say "no". If he proposed to replace you, your brain, and the
whole world with which you will ever interact, i.e. a virtual you in a
virtual world, would you say "yes"? You'd probably wonder how he was
going to compute that whole world with which you will interact?
And also assuming computationalism... Any implementation that does the
job... effectively does the job. That means while it's true there are
additionnal steps in the two case computer... it's just another
*valid* implementation of the same computation on one computer,
assuming computationalism that change *nothing*, arguing otherwise is
denying computationalism (maybe it's right and computationalism is false).
It also can't be a difference that would disrupt the
completion of a task or thought that requires continuity of
consciousness spanning S1-S2, since again the subject cannot
have any
evidence that such a disruption occurred.
Unless we have a theory of how consciousness is related to the
physical computation I don't think we can conclude that. We
already know that subliminal perceptions can affect conscious
thoughts - so why not subliminal memories.
We don't, but what Bruno is showing is the consequences *if* we are
turing emulable.
Bruno gives himself the luxury of considering turing emulablity at
arbitrarily low levels, including emulating your whole world. In fact
he sidesteps the doctors problem above, by simply emulating all
(arithmetically) possible worlds.
But this thread started with my questioning the idea of discrete
computational states, which are inherent in a turing emulation, as being
"thoughts" or "observer moments" and such moments having no order except
something inherent based on their content. I thing that thoughts have
duration in time and in computational steps and therefore can overlap
with other thoughts and this can provide an ordering not dependent on
the content of single computational states. I find this more convincing
because it doesn't rely on memories which in general are not part of
consciousness.
As I said I'm interested in what it takes to make a conscious AI. In
terms of pure computational capacity I expect that producing human level
consciousness may be within the capacity of the fastest computers within
the next 50yrs. But what to have them compute? I don't think a UD is
the way to go and even if it were it would give us no insight into
consciousness.
Brent
If we are turing emulable, all your above objections are not valid
because your objections are a level way too high (they are completely
valid objections at the level you describe, but assuming comp, those
are *still* computed at a lower level and hence are *part* of a
computation that generate consciousness, see the generalized brain
argument of Bruno).
Quentin
Brent
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to
[email protected]
<mailto:[email protected]>.
To unsubscribe from this group, send email to
[email protected]
<mailto:everything-list%[email protected]>.
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
--
All those moments will be lost in time, like tears in rain.
------------------------------------------------------------------------
--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en.
