How to measure what remains of the hard problem

Malcolm Lett

Let's say that there is a theory that explains consciousness in a computational way that relies entirely on physical phenomena using the science that we have today. It explains the outward behaviour of conscious beings, the inner state of the mind and the computational processes that operate against those states. Importantly, it explains in a computational way, why the being is aware of their own thoughts, the boundaries of those thoughts (ie: what can be represented in them vs not), and the processing steps involved to conclude consciousness.

However, the theory simplifies the whole thing to just data in, data out, and processing logic. It could be emulated entirely on a computer CPU -- not "human" consciousness itself, with emotions and the human level of intelligence, per se, but the same mechanics in a scaled down way.

Yet, when we now consider that explanation, we still think there's something missing. We think that a scaled up version of that theory would merely produce a sort of 'unconscious processor' that produces all the same behaviours, both externally and internally, as a human, but does not have phenomenal experience. Basically a specific kind of p-zombie.

In summary, we (well, some) think that human experience includes an extra phenomenal experience that is beyond the computational mechanics. Even if all of the data representations of a human concluding themselves as conscious can be emulated, we still think there is some other aspect to our experience that isn't represented as a 'data state' or a 'data input'. There's a gap - something that we aren't measuring in our computational analysis.

I'm wondering what theories there are that specifically address the question of measuring this gap.

apokrisis

I'm wondering what theories there are that specifically address the question of measuring this gap. — Malcolm Lett

(Given I’m responding to your second good post, I should say welcome to the chat. :up: )

Anyhow, you could say that computer science kind of does measure this. And what it has discovered is that all the talk about representation and data was the wrong way to go about getting anything resembling biological intelligence.

Neural network or Bayesian Brain architectures start to approach the issue in a more biologically realistic and ecologically embodied fashion. The gap then begins to measureably close.

So in a negative fashion perhaps, computer science does point towards the need to understand biology as something inherently intelligent and purposeful.

You can’t start off with what seems the “output” - a mental “representation” - and implement that in some simplistic computational fashion. You have to keep heading in the direction of biological science to have a hope of getting to the root of this question.

Wayfarer

In summary, we (well, some) think that human experience includes an extra phenomenal experience that is beyond the computational mechanics. — Malcolm Lett

I suggest that this element is what is designated by the term 'being' in the compound word 'human being'.

But it's a mistake to say that it's 'phenomenal'. 'Phenomenal' means 'what appears', whereas 'being' is what appearances appear to. 'Beings' appear to us as 'other beings', however both their being and ours is not 'what appears', as such. Likewise, being is not 'an experience' but 'the capacity for experience'; 'beings' are 'subjects of experience' but they're not themselves only experience, as they also comprise the elements that order and interpret experience (per Kant).

There's a gap - something that we aren't measuring in our computational analysis. — Malcolm Lett

This is referred to as the 'explanatory gap'.

Ask yourself what 'an explanation' provided by computational analysis would comprise. With scientific hypotheses, generally, and at a high level, there is the left-hand side - which is the equation or prediction - and the right-hand side - which is the observation or result. Scientific method demands that the prediction or equation be validated against the result or observation.

However, in this case, the object of analysis is also the subject doing the examining. It's precisely because you can't stand outside or, or 'objectify', the object of analysis that is the cause of both the 'hard problem' and 'the explanatory gap'. This is why it is in principle outside the scope of empirical analysis, and why, for example, accounts such as Dennett's must insist that it be eliminated altogether.

I'm wondering what theories there are that specifically address the question of measuring this gap. — Malcolm Lett

It's only 'a gap' by way of analogy; in actual fact, it's more an incommensurability between the methods of scientific naturalism and the subject of the analysis.

(See It is never known, but it is the knower, Michel Bitbol.)

TheMadFool

Basically a specific kind of p-zombie. — Malcolm Lett

I'm afraid no computer will ever qualify as a p-zombie. The idea of a p-zombie is that it has to be physically identical to a human and no computer is or can be such.

That said, there's no reason to believe consciousness isn't replicable on another kind of substrate, something non-biological. I say this because consciousness, to me, is simply data processing and anything capable of handling data is, in principle therefore, also capable of consciousness.

apokrisis

I say this because consciousness, to me, is simply data processing and anything capable of handling data is, in principle therefore, also capable of consciousness. — TheMadFool

My iPad handles data. So that doesn’t feel particularly convincing.

