Stathis Papaioannou wrote:
> Peter Jones writes:
>
> > > > Computer always have counterfactuals, because there changing
> > > > one part of them (whether data or programme) has an effect on
> > > > the overall behaviour. Changing one part of a recording (e.g splicing
> > > > a film) changes only *that* part.
> > >
> > > I don't think you can distinguish between recording and computation on that basis. By "recording"
> > > I don't mean just the film, but the film + projector as system. The film is the computer's fixed input
> > > and the computer is the projector in this case. A closer analogy would be a software media player
> > > playing an .mpg file: the output is rigidly fixed by the input, although the media player handles
> > > counterfactuals in that the output would be different if the input were different. But the same is
> > > true of any physical system sensitive to initial conditions.
> >
> > I don't see how that helps you argument. Such a "recording" is not just
> > a string
> > of unrelated states. There is something beyond the manifest, active
> > state that
> > explains how one state makes the transition to another. (Of course
> > I would say that "something" is matter/physical laws).
>
> (I'm assuming physicalism for the sake of argument here.)
> Remember, I am not restricting the term "recording" to the input data alone, but to data + program
> as a system, because there is always a super-program which includes the data as part of the program,
> or a super-machine which includes the data hardwired. For example, Winzip accepts .zip files as input
> and produces as output the original uncompressed file, but it is also possible to have a single self-expanding
> .exe file which is effectively a combination of Winzip + .zip file. Executing the .exe file puts the computer
> through much the same sequence of physical states as using Winzip with the appropriate .zip file, but the
> .exe file accepts no input and is completely deterministic. In other words, you can always make a small
> adjustment to a system by including data as program, so that the new system goes through the same
> physical events as the original one, but the if-then statements are just redundant code. I don't see how
> you can say that one system implements a computation (and is potentially conscious) but the other does not.
I don't: I say that both systms have counterfatuals.
Maudlin's sabotage doesn't remove all the counterfactuals. Any causally
linked sequence of evetns has countefactuals: if A had been different,
B would be different.
> > > > > The latter seems
> > > > > obvious to me from the fact that an entity experiences only one stream of consciousness at a
> > > > > time, regardless of how many actual (in the multiverse) or possible (in a single universe model,
> > > > > with or without true randomness) braches there are in which that entity is conscious.
> > > >
> > > > That doesn't follow. A counterfactual is a COUNTERfactual - -it is
> > > > something that could have happenned but didn't. There is no
> > > > reason why we should be conscious of in things
> > > > we coudl have done but didn't. (Unless counterfactuals
> > > > are itnerpreted as alternate worlds, but then they
> > > > are not really COUNTERfactuals -- they actually
> > > > did happen, buit "somewhere else").
> > >
> > > That's just the point I am making: there is no reason why we should be conscious of things we
> > > could have done but didn't,
> >
> > Well, there is if you start from the premisses that
> > 1) consciousness is a type of computation
> > 2) computations provide counterfactuals
> > 3) In an immaterial universe, counterfactuals are provided by "other
> > worlds".
>
> I accept (1), although I'm not completely certain about it.
>
> I think (2) is problematic, because by fixing the input you fix the output as surely as if you excised the
> if-then statements, and the computer goes through exactly the same sequence of physical states as if
> you had excised the if-then statements.
In a particular situation, it will always go through a certain sequence
anyway.
The question is whether changing one thing (an external input,
or hardcoded data, or part of the programme itself) will have knock-on
effects.
> Conversely, any physical system, however rigidly deterministic,
> could be seen as implementing if-then statements because *if* some part of the system had been
> different *then* by following the laws of physics some other part of the system would have been different.
Of course. The ablity of computers to implement the counterfactuals
isn't some maguic added by the software, it is inherited from the
physical causality of the hardware.
> If you look beyond the lines of code you will see that a computer loaded with software is really just a
> physical system that moves this way if you push it here, that way if you push it there, all the while following
> the laws of physics. How could it be otherwise?
Exactly. The questions is whether and how that can
be reproduced immaterially. If there is no pushing
and pulling in Platonia, there is no computation, and
no computational minds.
> > > and there is no reason I should notice anything strange had happened
> > > if all my copies in other multiverse branches suddenly drop dead. Even if as a matter of fact it can be
> > > shown that consciousness is always associated with the actual or potential implementation of
> > > counterfactuals, it does not follow that we are conscious *as a result* of this.
> >
> >
> > The claim of computationalism is that we are indeed conscious
> > as the result of running a program. If you are going
> > to reject compuationalism, what other route do you have
> > to a immaterial, Everythingist universe ?
>
> What I'm rejecting is the notion that we are conscious *as a result of the counterfactuals*, whether
> actually implemented in other worlds or potentially implemented in a single world.
Because it is false that computation requires counterfactuals,
or false that cosnciousness is computation ?
> A computer with fixed
> input is as rigidly deterministic as a physical system can be.
Indeterminism is a seperate issue form counterfactuallity.
> It may *look* like it has if-then statements,
> but these cannot be implemented.
They can if you change the internals.If you can hardcode
external data into the program, I can reach into the code
and fiddle with them.
