[ExI] Fwd: Mental Phenomena

[Brent Allsop]

Hi Stathis,

You continue to use the word: "See redness".  The word "see" means using
eyes to detect light.  Seeing has a long chain of different physical causal
things between the target and the knowledge.  So, to me you are saying
something like there is a homunculus inside my brain "seeing" this redness
through some chain of different causal events.  Seeing is an
objective process, the final result of which is physical knowledge, which
is very different than, or only a representation of the target being
"seen".  We don't 'see' this knowledge, it is the result of seeing.

On Fri, Jan 31, 2020 at 12:14 AM Stathis Papaioannou via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

>
>
> On Fri, 31 Jan 2020 at 14:06, Brent Allsop via extropy-chat <
> extropy-chat at lists.extropy.org> wrote:
>
>> OK, here are a few of the more important things you are missing.  You
>> said I would say:
>>
>> “Finally, there is the actual conscious experience, such as seeing the
>> redness of the strawberry.”  I would not say: “see the redness”.  There
>> are two ways to know about the same physical facts.  1. To objectively
>> perceive them through a cause and effect chain of different physical
>> events.  You don’t “ignore” the different physical qualities of all these
>> intermediate things, you just think of all these different things in a
>> substrate independent way, thinking or interpreting each of them as if they
>> are only representing “red” instead of modeling how they are physically
>> nothing like red.
>>
>
> Here I don't  quite understand what you mean by 'thinking or interpreting
> each of them as if they are only representing “red” instead of modeling how
> they are physically nothing like red'. If glutamate has a redness qualia
> property, and I only look at glutamate's chemical composition, solubility,
> density and so on, am I 'interpreting' glutamate as 'only representing
> "red"'? What I would like to do is consider all the properties of glutamate
> EXCEPT the postulated redness property. Let's say it's there, but we just
> ignore it, for the sake of the argument.
>

Would you agree that if we consciously know something, there must be
something that is that knowledge?  In computers, we say there must be
something that instantiates the 1s and 0s.  But this "instantiates" is
separating whatever physics we are thinking of as being 1s or 0s.  The only
way to get the 1s, from that particular set of physics, is if you have a
working dictionary mechanism to reverse the "instantiation".  With
consciousness redness just is a quality of our conscious knowledge.  We can
think of this physics as "instantiating" red knowledge, but to get the word
"red" from redness, just like getting a 1 from +5 volts, requires a working
dictionary mechanism.  You don't need a dictionary for redness, it is just
a quality.   You don't see redness, it is computationally bound in with
other physical knowledge to make composite qualitative experiences like
conscious knowledge of a strawberry.

If we have an RGB video cable with 3 RCA connectors on each end.  Each of
the connectors are covered in black plastic, but has either an R, G, or B
label.  This is just a set of 3 copper wires, covered in black
insulation/plastic, with gold connectors.  The one with the R letter is in
no way physically 'red'.  We just think of it as being the one that caries
the 'red' part of the signal.;  And of course, the varying voltage in that
wire, isn't anything like a red strawberry, nor is it anything like redness
knowledge.  We just think of the signal in the R wire, as "representing"
red things in the video image.  We can model the actual physical make up of
the wire, and the redness signal, with an objective description of it.  But
the only thing important is modeling this wire as the one carrying the
'red' signal, and that some place down the line, we will end up
interpreting this 'red' signal, by instantiating it with redness knowledge.


