Re: Problem for physicalist evolutionists

From: someone2 <glenn.spigel2@xxxxxxxxxxxxxx>:
On 21 Mar, 02:01, n...@xxxxxx (Eric Rowley) wrote:
From: someone2 <glenn.spig...@xxxxxxxxxxxxxx>: >




On 20 Mar, 19:10, n...@xxxxxx (Eric Rowley) wrote: > >> From:
someone2 <glenn.spig...@xxxxxxxxxxxxxx>: > >> > On 20 Mar, 01:31,
Rupert Morrish <rup...@xxxxxxxxxxx> wrote: > >> Hi,
Welcome back!
You haven't quite understood what I was stating with
regards to > >> > representation, though as I said, it isn't
important to be able > >> > to see that the metaphysical
assumption that if in principle, > >> > all the physical state
were known, and any rules in terms of the > >> > physical which
governed its interactions were also known, the > >> > change in
state of the physical over time (the behaviour) could > >> > be
explained was wrong. Though we can try once more, but don't > >>
get hung up on it if you are having problems getting it. > >> >
Consider the sequence of 32 bits (1's and 0's) > >> > In one
representation the first 8 bits represents the number of > >> >
pixels, the next 8 bits, its red value, the next 8 bits its > >>
green value, and the next 8 bits its blue value. > >> > In
another representation the first 4 bits represents an sound > >>
type, the next 4 bits represents represents the frequency, this
is repeated 4 times, such that 4 sound types and their
frequency > >> > can be represented. > >> > Now consider the
input to the robot, it is receiving data in > >> > 32bit chunks,
processing it, and outputting it. Is what is > >> > happening
representing it receiving visual information, and > >> >
outputting a report which will go to a speaker system, or is > >>
what is happening it receiving sound information, which when >
processed determines which fairy lights go on and off. > >>
There can be no consideration of what would be the point of it
switching fairy lights on or off, or the millions of other
things such a system could represent given the source of
input, > >> > and destination of output isn't specified. As what
it would > >> > actually represent would be whatever the context
was given the > >> > input source, and destination source.
But what is the point of pointing that out? > >> In a human
brain the context is given by which nerves the > >> signals enter
and exit by, the robot ANN might well run on similar > >>
principles, or, if it had a common data bus, the data packets
would have to have some form of address or >
identification. > >> I really don't understand what you're
saying with picture 5 > >> and Theory R, how do you take data
recorded from a visual > >> sensor and "consider that the
information was no longer > >> intended to represent visual
information, but that it was > >> intended to represented
auditory information"? > >> Are you saying that you feed the data
to the ANN through the > >> wrong input? > >> Or that one ANN was
programed or trained to process visual > >> data and the other
(identical) ANN auditory data? > >> In either case, are you
expecting the two ANNs to process > >> the data in the same way?
I certainly wouldn't! > >> In the first case I wouldn't
expect the part of the ANN > >> used to process visual data to be
identical to the part > >> used to process auditory data. > >>
And in the second case I wouldn't expect the two ANNs to > >> be
in the same state after programing/training.
If you can't get this then it >
isn't important for this first point. >
The following is however. > >> > You say that you don't
believe that the three theories in the > >> > article cover all
physicalist theories, but do you agree that > >> > physicalist
theorised realities either fall into the category > >> > where
all that exists is the physical AND EITHER in principle, > >> >
if all the physical state were known, and any rules in terms of >
the physical which governed its interactions were also
known, > >> > the change in state of the physical over time (the
behaviour) > >> > could be explained OR it couldn't. > >> > If
you do, you then need to understand that we are only > >> >
investigating the implications of the first case (that in > >> >
principle, if all the physical state were known, and any rules >
in terms of the physical which governed its interactions
were > >> > also known, the change in state of the physical over
time could > >> > be explained) for the current point. > >> Ok.
Given this hopefully you can see that what was being
illustrated > >> > in the article was that it wouldn't matter
what physicalist > >> > philosophical theory was offered, its
behaviour can be explained > >> > simply with knowledge of its
physical state, > >> But where have you shown that its phenomenal
experience isn't > >> part of its physical state?
and the rules in > >> > terms of the physical which govern
its interactions. Therefore > >> > which philosophical theory was
correct with regards to what it > >> > would be like to be the
robot, if it was like anything, is
irrelevant to its behaviour. > >> Only if the robot
doesn't base its behaviour on its experiences,
why wouldn't it? > >> > Sorry to labour the point here, as
you may have already > >> > understood, but I just want to make
sure that you are clear that > >> > if the assumption we are
investigating were true, that it would > >> > be irrelevant to
behaviour what it was like to be the robot, if
it was like anything at all. > >> Only if you assume that
that it would be irrelevant to behaviour > >> what it was like to
be the robot, which makes your argument > >> perfectly circular.
Read the article, >
I have! >
especially the point Professor Place makes.