> It is like saying that a billiard ball has if-then statements, because if
> it were struck differenly by another billiard ball, it would move differently.
If physical systems didn't have counterfactual behaviour, programmes
wouldn't
be able to. The only thing computers have that
physical systems don't is the intentionality of a programmer.
(But I still think it is false that any physical system implments any
computer).
(That being the case, the computationalist claim isn't trivialised,
because computationalists think minds are a partiuclar kind of
programme).
> You could probably build a
> computer out of billiard balls arranged on a huge table, with the "hardware" being the balls, the "program"
> being the arrangement of the balls, and the "input" being how you hit the balls with the cue. Ignoring the
> effects of chaos (something we try to avoid in real computers), this billiard ball computer will respond in a
> perfectly deterministic way, in that if the input is fixed, so is the output.
Yes, you probably could.
> The counterfactual behaviour is
> intrinsic at the most basic, most stupid, physical level, and is unavoidable in any physical system.
> > > > The claim that consciousness requires counterfactuals
> > > > stems from the argument that consciousness is
> > > > comptutation, and computation requires counterfactuals.
> > > >
> > > > It doesn't stem from an expeiential insight into counterfactual
> > > > situations.
> > >
> > > A practical computer requires counterfactuals in order to interact with its environment.
> >
> > A computer programme has counterfactuals because
> > in general it has if-then branches, and in gnereal
> > it doesn't execute them all. That is a quite
> > separate consideration from "interacting with the environment".
>
> If it doesn't interact with the environment the counterfactuals are never implemented, not even
> potentially.
It is potentially the case that if some part of the data+programme were
changed, it would have knock-on effects. That indicates
that the sequnece of events is causally linked,
and causal linkage -- even of completely fixed and deterministic kind
--
distinguishes the data+programme from a recording.
> > > The
> > > problem with this idea is that firstly *any* physical system interacting with its environment
> > > handles counterfactuals,
> >
> > Why is that a problem ? The claim is that programmes have
> > counterfactuals,
> > not that everything with a counterfactuals is a programme.
>
> Perhaps you did not say so explicitly, but my impression was that you think counterfactual behaviour is
> what makes computer programs special. There is some confusion as to the status of a "regular" program
> accepting input from the environment, compared to the same program with fixed input. I would call the latter
> a recording and I would say that it is just as conscious (or not) as the regular program on a specific run.
The latter is not arecording because its stages are causally linked.
Changing one stage, however hypothetically, would change others.
> > > and secondly there is no reason to assume that the handling of
> > > counterfactuals is somehow responsible for consciousness.
> >
> > The purported reason is computationalism. You culd
> > abandon it, but where does that leave you ?
> >
> > > A rigidly determined computation
> > > may be a "trivial" case of a computation but it does not mean it is not a computation
> >
> > A determined system can still have counterfactuals.
> >
> > > . A machine
> > > hardwired to compute the digits of pi, and nothing else, is still computing the digits of pi even though
> > > it isn't much use as a general purpose computer.
> >
> > And it would computer something else if
> > it were hardwired differently. It still fulfils
> > the "internal counterfactual criterion" -- if
> > one of its states were changed, the subsequent sequence of states
> > would change. Unlike movie.
>
> A movie would be different if the patterns on the film were different, or the projector's speed were
> different.
Changin one part would not have an effect on other parts.
> If you say that a movie does not implement a computation, it's not because it lacks counterfactual
> behaviour.
It is because there is no causal linkage between it stages.
We appeal to counterfactuallity, in a broad sense,. to
establish causality.
> > >Similarly, we can imagine beings who are still
> > > conscious even though their lives are rigidly determined.
> >
> > Counterfactuals are *not* the same thing as indeterminism.
>
> OK, you've made that clear. But since every physical system implements counterfactuals, it leaves the
> possibility open that every physical system might implement a computation, perhaps a conscious computation,
> under the right interpretation.
Computationalsim is not the claim that *any* computation is conscious.
There is no objective sense in which every system implements every
computation.
Objectivity requires simplicity and observer-independence. I am not
consicous because some observer has decided I am!
> The implementation of a computation occurs as a result of physical activity
> (according to standard computationalism) but its meaning is a static, timeless thing, perhaps printed in a
> manual, perhaps held in the mind of the programmer, perhaps not physically available at all.
If computation has external semantics, it does not exist
in Platonia. If it does not, it doesn't have a "static, timeless"
meaning -- just an objective counterfactual structure.
> In the latter
> case - for example, if the manual is destroyed and the programmer dead - there is no reason why the computer
> should be any less valid a computer, or any less conscious if it were conscious to begin with.
But being (objectively) a computer means counterfactuals, and
counterfactuals
mean casuality.
> Stathis Papaioannou
> _________________________________________________________________
> Be one of the first to try Windows Live Mail.
> http://ideas.live.com/programpage.aspx?versionId=5d21c51a-b161-4314-9b0e-4911fb2b2e6d
--~--~---------~--~----~------------~-------~--~----~
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
-~----------~----~----~----~------~----~------~--~---
Received on Thu Aug 31 2006 - 10:21:10 PDT