>
>> The second way to know about physics is to directly subjectively
>> experience them.  The first is the “objective” way, requiring lots of
>> interpretation, which can be mistaken or a seeming.  The second is the
>> direct subjective way, which can’t be mistaken and there is no
>> interpretation.  Physical redness is just a physical quality which we
>> can directly experience and are objectively describing, when we abstractly
>> describe glutamate reacting in a synapse, possibly including the way you
>> abstractly describe glutamate as changing the shape of the receptor.
>>
>
> The use of the word "interpretation" is confusing. If I say "glutamate has
> a density of 1.46 g/cm^3" is that an objective observation, and if so what
> am I "interpreting" when I make it? And when you say it could be "mistaken
> or a seeming" do you just mean that my scales might be calibrated
> incorrectly, and hence the real density might actually be 1.53 g/cm^3?
>

"has a density of 1.46 g/cm^3" is a set of abstract symbols objectively
describing a physical property.  Using abstract descriptions like this, we
can fully describe and model and predict everything about glutamate.  The
prediction is, that we can do the same thing, fully abstractly describing
redness.  But these abstract descriptions of redness are not the glutamate,
nor are they these descriptions redness.  The only thing that has a color
quality is subjective knowledge.  There are two images in the camp
statement of "Representational QUalia Theory
<https://canonizer.com/topic/88-Representational-Qualia/6#statement>".  The
only thing in these images are the subjective knowledge. Everything else is
in black and white, indicating all our subjective knowledge of it provides
no color information.  That is done with our brain, when it interprets the
objective information and represents it with something that has a redness
quality.  Just like Frank Jackson's Mary, we can objectively know and
describe and predict everything about 'red'.  But until we directly
experience what all the objective knowledge is knowledge of, we can't know
which of it is redness or greenness.


> The use of the word "abstract" is confusing. Do you just mean "objective"?
>

"Interpretation" (used earlier) is when you convert or transduce one
physical representation into another different one.  Our brain interprets
an abstract word like "red' to the redness quality.  All "objective"
information is just something physical, representing something different
than it is.  "Abstract" is the way many people use the word 'red' to
represent anything to do with the perception of red.  Again, such an
abstract description gives you no qualitative color information.


>
> The contrast of "objectively describing" and "abstractly describe" is
> confusing. It seems that by "objectively describing" you mean what I would
> call "subjectively describing", and what you call "abstractly describe" is
> what I would call "objectively describe".
>

"abstractly describe", "objectively describe", "symbolically represent" are
all synonyms.  They are methods of modeling or representing things in
symbolic ways.  Without a dictionary defining these symbols or
representations, there is no qualitative color information in any of it.


>
> Next: “from here on I get a bit lost. It seems to me that you are saying
>> that since redness is a physical property of glutamate, it must have
>> physical effects, and therefore it is not possible changing it would leave
>> the qualia and behavior `unchanged.”
>>
>> The thing that leads you astray, is you interpret what I’m trying to say
>> as if glutamate and redness are separate things.  You are clearly saying it
>> here in a way that separates these two.   They are not separate in any
>> way.  Objectively, you see the glycine firing, instead of glutamate,
>> resulting in you saying: That is not red.  Subjectively, you directly
>> experience this same physics as redness, or greenness, enabling you to say
>> what color they are.  The objective and subjective facts are all the same
>> thing causing you to say: “That is not red.”  Saying redness causes you to
>> say: “that is red” is the same thing as saying glutamate, (possibly
>> changing the shape of the receptor) causes you to say: “That is red”.
>>
>
> I think you are saying: if glutamate in fact has redness as a fundamental
> property, then subjectively when glutamate is released in your own synapses
> you will see red, and objectively if you see glutamate released in someone
> else's synapses (without directly sharing their experiences) you can
> confidently say that they are experiencing redness; and analogously for
> glycine and greenness.
>

I think this is accurate, if your sloppy terminology is interpreted
correctly.  As I said above, It would be less sloppy if you replace "see
red" with "directly experience your redness."


>
> I object to the statement that 'Saying redness causes you to say: “that is
> red” is the same thing as saying glutamate, (possibly changing the shape of
> the receptor) causes you to say: “That is red”'. This is because changing
> the shape of the receptor provides a complete causal explanation for saying
> "that is red". That is, an advanced scientist could examine the glutamate
> and associated systems without any understanding of what redness is, what
> qualia are, what human language is and so on, and predict that the sounds
> "that is red" will come out of the subject's mouth. Because redness is not
> necessary to predict what the system will do, it cannot be included in the
> list of causal properties. At a higher level of description it is true that
> the subject reported seeing red due to the redness, but not at the low
> level where the physical interactions occur.
>

"changing the shape of the receptor provides a complete causal explanation
for saying 'that is red'."  Again, though this is a "complete causal"
description, you can't know the quality it is describing.