Why not reserve your admiration for a system that shows itself capable of handling the world?

TheMadFool

My iPad handles data. So that doesn’t feel particularly convincing.

Why not reserve your admiration for a system that shows itself capable of handling the world? — apokrisis

Well, what is consciousness if not data processing? Think of the times when we all agree that a person is not conscious e.g. when sleeping or when s/he has fainted or when s/he's dead. These three states of unconsciousness have one thing in common - the absence of thoughts and what are thoughts but data being processed?

apokrisis

Well, what is consciousness if not data processing? — TheMadFool

If it ain’t data processing then it ain’t data processing.

The question is why you would think it was?

Think of the times when we all agree that a person is not conscious e.g. when sleeping or when s/he has fainted or when s/he's dead. These three states of unconsciousness have one thing in common - the absence of thoughts and what are thoughts but data being processed? — TheMadFool

There is thinking - of a desultory and ruminative kind - even in deep sleep as it happens. It is just unremembered and disconnected.

But the real issue here is in what sense do you think that the brain does “data processing”?

That is fine as a vague metaphor. But the brain isn’t designed to be a Universal Turing Machine - the standard formal definition of data processing.

If you have some other precise definition of data processing, now is the time to reference it.

Wayfarer

Computers, quite literally, process information – numbers, letters, words, formulas, images. The information first has to be encoded into a format computers can use, which means patterns of ones and zeroes (‘bits’) organised into small chunks (‘bytes’). On my computer, each byte contains 8 bits, and a certain pattern of those bits stands for the letter d, another for the letter o, and another for the letter g. Side by side, those three bytes form the word dog. One single image – say, the photograph of my cat Henry on my desktop – is represented by a very specific pattern of a million of these bytes (‘one megabyte’), surrounded by some special characters that tell the computer to expect an image, not a word.

Computers, quite literally, move these patterns from place to place in different physical storage areas etched into electronic components. Sometimes they also copy the patterns, and sometimes they transform them in various ways – say, when we are correcting errors in a manuscript or when we are touching up a photograph. The rules computers follow for moving, copying and operating on these arrays of data are also stored inside the computer. Together, a set of rules is called a ‘program’ or an ‘algorithm’. A group of algorithms that work together to help us do something (like buy stocks or find a date online) is called an ‘application’ – what most people now call an ‘app’.

....I need to be clear: computers really do operate on symbolic representations of the world. They really store and retrieve. They really process. They really have physical memories. They really are guided in everything they do, without exception, by algorithms.

Humans, on the other hand, do not – never did, never will. — Robert Epstein

Remainder here

TheMadFool

There is thinking - of a desultory and ruminative kind - even in deep sleep as it happens. It is just unremembered and disconnected. — apokrisis

This is self-contradictory. How do you know you were thinking "even in deep sleep" if you don/t/can't remember it?

But the real issue here is in what sense do you think that the brain does “data processing”? — apokrisis

Data processing in the sense that we play with ideas - explore associations, logical connections and possibly other things I can't think of right now.

That is fine as a vague metaphor. But the brain isn’t designed to be a Universal Turing Machine - the standard formal definition of data processing. — apokrisis

Well, the way I see it, all that needs to be done is, like the brain, we need to have in place hardware capable of logic and memory. After that, consciousness is simply a matter of feeding such a system with data.

apokrisis

This is self-contradictory. How do you know you were thinking "even in deep sleep" if you don/t/can't remember it? — TheMadFool

You can catch it just as it fades if you are awake quickly enough and are primed to make the effort.

This was discovered in experiments where subjects were woken in slow wave sleep and asked the question. You can notice it yourself but it takes a little practice.

A memory has to be moved from working memory to get fixed as a long term memory. So that is the step that gets shut down during sleep.

Well, the way I see it, all that needs to be done is, like the brain, we need to have in place hardware capable of logic and memory. After that, consciousness is simply a matter of feeding such a system with data. — TheMadFool

Well that is describing the brain in computational jargon. A neurobiologist would want to put quotemarks around all those terms for good reason. They are fundamentally misleading once you pursue the metaphors any distance at all.

TheMadFool

You can catch it just as it fades if you are awake quickly enough and are primed to make the effort.

