Kory Heath wrote:
>
> On Nov 24, 2008, at 3:28 AM, Bruno Marchal wrote:
>> MGA 1 shows that MEC+MAT implies lucky Alice is conscious (during the
>> exam). OK?
>> MGA 2 shows that MEC+MAT implies Alice is dreaming (and thus
>> conscious)
>> when the film is projected. OK?
>
> I don't mean to hold up the show, but I'm still stuck here. I don't
> understand how Lucky Alice should be viewed as conscious in the
> context of MEC+MAT.
>
> In a different message, you said this:
>
>> But to go in the
>> detail here would confront us with the not simple task of defining
>> more precisely what is a computation, or what we will count has two
>> identical computations in the deployment.
>
> As complex as that task may be, I'm beginning to think that I can't
> get past MGA 1 without tackling it.
>
> Imagine that you have a grid of bits, and at each tick of the clock,
> each bit is randomly turned on or off using a pseudorandom number
> generator with a very long periodicity. Imagine that for some stretch
> of time, the bits in the grid act "as if" they were following the
> rules to Conway's Life. Are Conway's Life computations in fact being
> performed? I thought "obviously no". The majority answer here seems to
> be "obviously yes".
>
> Suppose that we perform a very complex computation, and the result is
> the integer "5". Should any computation that results in "5" be viewed
> as performing the former computation?
>
> Chalmer's paper "Does a Rock Implement Every Finite-State Automaton?"
> seems directly relevant to all of these Lucky Alice thought
> experiments. (Is it?) I need to re-read that paper.
>
> I have no doubt that my thinking on these topics is confused. Where
> should I begin?
>
> -- Kory
I share your reservations, Kory. In outline, Burno's argument so far seems to
be (I'm sure Bruno will correct me if I get this wrong):
1. Assume that consciousness supervenes on the material realization of some
complex computations.
2. These computations could be performed stepwise by some machine that only does
arithmetic and consciousness would still supervene.
3. The order of the steps matter, but not the time interval between steps. So
even if the steps are discrete and separated in time consciousness will still
supervene.
4. Since many different mechanisms can realize the sequence of steps the
consciousness must supervene on the computation however the sequence is realized.
5. The sequence of steps could be realized by accident, i.e. a random number
generator.
6. The sequence of steps could be realized by a recording of the original,
conscious sequence.
7. 5 & 6 supra are absurd (i.e. "false") therefore there is an implicit
contradiction in 1.
But I don't find this compelling. First, 5 & 6 are not contradictions - they
just violate our intuitions about what consciousness should be like. But what
is it about them that violate our intuitions?
(a) They have divorced consciousness from it's context, i.e. it's potential or
actual interaction with an environment.
(b) They eliminate the temporal continuity, so that the consciousness is sliced
into discrete "observer moments" which are regarded as states in a state machine.
(c) They eliminate causal connections within the process that is supposed to
realize consciousness.
The causal connections are broken by imagining coincidents that are so
improbable that their probablity of happening within the lifetime of the
universe is infinitesimal - in other words at a level where we have no way to
distinguish improbable from impossible.
Having shown there is something counter-intuitive implicit in 1 thru 7 supra,
we're invited to conclude that consciousness supervenes on pure, abstract
computation which takes place in an arithmetical Platonia. But that also
violates a lot of intuitions. Of course I'm not against violating intuitions,
but I expect some predictive power in exchange.
Brent
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list.domain.name.hidden
To unsubscribe from this group, send email to everything-list+unsubscribe.domain.name.hidden
For more options, visit this group at
http://groups.google.com/group/everything-list?hl=en
-~----------~----~----~----~------~----~------~--~---
Received on Mon Nov 24 2008 - 20:19:42 PST