"Because redness is not necessary to predict what the system will do,"
This is the part you erroneously continue to ignore.  What the subjective
experience is like is all important.  In a future world we will discover
which physics it is that has a redness quality (let's assume it is
glutamate) resulting in a dictionary mapping subjective redness to
conscious knowledge represented with glutamate.  So the most important
behavior is, what does it say when you ask the subject what it's redness is
like?  If the subject has this dictionary information, the subject will be
able to eff the ineffable nature of his subjective experience and report
it's knowledge has a glycine quality, which is what you represent greenness
with.  The behavior must include things like reporting on the quality of
their subjective knowledge.  No other behavior matters but this.

"At a higher level of description it is true that the subject reported
seeing red due to the redness, but not at the low level where the physical
interactions occur."  We must also think of things from a subjective point
of view.  Conscious visual knowledge includes thousands of individual pixel
elements, each of which has an elemental colorness quality.  So, for each
pixel, there must be something you can objectively observe that is that
particular subjectively elemental pixel of colorness.  When that one pixel
changes from redness to grenness, there will be an objective description of
the physics, that is that subjective redness changing to greenness.  Again,
you keep eliminating these necessary elemental parts of consciousness from
the neural substitution, resulting in all the problems.



> And you completely missed the second necessary half of what I model
>> consciousness as being.  This is strong evidence that you are completely
>> blind to this half.  And it is THIS half that reveals all the problems in
>> the neural substitution argument.  In my view, this is a necessary part of
>> having composite qualitative experience or consciousness which we
>> absolutely have.  When you do the neural substitution in the way you do, it
>> completely leaves out this necessary half of the functionality and this is
>> part of the cause of all the craziness and “hard problems” that don’t make
>> any sense.
>>
>
> I guess you are referring to the binding system, but I have said before
> that any function whatsoever of neurons can be substituted leaving
> behaviour exactly the same.
>

This terse dismissal of the binding system, basically saying:  "That isn't
important, because of what the neural substitution argument demonstrates."
is the core to what I see as our misunderstanding.  I wouldn't mind if you
ignored everything else and we just focused on this.  The neural
substitution, as you describe it, can't account for the necessary binding
functionality making it possible to have composite qualitative conscious
experiences.  So doing any substitution on any system that lacks the
functionality required to be conscious, is irrelevant.  If you include a
binding system in whatever you are substituting, it will all make sense.
If you don't, that's when all the 'hard problems" arize.



> Oh, and you also completely left out the fact that when doing the neural
>> substitution without this necessary half of the functionality, you can
>> prove that nothing, not even “functionality” nor “magic” can have a redness
>> quality, for the same reasons glutamate can’t have redness.  You just never
>> notice this, because no matter where I try to put redness to make you
>> happy, even if this is op amp functionality or magic, you change the
>> subject and redefine everything as if the redness is someplace else, in the
>> system.  You never model what it is that has the redness quality, even if
>> it is functionality.  Because, if you did, I could prove to you that it
>> can’t be that, for the same faulty logic causing you to think it can’t be
>> glutamate.
>>
>
> This may be hard to conceptualise, but the redness must be in the minimal
> system that is required for redness,
>
This sounds right to me.


> and not localised in any part of the system.
>

Again, you're ignoring some necessary facts about subjective experience,
and trying to get out of objectively describing redness. (because if you
did, it can't be that, for the same reason glutamate can't be glutamate.)
Regardless of whether an individual pixel is localized to glutamate, or
distributed across some kind of network, subjectively, it is isolated to
that one elemental subjective pixel.