This was discovered in experiments where subjects were woken in slow wave sleep and asked the question. You can notice it yourself but it takes a little practice.

A memory has to be moved from working memory to get fixed as a long term memory. So that is the step that gets shut down during sleep. — apokrisis

Well, prima facie this means thinking equated to data processing alone is not sufficient for consciousness but take a close look at what's missing - awareness of self, of thinking - and these are, if you really give it some thought, just a different level of data processing. Meta-thinking, if you will.

Wayfarer

↪TheMadFool

But, Mad Fool, you’re simply wrong, mistaken, incorrect. When you think, no bits are processed, no switches are flipped. It’s an entirely different thing. It’s really, to put it crudely, like believing that images in mirrors are capable of independent thought and action. (Now wouldn’t that be mind blowing? You’re straightening your tie and the image turns around and walks off....)

Ever head the saying ‘forgetfulness of being’? It’s associated with some German philosopher. Anyway - you’re evincing it (and not often I get to use that word!)

TheMadFool

But, Mad Fool, you’re simply wrong, mistaken, incorrect. When you think, no bits are processed, no switches are flipped. It’s an entirely different thing. It’s really, to put it crudely, like believing that images in mirrors are capable of independent thought and action. (Now wouldn’t that be mind blowing? You’re straightening your tie and the image turns around and walks off....)

Ever head the saying ‘forgetfulness of being’? It’s associated with some German philosopher. Anyway - you’re evincing it (and not often I get to use that word!) — Wayfarer

I'm not saying brains are computers jot and tittle. I made it clear that working from the fact that consciousness seems to result from logic being applied to ideas, abilities computers possess, there doesn't seem anything undoable about consciousness.

Sorry but I didn't get the "forgetfulness of being" part. :chin:

Wayfarer

↪TheMadFool

It’s because humans are beings - that is our designation ‘human beings’ - and computers are devices. They’re mechanisms, instruments, great banks of switches operated electronically at fantastic speed. But they’re not beings. When asked why not, I find it hard to answer, not because I don’t know, but I find it hard to imagine how someone can ask the question. In other words, if you ask the question, there’s something fundamental about the nature of being that you don’t see. That is what I mean by ‘forgetfulness of being’.

apokrisis

Well, prima facie this means thinking equated to data processing alone is not sufficient for consciousness — TheMadFool

The problem is that data processing is a completely mechanical way of looking at it. The Chinese Room argument blows that out of the water.

The best general theory of mind and life is that it is a semiotic process. A modelling relation.

So there is good news. There is a decent answer now. We don’t have to keep searching or trying to make bad metaphysics fit the known neurobiology.

TheMadFool

It’s because humans are beings - that is our designation ‘human beings’ - and computers are devices. They’re mechanisms, instruments, great banks of switches operated electronically at fantastic speed. But they’re not beings — Wayfarer

Funny that you should say that. Reminds me of speciesism and racism. What should we call this brand of discrimination? Beingism?

To be frank, I'm not saying anything that isn't mainstream science. Astrobiologists are known to entertain the possibility of non carbon-based life with consciousness not being ruled out in any such imagined scenario. The opinion I express here is just a natural extension of this.

In other words, if you ask the question, there’s something fundamental about the nature of being that you don’t see. That is what I mean by ‘forgetfulness of being’. — Wayfarer

:up: What exactly is this "something fundamental about the nature of being" that I don't see?

TheMadFool

The problem is that data processing is a completely mechanical way of looking at it. The Chinese Room argument blows that out of the water.

The best general theory of mind and life is that it is a semiotic process. A modelling relation.

So there is good news. There is a decent answer now. We don’t have to keep searching or trying to make bad metaphysics fit the known neurobiology. — apokrisis

All this reminds me of Leibniz's principle of identity of indiscernibles. If we can't tell apart a person who can speak Chinese and a Chinese Room in the thought experiment then they must be identical, no?

Wayfarer

What exactly is this "something fundamental about the nature of being" that I don't see? — TheMadFool

Reporter (to Louis Armstrong): ‘What exactly is jazz?’

Armstrong: ‘Lady, if you don’t know, I can’t tell you.’

apokrisis

If we can't tell apart a person who can speak Chinese and a Chinese Room in the thought experiment then they must be identical, no? — TheMadFool

Sounds legit.