Do you have an actual quote?
I find myself wondering how much of that point comes directly >
from Professor Place and how much is your interpretation. > I
remember from your previous appearances on T.O. that you > had a
tendency to mix quotes with your own material with no >
indication of where one ended and the other started. >
Even if the whole thing is the professors material I wonder
what you think it implies? >
It seems that it could be interpreted as simply saying that
there > are two different ways of looking at the same thing while
I get the > impression that you take it to mean that there are
two separate > things involved? >
Then you
will see when you get to the part I was discussing Rupert
Morrish > >above, that the point has nothing to do with showing
that the robots > >phenomenal experience, if it had one, to not
be considered part of its > >physical state. >
Then why wouldn't it affect behaviour?

If behaviour is determined by the interaction of the physical
state > and external inputs, and phenomenal experiences are part
of the > physical state then they ought to affect behaviour. >
The only thing I found that seemed relevant was your wierd
assertion > that perceiving a multitude of discrete bits of
information simultaneously > is a physical impossibility with the
system in your thought experiment. >
What property of the system makes it impossible? >
And why can't the system be upgraded to a more capable system?

The human brain is massively parallel, consisting of 100
billion > neurons, each linked to as many as 10,000 other
neurons. > If your ANN can't match that I suggest you wait a
decade or two > and run the experiment again. >
Would you mind answering my questions about picture 5 and
Theory R? >
How do you take data recorded from a visual
sensor and "consider that the information was no longer >
intended to represent visual information, but that it was
intended to represented auditory information"? >
Are you saying that you feed the data to the ANN through the
wrong input? >
Or that one ANN was programed or trained to process visual >
data and the other (identical) ANN auditory data? >
And what exactly is the point supposed to be?

What is the implication of the data being meaningless when
taken out of context? >
In a normaly functioning robot the data would surely be in
context? >
In either case, are you expecting the two ANNs to process
the data in the same way? >
I certainly wouldn't! >
In the first case I wouldn't expect the part of the ANN > used
to process visual data to be identical to the part
used to process auditory data. >
And in the second case I wouldn't expect the two ANNs to
be in the same state after programing/training. >

How can you claim to have read the article and then suggest that
I was taking what Place said to suggest that there were two
different things involved when I clearly state:

Very little of what you write is clearly stated. :-(

"Thus the brain could be identified in two ways, one which would
be the actual brain which could be looked at, if for example the
skull that housed it was opened up, and the other would be the
identity of what it would be like to be that brain."

Thus both identities are composed of brain.

This still doesn't clearly state that the two are the same thing,
but I'll take your word for it.

Questions such as why wouldn't what it would be like to be the
robot affect behaviour. Did you understand in the article that
three theories were put forward for what it might be like to be
the robot?

Yes.

Did you understand that theory C would be able to
explain the behaviour?

No, F is the only one I think would work for a ANN (or brain)
with a level of complexity approching the human brain.

If you did presumably you would have
understood:

But I didn't so I don't.

"Which would mean that scientifically, the implication is that if
a phenomenal reality were to be theorised for the robot, that the
theory would also need to imply that the phenomenal reality had
no influence on behaviour, else it would have to predict a
different behaviour from a theory (Theory C for example) in which
no such identity was posited,

That would indeed be the case if all three theories worked, I don't
think they do.

or contest the explanation of that
theory (the mechanism simply following the laws of physics)."

Here we go again!

I know that you assume that experiences affecting behaviours can't
possibly do so by being part of a physical mechanism that simply
follows the laws of physics, but you really can't expect a physicalist
to accept that unless you can support it somehow.

With regards to the representation argument, I'll simply repost
what I had just posted to someone else: --------------------Did
you understand that composition could hold no representation
other than in context?

I'm not quite sure what you mean by "composition "?
But otherwise, yes, so what?
Why would the data ever be out of context?

If you did, then you might have understood, that whether the
processing in a robot brain in a vat with lead walls 100 miles
thick represented visual information from the input, and
processing of visual information, and outputting of a report on
the visual information to a speaker system, or whether the
processing represented auditory information from the input, and
the processing of it used to determine which fairy lights should
light up, and the output being output to a fairy light controller
system, would depend on context.

Yes, normally I would expect visual data and auditory data to
come in on different input channels (eg the optical vs auditory
nerves in the human brain) and the outputs to the speaker and
lights to be output on separate channels.

