The reason why I very strongly insist that subjective "private" experiences are a matter of physics is because our experiences are highly specific physiological responses to specific kinds of physical stimuli.
OK, then, point to whatever specific physical thing encodes an experience and explain why only that physical thing can encode that experience.
Information in the abstract does not trigger the sensation of pain or the perception of red; specific chemical signals in the body are required to trigger these experiences.
The specific chemical signals in our body are an artifact of our evolutionary history. As the Wasp rightly pointed out, there is no reason to believe that the specific chemical signals (I will assume you are talking about neurotransmitters here) we use are required for the job -- in theory we could swap them out for an entirely different set of neurotransmitters and receptors and get functionally identical results.
As a thought experiment, would you agree that we could (in theory) replace the neurons in the brain one at a time with nanomachines that are functionally identical at a neuron-by-neuron basis? Why or why not?
If a person is exposed to psychoactive substances their change in mental states is due to the nervous system's physical reaction to the reagent and not a magical emergent property of algorithmic code execution.
I am personally subjectively familiar with the process.
We already know that neural networks do not process information the same way that our usual Von Neumann/Harvard architecture computers do -- no surprise there. That does not stop us (in theory) from emulating neural networks using algorithmic processes to whatever degree of detail needed. In practice of course, there are huge engineering challenges, but they are just that --
engineering challenges.
We are also pretty sure that psychoactive substance work because they (or their metabolites) either impersonate neurotransmitters, or they mess with the release/reuptake systems in synapses for particular neurotransmitters, thereby messing with the usual synaptic weighting (and, therefore, firing rate of the target neuron) for the sites the drugs affect. If that network happens to participate in a process involved with consciousness, then consciousness will be affected or altered, but as a result of the drug messing with the synaptic weights or firing rates of the neurons involved, not because the drug inherently possesses some magical consciousness altering property.
The different chemical compositions of the substances that bind to cellular receptors determine the type of subjective response, if any, that a given conscious person will experience.
First, the specific substances (I will assume you are talking about neurotransmitters and substances that impersonate them here) and their receptors are artifacts of our evolutionary history -- there is no reason to assume that the specific ones we use are the only ones possible.
Second, I see no reason to expect that we will find consciousness at the level of synaptic activity and neural excitation levels, especially since pretty much the same neurochemistry is used by everything that has a nervous system, whether or not we recognize them as conscious.
Somehow, properties of the signal molecules are conveyed by the EM signals along the cell membrane which, in turn, appear to have direct affect on the subject's conscious experiences.
No, the neurotransmitters do not have any special properties beyond the fact that they bind to neurotransmitter receptors. Even then, we have no reason to directly equate neural polarization levels and excitation thresholds to consciousness and subjective experience, and even if we did there is no reason that we could not in theory simulate those features in an artificial neural network.
There is no way a computer scientist can properly address the
questions I raised in terms of I/O switching functions because they are inherently
biophysics questions.
I have good reasons to doubt that, as outlined above.
Our consciousness and conscious experiences are undeniably a result of the physical conditions and biological mechanisms of our brains
Indeed.
and it's essential properties cannot be seriously tackled purely in terms of IP functional language.
Not so much.
At some point we're going to have to deal with the actual physics of what the brain is doing instead of arrogantly -- lazily -- chalking it up to "computation/information processing/SRIP/etc." because the prospect of unknown science makes our brains hurt.
Well, we have no indications that there is an "unknown science" at play, at least not at the level of physics or chemistry. I understand you believe differently, but it is just a belief right now.
Do you -honestly- believe that any substrate of any composition will have subjective experience merely because its implementing a particular switching pattern?
As long as it meets the requirements I outlined, then yes.
But you could determine this if you knew it were implementing a particular line of code?
A particular line of code in isolation? No, not any more than you could determine that a system is conscious by looking at a single neuron.
If consciousness itself could be reduced to something apart from all of those cognitive capacities what makes you think that its simply a matter of computational coding?
What makes you think it is not? Nevermind, you think it is some special property of our neurochemistry.
So all substances are conscious and it's simply a matter of waking them up with the correct algorithmic procedure?
No, I think that systems that meet the criteria I outlined are conscious, no matter what they are built out of.
We just established that linguistic and/or motile behavior may not be possible for a some conscious entities. Absent such a behavior test, how else could one discern whether or not they're conscious?
In general, we cannot. We could probably build a test for consciousness that is not behavioral if we know the details of how consciousness is implemented in entities of whatever type, but such tests would not be general.
As far as the subject is concerned, their sensory experience is very much like an output; some think of it as something akin to a fully immersive theatrical experience. Sensory stimuli that do not make it onto this stage of conscious experience are what we call subliminal. In any case, if we identify consciousness we'd have identified the experiencer.
The Cartesian Theater called, they want their homunculus back.
The problem is that you're putting the cart way before the horse.
You are insisting that our designs for a cart replacement are ridiculous and can never work because we swapped the horse for an internal combustion engine.
We have not yet identified what physically constitutes our consciousness and computational descriptions are not a sufficient substitute.
Yes, and?
At this point, computation has just become a "god of the gaps" explanation. Computationalism isn't science, its a placative ideology serving to distract AI researchers from the fact that they really have no idea what consciousness is.
That is your opinion, certainly.