3017amen

summary, we (well, some) think that human experience includes an extra phenomenal experience that is beyond the computational mechanics. — Malcolm Lett

Computers: a priori
Consciousness: a posteriori and synthetic a priori

TheMadFool

Sounds legit. — apokrisis

The problem:

1. X = Native Chinese Speaker

2. Y = Chinese Room with a person who doesn't understand Chinese following mechanical linguistic rules

3. Me, conversing with X and Y

I can't tell the difference between X and Y. Either X doesn't understand Chinese OR the Chinese Room understands Chinese. We're certain that X understands Chinese. It must be then that the Chinese Room understands Chinese.

:chin:

TheMadFool

Reporter (to Louis Armstrong): ‘What exactly is jazz?’

Armstrong: ‘Lady, if you don’t know, I can’t tell you.’ — Wayfarer

:rofl: :chin: Deepity :smile:

apokrisis

It must be then that the Chinese Room understands Chinese. — TheMadFool

So this thought experiment proved to you that the room understands rather than that the person outside the room had a false understanding that the room understands.

Go for it!

Malcolm Lett

Well, the way I see it, all that needs to be done is, like the brain, we need to have in place hardware capable of logic and memory. After that, consciousness is simply a matter of feeding such a system with data. — TheMadFool

This is pretty much my view too. Almost Dennet-like, I suppose. That in the long run we'll figure out the mechanisms and we'll see all of consciousness as a mechanical process. But I also see the explanatory gap as needing explanation.

I'm reading through Michel Bitbol's It is never known, but it is the knower (thanks @Wayfarer). He claims that scientists naively infer from our past scientific successes that we'll also succeed in explaining consciousness through physical mechanistic principles. I disagree. I think we will eventually explain it as a physical mechanistic process because the majority of evidence is that everything physical in the universe is a physical mechanistic process, and the majority of evidence is that we are physical.

But to my mind, current theorists who propose a mechanical process and claim that it explains everything about consciousness are indeed naive. There is definitely a something needing explaining. Like Bitbol's thesis on the importance of taking subjectivity seriously, any theory on the mechanics behind a subjective conscious experience is incomplete until it explains how the objective mechanics produces the subjective.

Ultimately, like for much of scientific discovery, we'll improve our understanding of "mechanical process" while taking the path towards the nirvana of understanding consciousness.

(BTW, I'm not quoting you because I assume you think that's all there is. It was just a convenient starting point for making my own point)

Malcolm Lett

We're certain that X understands Chinese. It must be then that the Chinese Room understands Chinese — TheMadFool

The problem is the assumption that 'understanding' is binary.

A calculator understands maths in much the same way as the room in the Chinese Room analogy understands chinese. It has some non-negligible understanding of the maths that it's programmed to work with. If it had no understanding, then it wouldn't suffice as a culculator.

We take say that humans "understand" a concept because we build detailed models around that concept. We model not just the end result of how to apply a concept, but also layered theories and explanations. We attach all sorts of context to the concept: how we "feel" about that concept, when/when not to apply it.

All of that can be explained using the same underlying computational processes that the calculator uses.

Is there something 'special' about the human understanding vs the calculator understanding that isn't just a matter of degree? Well, I personally think not, but I'll leave that as an open question for now.

What I will suggest though, is that the word "understand" is socially understood to mean a certain thing only because that's our human-centro definition of it.

apokrisis

That in the long run we'll figure out the mechanisms and we'll see all of consciousness as a mechanical process. But I also see the explanatory gap as needing explanation. — Malcolm Lett

The explanatory gap is what a mechanical conception of nature creates. So “more mechanism” is never going to bridge that gap.

Hence why biologists and neuroscientists are arriving at semiotics as an alternative conception of nature.