If the input source was indeed
some type of automated synthesiser keyboard, and the outputs did
indeed go to a fairy light controller, then the processing did
indeed represent which fairy lights would go on based on the
sound input. How could the robot brain in a vat experience what
the processing represented without knowing the context, and how
could it know the context, given the 100 mile thick lead walls,
through which the inputs and outputs were fed.

How would the robot brain in a robot know the context, given the
2 mm thick titanium walls through which the inputs and outputs
were fed?

How do you know the context, given the 5mm thick bone walls
through which the inputs and outputs are fed?

If there is a arbitary fixed rule to how certain activity is
experienced, then this rule could have been different. If
conceptually what is experienced could be different dependent
upon what the rules were, then the experience would change if
there was a flucuation in the rules. There can be no reason given
why it would be one representation (devoid of context) in
physicalism,

Human babies (and quite possibly intelligent robots) learn the
contexts and fix the rules as they develop mentaly.

and therefore no objection to it have been another
representation,

Unless one is arbitrarily assuming that the experience has no
effect on behaviour there is the objection that an inappropiate
representation produces inappropiate experiences which lead to
inappropiate behaviour and would therefore be selected against
by any evolutionary or learning process involved in selecting
representations.

therefore no objection to fluctuation in
representation. With no objection to fluctuation in variation,
there exists at least one alternative explanation to a variation
in experience, other than a variation in the underlying
substrate. Therefore the assertion that there is no other
explanation that a variation in the underlying substrate is
false.

Therefore your conclusion fails.

As pointed out earlier, I don't have to explain the spiritual, I
simply have to show, that the story that all there is is the
physical, is at best implausible, and the most common formulation
of it simply wrong,

Which you have never done.

and that there is no logical objection to there being a spiritual,

Fair enough.

such as there being no alternative to the objections to the
physical having to also hold for the spiritual. So hopefully you
can see, it isn't that a flucation in the rules is being put
forward as an answer, I'm not attempting to explain the
spiritual, especially given language based upon the rule driven
physical world we have been given to experience. It is simply
used to show the assertion which is used as an objection to be
false. I would have a hard time explaining my sensation of free
will, as neither determinisic nor random, yet neither
deterministic nor random describes the sensation. Though unless
unless "willed" for example was allowed for an alternative, I
haven't the language to express it. --------------------

Might I suggest you take more time in reading the article, as it
seems as though you scanned it, simply looking for questions, and
didn't notice they were already answered.

I don't think most of my questions were answered.
Perhaps you could point out the answers to a couple?

I did find "They point out that while we may know that
a given neural cluster within the robot might represent
visual information, and that another might represent
auditory, it is not plausible that this could be recognised,
as there is no fixed representation of such a composition to
recognise."

Why could the ANN not recognize that data in the first cluster
represents visual information and data in the second represents
auditory information?

After all it has to recognize the difference somehow, otherwise
how can it apply the proper processing to the data?
Since it obviously has the needed information, why would it not
use it when producing experiences.
And having produced the experiences, as neural states, why
would it not use them, like any other data, in the production
of behaviours?

You seem to have the weirdest ideas about how brains/ANNs work.

And I already mentioned your assertion that "perceiving a multitude
of discrete bits of information simultaneously is a physical
impossibility with the system in my thought experiment."
but as long as you avoid explaining why that would be so I don't
really consider it a useful answer.

Eric


.

Relevant Pages

  • Re: Problem for physicalist evolutionists
    ... the robot ANN might well run on similar>>> ... ANN auditory data? ... if all the physical state were known, and any rules in terms of> ... follows the laws of physics, but you really can't expect a physicalist ...
    (talk.origins)
  • Re: Problem for physicalist evolutionists
    ... something to be the robot. ... deviation from the *currently known* laws of physics was built, ... output to a fairy light controller system, would depend on context. ... then the processing did indeed represent which fairy lights would go ...
    (talk.origins)
  • Re: Significance (contd)
    ... robot build following the same laws of physics as things that it isn't ... built following the same laws of physics as a microwave oven. ... expect the toaster to exhibit Y, am I claiming simply that the same laws ...
    (talk.origins)
  • Re: Explain the Law of conservation of matter and energy
    ... >> such thing as force at all in modern physics. ... > difficulty in understanding context. ... > after form gives it an entirely different meaning. ... Does normal in 3rd normal form mean average? ...
    (sci.physics.relativity)
  • Re: Implausibility that we can be explained by evolution
    ... experiencing" on such a robot, because of the behaviour it displayed? ... are simply be mechanisms following the laws of physics. ...
    (talk.origins)
Loading