[ExI] Digital Consciousness .

Brent Allsop brent.allsop at canonizer.com
Sun Apr 28 17:25:40 UTC 2013

On 4/28/2013 10:01 AM, Stathis Papaioannou wrote:
> On 29/04/2013, at 1:14 AM, Brent Allsop <brent.allsop at canonizer.com> wrote:
>> No, No, No.  You guys still aren't getting it.  Let me try, yet again, in hopefully a better and more complete (thanks to your continued help) way of describing what I believe.
>> This 'third' class of possibilities you are talking about is obviously accepted by everyone I know to be absurd.  Yet everyone in your camp seems to spend all their time fighting this strawman, as if everyone was giving it serious consideration.  James Carroll wrote and published an entire paper fighting this straw man 'third' class of epiphenomena possibilities idea.  You guys need to stop making fun of everyone not in your camp, as if they are talking about this 'third' class of possibilities.  Including things where you don't notice fading qualia, or any other kind of" epiphenomena" that has no causal correlate in reality.
>> (Note:  Since I mentioned James Carroll, and his paper, and since I'd like his input on this, I've CCed him.  And I've CCed Steven Lehar, in case we can get some of his always brilliant feedback on this epiphenomena issue.)
>> Even Chalmers wastes a huge portion of the words in the paper focusing on this absurd idea that nobody thinks is in any way important or credible, or worth wasting time on.  Because chalmers wastes so much time on this possobility, everyone redicules it as if he does give it serious consideration.  All this is so frustrating.  That is why Canonizer.com is so important, so everyone doesn't waste their time on arguments like this, which nobody takes as a serious possibility.
>> If anyone here is giving this 'third' epiphenomenal class of possibilities any serious consideration, please speak up now, so we can quickly nip any such absurd idea in the bud, sufficiently so we don't have to return to it, yet again.  Either way, will all of you stop assuming I do believe in this epiphenomena possibility, and arguing against it, assuming you are refuting my working hypothesis by doing so.
>> Property Dualism theories are predicting that there is very real causal properties to a redness qualia, and that through these causal properties, they have a huge influence in the world, through our consciousness.
>> Property Dualism is also predicting that these causal properties of redness are some necessary and sufficient subset of the causal properties we already fully understand, at least abstractly and behaviorally (just not yet qualitatively).  We are not predicting there is some additional new magic, quantum, or whatever causal properties that just haven't been discovered yet.
>> The qualitative natures of these causal properties, suffer from the 'quale interpretation problem', making them blind to traditional cause and effect based observation, because, by definition, all such observation involves multiple arbitrary mediums, being interpreted as something they are not in some abstracted way (requiring interpretation).  And if you don't know how to properly interpret that something is representing a redness qualitative nature, you will be blind to it, while still getting the abstracted and or behavioral information they are being properly interpreted as representing.
>> The system will be aware of the phenomenal qualities of both redness and greenness, and it will be saying glutamate (or some Functionally active pattern, or whatever you predict are the neural correlates of redness) is the only thing that has a redness quality to it.  In other words, when you look at the system, via abstracting cause and effect based observation, all you will see is a high fidelity detector of glutamate, that picks it, because of the causal properties of glutamate.  Nothing else but real glutamate will produce a "yes, that has my redness quality" abstracted response.
>> Of course, you will be able to implement some additional hardware in the system, such as something that will communicate to the system that knows what a redness quality is like, and you could do something like, wire it so that it knows that this other thing is a greenness quality, (or worse some abstracted set of bits, like 0xFF0000, that is meant to be interpreted as if it has a redness quality (in addition to interpreting as if it is the abstract idea of 'red') despite the fact that it, by definition, does not have the redness quality.
>> Only when you do obviously lying and blatant enginered interpretation of things that do not have a redness quality, as if they are something that does have a redness quality, will you be able to get the entire system to start behaving in a way that can be interpreted as if it really is aware of both the qualitative nature of it's knowledge, and the abstracted nature.  All of this will have profound and obvious 'interpretations' being made, and have huge effects on what it will be like, subjectively, and qualitatively, for the system - all producing very different behavioral responses.
>> Another critically important issue is how property dualism is predicting this stuff can be effed, in various strong and week ways.  In other words, once you know the necessary and sufficient causal properties that reliably produce a redness quality, you will then know how to interpret the abstract informaiton of the behavior you are observing.  This grounding of the symbols will then let you know the qualitative nature of the redness quality (and it's neural correlate causal properties), and not just the abstracted and behavior idea of 'red'.  To say nothing of the stronger means of effing the ineffable, where two hemispheres, or two brains, are aware of the same causal properties that is behaving the way it does, because of the ineffable qualitative nature it has, which we can experience.
>> And, the prediction is, as I believe Stathis has admitted.  Once you replace significant portions of the brain (way more than just one neuron) with some quite sophisticated interpretation hardware which can interpret a serious of ones and zeros that by definition do not have a redness quality to them, as one that does, you will never be able to eff, or know that the ephiphinomena only you guys think others think could exist epiphinominally with such.  All of this will have effing profound and obvious causal and subjective behavior, as you attempt the neural substitution.
>> Can we now drop the worthless epiphenomena issue, once and for all, and stop wasting everyone's time and assuming anyone that thinks different than you is wrong because you think they believe in such, and from now on focus on what is important?  And that is what is the set of causal (or informational?) properties that has the redness quality we experience, which we represent and choose abstract information like 'red' with because of it's qualitative nature?
> Brent, I'm not sure what you are calling the "third possibility" and "epiphenomena", but I guess you don't believe that the qualia can fade without you noticing, and I guess you don't believe that the qualia can fade with you noticing but being unable to change your behaviour. Good! So since your behaviour cannot change due to the replacement you must agree that the qualia must be preserved.
> Having gone through this many times, I suspect that the real problem is that you still don't understand why your behaviour can't change due to the replacement, even though you agree that the behavioural properties of the brain are computable. Am I right in this?