> This might be clearer if we consider something like the understanding of
> language. The entire system needs to be working, not just one part of the
> system for a particular word.
>

Again, from a subjective point of view, it is the elemental colored pixels
representing the surface of the strawberry that are the letters that make
up the knowledge of the strawberry.  Just as you can have a single letter
of a single word, you can have a single pixel of redness, that is just
redness.  All you need is something that is objectively observable that is
redness, there must be something that is objectively observable as being
different, that is greenness, and there must be some mechanism that allows
it to be computationally bound into one composite qualitative experience.
The way you describe what you're neuro substiting, can't do these 3 simple
things.

What I’m trying to say is IF you include anything that could have a redness
>> quality we can directly experience, and if you would include the second
>> half of the necessary part of consciousness, and if you adequately modeled
>> both of these parts in your neural substitution, not only would everything
>> be obvious about what is going on, it would become clear how various slight
>> of hands can be done with this kind of substitution, to make it appear that
>> the neural substitution is doing one thing, when in reality it could be
>> physically or subjectively very different.
>>
>
> I don't recall you giving a yes or no answer when I have asked if you
> think replacing glutamate with an analogue that affects glutamate receptors
> in exactly the same way would result in: (1) a change to the behaviour an
> external observer would see, if he were free to do any tests or ask any
> questions; or (2) a change to the redness qualia.
>

It all depends on what the necessary and sufficient set of physics for
redness is.  Redness could be just the receptor response, which could be
duplicated by either variant of glutamate.  If it is the response of the
receptor that is the redness quality, then we need to substitute all my
usages of glutamate, with "glutamate receptor" as the objective description
of redness.  And you must be able to computationally bind this receptor
with other colors to make a composite qualitative experience.

I guess another part I’m not stressing enough (trying present things from
>> your point of view) is the prediction that no such neural substitution
>> would ever be remotely possible, except in only the most contrived and
>> absurd ways.  Once you got to a pixel neuron, presenting glutamate to the
>> binding system, enabling you to be directly aware of it with all the other
>> physical pixels and their physical qualities as one composite experience,
>> the only way to do a switch, in a way that would even remotely behave the
>> same, is if you, in one atomic switching operation, rebound everything,
>> including all memories and names of things that are red with whatever you
>> substitute redness with.  I’ve tried to describe exactly this, many times
>> in a neural substitution where the binding system is a single neuron –
>> resulting in exactly all the switching being made, in one atomic swap
>> between the real and the simulated.  In other words, when the binding
>> system was replaced, in one atomic switch, there would be dancing (between
>> red and green or absent) qualia.  The only way it would work is if you
>> replaced everything, all at once, and remove any knowledge from the system
>> including that the word redness, before the switch, was bound to redness,
>> while after it is now bound to greenness (or the word red).  To say nothing
>> of how hard it would be to get something that represented red things with
>> the word red, to behave as if it’s knowledge really was physical redness,
>> instead of an abstract word.
>>
>>
>>
>> Does any of this seem like evidence as to where the problem is, as it
>> seems like evidence of the problem to me?
>>
>
> I don't understand the problem here. If glutamate is replaced with a
> glutamate analogue then everything will work exactly the same, and if you
> do this with all the neurons in the brain, all the neurons will function
> exactly the same. I think you still imagine that removing the glutamate
> will somehow change the behaviour of the binding neurons, or the
> description of the qualia, or something; but it cannot, since that would be
> a logical contradiction given the initial assumptions. I am sure either you
> or I are missing something here.
>

It is still the necessary binding system that you continue to ignore
causing all the problems.  The discrete set of neurons you are describing
doing a neural substitution on, cannot accomplish the required ability to
have lots of pixel qualities, computationally bound into one composite
qualitative experience.  So whatever you can or can't do to it, is
irrelevant.  Consciousness must be something which includes the required
subjective binding functionality and qualities.  If you can point to
something in the system that is the redness quality, if you could describe
how this could be different than the greenness quality, and how both of
these can be computationally bound into a composite experience.  Everything
will just work and there will only be a color problem.  You are not doing
this, and that is the problem.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20200131/8914ee84/attachment-0001.htm>