Wayfarer

I think we will eventually explain it as a physical mechanistic process because the majority of evidence is that everything physical in the universe is a physical mechanistic process, and the majority of evidence is that we are physical. — Malcolm Lett

I don't think there can be evidence for that. It's a metaphysical attitude, or rather, a methodological postulate that is then interpreted as a metaphysical principle. Modern science has tended to want to see 'everything in the universe' as physical, because physical objects are amenable to the precise objectification and quantification that is central to its method. That was part of the conceptual revolution introduced by Galileo, Newton, and Descartes, among others, at the advent of modern science.

any theory on the mechanics behind a subjective conscious experience is incomplete until it explains how the objective mechanics produces the subjective. — Malcolm Lett

Well, an obvious philosophical criticism of that endeavour, is to point out the sense in which subject and object are mutually dependent. It is grounded in what Kant termed his 'Copernican revolution in philosophy', that 'things conform to thoughts rather than thoughts to things' It is often waved away as 'idealism', except for the inconvenient fact that physics, again, has had to included a reckoning of 'the role of the observer' in its calculations. We've had several discussions on this forum about Wheeler's idea of the participatory universe. Likewise Arthur Eddington's popular science book from between the wars, The Nature of the Physical World, was generally idealist in its orientation. There's an identifiable strain of idealism running through a lot of 20th century physics - Heisenberg, Bohr, Wigner and Schrodinger all wrote about it at some stage in their career. (Bitbol's exposition of Schrodinger's philosophical texts is also good.)

JerseyFlight

I don't think there can be evidence for that. — Wayfarer

???

It's a metaphysical attitude, or rather, a methodological postulate that is then interpreted as a metaphysical principle. — Wayfarer

What on earth? Matter is what we find existing, this is not an "attitude," or "postulate."

Modern science has tended to want to see 'everything in the universe' as physical, because physical objects are amenable to the precise objectification and quantification that is central to its method. — Wayfarer

Did you happen to have an alternative?

Wayfarer

Matter is what we find existing, this is not an "attitude," or "postulate." — JerseyFlight

I wasn’t talking about ‘matter’ but about the postulate that ‘everything is physical’ which is physicalism or scientific materialism, depending on who you talk to. I mean, saying ‘everything in the universe is a physical mechanistic process’ is problematical in light of the hypothesis that what is understood by physics only comprises 4% of the totality of the cosmos, the balance existing in the form of dark matter and energy, about which nothing is known (see Hempel's Dilemma). I simply think science itself has shown materialism to be untenable, but the wider culture hasn’t caught up with that yet.

JerseyFlight

Glad to see you agree that matter exists.

I wasn’t talking about ‘matter’ but about the postulate that ‘everything is physical’ which is physicalism or scientific materialism, depending on who you talk to. — Wayfarer

I am aware, you are talking about a postulate not reality. This is a sophistical game. It's leveraged on the idea that one is making a formal claim about the nature of reality, you then seek to rightly point out the unsustainable absolutism of the claim. All good and well. When I make this argument you are free to counter it as you see fit, but you will not be assigning it to me. What you are here using as a your leverage is the refutation of an abstraction, not reality.

I mean, saying ‘everything in the universe is a physical mechanistic process’ is problematical in light of the hypothesis that what is understood by physics only comprises 4% of the totality of the cosmos, the balance existing in the form of dark matter and energy, about which nothing is known. — Wayfarer

Well, you seem to know a lot about it. Never mind your Fairy-Dust-of-the-Gaps argument here, what specially was your alternative to matter?

Malcolm Lett

The best general theory of mind and life is that it is a semiotic process. A modelling relation. — apokrisis

Yes. That looks promising. I think it offers some useful tools for "measuring" more of the explanatory gap.

Biosemiotics basically says three things:
1) it is not sufficient to define the living world via its physical mechanisms,
2) you also need to consider the 'data' that the mechanisms produce - aka symbols,
3) and the two are intrinsically linked because, as it happens in all dynamic living systems that we are aware of, you cannot have one without the other and still produce the kinds of behaviours that we expect of a dynamic living system.

But what's most useful from that is that it provides a framework for measuring the effectiveness of a system to produce self-referential conscious-like processing capabilities, and its efficiency.

It occurs to me that one way of using semiotics is kind of similar to Tononi's Phi theory, in that it provides a way of characterising different systems - how well does the system follow the circular process of physical mechanics interpreting symbols and creating more physical mechanics from those systems.

In another view, it's a kind of (slightly open ended) anthropic principle applied to the underlying mechanisms of living organisms. In The Necessity Of Biosemiotics Matter-Symbol Complementarity, Pattee explains that living organisms on earth use the particular DNA/RNA processes that they do, because that's what works.

How to measure what remains of the hard problem

Welcome to The Philosophy Forum!

Categories

More Discussions