Hi Stathis,

Thanks for putting so much effort towards this, and I apologize that I 
still, despite significant progress thanks to everyone's help, have so 
much difficultly communicating.

Yes, I don't believe that "qualia can fade without you noticing" and 
that it will not be possible for you to notice, without changing your 
behavior, or any other way qualitative natures are disconnected form 
consciousness, and it's underlying neural correlates.

You still believe that my real problem is that I still don't understand 
why our behavior can't change due to the replacement.  I fully 
understand all this, and you're still jumping back to a straw man that I 
also do not accept and agree with you.  There is yet another option that 
you aren't fully getting yet, that is not anything like any of these 
epiphenomal qualities that are disconnected from reality.  In this way, 
the qualities are very real, and they have very real causal properties.  
These causal properties are properties we already know, abstractly, all 
about it's behavior.  We just don't know about it qualitative nature.  
We will think the system is detecting glutamate, merely because of it's 
causal behavior, when in fact, it is detecting it because of the 
qualitative nature, of which the causal behavior is a mere symptom, and 
all we know are abstracted interpretations of the same.

Let me try to put it this way.  James has admitted that an abstracted 
representation of the causal properties of glutamate is just something 
very different than the causal properties of glutamate, configured in a 
way so that these very different causal properties, can be interpreted 
as real glutamate.  In other words, he has admitted that the map is not 
like the territory, other than it can be interpreted as such.  Do you 
agree with that?  And if you do, is it not a very real theoretical 
possibility, that the reason glutamate is behaving the way that it is, 
is because of it's redness quality.  And also is not a very real 
possibility that the real glutamate has this ineffable quality (blind to 
abstracting observation or requires mapping back to the real thing) for 
which the abstracted representation of the same, though it can be 
interpreted as having it, doesn't really have it.

Brent Allsop

More information about the extropy-chat mailing list