Re: The GLUT and functionalism

From: Stathis Papaioannou (stathisp@gmail.com)
Date: Wed Mar 19 2008 - 00:39:37 MDT


On 19/03/2008, Lee Corbin <lcorbin@rawbw.com> wrote:

[To recap, we have a machine and operator in Australia and another
machine and operator in Vienna. The question is whether the Viennese
machine can continue the computation of the Australian machine, and
whether any associated consciousness also continues without
interruption. S(n) is the last state of the Australian machine which
the Australian operator notes and passes on, either explicitly or
hidden in false data, to his Viennese counterpart, who then endeavours
to calculate or look up the next state S(n+1) to input into his
machine.]

> I see. In other words, it's as though the Vienna operator, a
> very smart dude, works out *himself* with, say, his pocket
> calculator, S(n+1). Quite all right with me. I don't think it
> ever matters where a computation is done, nor on what
> hardware. Just so that it's a real computation, and so
> sustains real causality and information flow.

And you will hopefully allow that in this step, working out the next
state could involve a lookup table. This low level lookup simply
mimics the physics of the machine, and is not like looking up the
answer to a computation in order to bypass all the intermediate steps.

> > But if the Australian operator simply provides a list
> > of all possible states, knowing but not telling which one is S(n),
> > that seems to me exactly the same as providing no information at all;
>
>
> I agree.
>
>
> > i.e. the same as if the Vienna operator knew nothing about the
> > antipodean device and simply tried random states, one of them just
> > happening to be S(n+1). How does this affect whether consciousness is
> > interrupted?
>
>
> Very good question. In that case there is an infinitesimal interruption in
> consciousness, because there was no genuine information flow or
> causality, in other words, to me, not a real computation at all. All
> possible things are tried in Vienna. I think I see where you are going,
> but I'll let you deliver the next blow! :-)

Would it make a difference if the Viennese operator worked out the
appropriate state? Here are several possibilities:

(a) The file I've received says that the final state of the Australian
machine was S-6754. The successor state for that is S-2037, so if I
input that into my machine, the computation won't know it has been
spread across two continents.

(b) The file I've received gives the final state of the Australian
machine as either S-6754 or S-789. Their respective successor states
are S-2037 and S-9175, so I'll try both of those, and when I get the
right one, the computation and its consciousness will have been
implemented without interruption.

(c) The file from Australia lists all possible machine states S-1 to
S-9999. I have a lot of work ahead of me, but one of them has to be
the right one. First, I'll input S-1, the successor state of S-3412;
then S-2, which as everyone knows is the successor state of S-439;
next, S-3, successor of S-2031 (or was it S-3021? I always get those
two mixed up but it's definitely one or the other)... and so on to
S-9999.

You can see that the Vienna operator's muttering as he goes about his
work in the last example has no effect on the mechanics of the
computation. As he inputs S-2037 he could be thinking that it is the
successor of S-6754, or he could be thinking about what he will have
for dinner, and it will make no difference to his behaviour as he is
just inputting all the states anyway.

In which of the above cases is the causal link between the two
machines preserved? In which does consciousness continue
uninterrupted?

-- 
Stathis Papaioannou


This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:01:02 MDT