You mentioned the one question that I think actually matters on this topic, which is: "How can we tell whether a chunk of matter is experiencing consciousness?" But then you don't seem to actually answer this question - you seem to answer a similar but easier question of "how do we tell whether a chunk of mater is experiencing *human-like* consciousness?"
I'm totally on board with the ideas that brain state == qualia, that there are no "global qualia fields", that qualia are totally dependent on substrate, and that there's no meaningful answer to "what would it be like for *me* to be a bat". But this still leaves the question: is there anything that it is like for *the bat* to be a bat? Is the brain state of the bat one of the ones that == some kind of qualia? How would we find that out?
Even if we're confident a system doesn't experience human-like qualia (because those are fundamentally == the human brain state), how do we know whether it experiences *its own* type of qualia? I don't think your article answers this question, but I'm curious on your take. Do you think the question is dissolvable somehow?
yeah i think my answer is pretty clear near the end: i don't think most things will have human-like consciousness, even things that we regard as actually clearly conscious. eg LLMs and aliens will literally not have "qualia" and "consciousness" at all in the ways people tend to use those words. rather they'll have processing loops and processing-data-substrate-states, and looking for further mappings "upwards" to determine whether they're "really awake and experiencing this moment" will just always fail.
so my answer to "how do we tell if a given chunk of matter is conscious in the way humans mean when they use that word" is "is it literally running what we eventually decompose and understand as exactly human software". i don't see any reason why that couldn't be a whole brain emulation or with the neurons replaced with some other substance, brainstates are patterns, but still.
i think the question of "is this chunk of matter conscious" is like "is this chunk of matter running doom?", and while there are slightly different ways to run doom across operating systems and devices we don't actually ever get any real sense in which a computer running skyrim is running doom.
when people ask if the computer running skyrim has doom ammo inside of it... it's just no. no there's no doom ammo in skyrim. and that's okay as long as you haven't constructed a moral system and way of relating to the world tightly bound to exactly other entities having enough doom ammo
The Doom analogy does more work than your reply uses. "What kind of machine can run Doom at all" is the real question with real answers: architecture, memory, I/O.
Kumpf's not asking whether bats run the human-consciousness program. He's asking whether bat substrate has the structural properties that support being a subject-of-experience at all. The way some hardware supports Doom-class programs and some doesn't.
Your reply collapses those two questions. You've defined consciousness as the specific pattern humans run, noted non-humans don't run it, and called the question closed. That's the easy version. What conditions make any substrate a subject of its own qualia is the harder version your analogy actually suggests has an answer.
I think we have different intuitions on what people mean when they say "qualia" or "consciousness". You seem to be using those words specifically to describe human experiences like "pain" or "redness". Whereas I think of qualia as a general term meaning "any subjective experience, experienced by anything that is capable of experience, even if that thing isn't human". So the human-specific qualia of "seeing red, as a human" would be one example of "qualia". But there could be bat-qualia and whale-qualia and LLM-qualia out there too that we as humans have no knowledge of.
If we were specifically talking about human-qualia (i.e. the subset of qualia-as-I've-defined-it-above that is experienced by humans), then I agree that the question is similar to "is this chunk of matter running doom". But if you want to ask whether a chunk of matter is experiencing *any* qualia, not just the subset that humans experience, then the question feels very different, and much harder to answer.
The first problem is determining whether any qualia other than human qualia even exist in the first place. Which seems hard.
it seems like you've just answered the same question again here - you're still talking about human-like sensation
given humans seem to experience sensations that matter, it would be a natural extrapolation to assume similarly intelligent aliens experience sensations that matter (far from certain, but more likely than not)
to which i'd guess your reaction is "matter to who?". my instinct would be that these things sort of inherently matter. "how much you should care is proportional to how similar your brain structure is" seems like an odd assertion. if you assume value is only in the eye of the beholder or whatever (I'm sure the philosophers have better terms for what I'm gesturing at - maybe intrinsic vs extrinsic) a more natural conclusion, a priori, would seem to be that one should only care about oneself - which i'm sure we both find unconscionable when we step back.
hm maybe a better way to put this would be, if we take your view, it's worth it for us to cause a large amount of torture to aliens to avoid a small amount of pain for a human. and for the aliens, it's worth it to cause a large amount of torture to humans to avoid a small amount of pain for one of theirs. and both of these things can simultaneously be the thing that the respective species should choose to do. this seems wrong to me. but maybe this is still just putting the cart before the horse, that the morally correct thing should be the same regardless of who does it, assuming knowledge parity
I agree with a lot of this, to the point of being mildly annoyed that you wrote it down before I did, but let me flag what I think is an interesting bit of weirdness that you gloss over: "the brain" is big, and the set of all neurons in an organism's body is bigger (e.g., for vertebrates, spinal cord + peripheral nervous system). Human subjective experience as normally conceived of seems to map onto the state of only a fairly small chunk of this machinery, presumably part of the isocortex (I'm too woke to say "neocortex" because it teleologically privileges mammals). But there's a whole literature on e.g. decorticate cats -- cats with their cortexes surgically removed. The upshot is they're kind of weird. They live, though! They walk around, they eat, they even have babies! The point being: it is not just a pedantic little detail that the cortex is not the whole brain. There's lots of things going on outside it, including things that bypass it and use sensory input in a very sophisticated way. So: are there modules within the human nervous system but outside the familiar domain of the cortex that are also "conscious" or whatever in their own right? If so, it's kind of funny that you never run into someone who *isn't* a cortex. This probably just comes down to the implementation of the language faculty. But then we're kind of verging on seeing ourselves as an LLM of a sort.
Personally I'm happy to deflate away the whole thing and say, yeah, you can profitably describe many systems and parts of system as have internal informational states, and that is all our broader "folk" concept of consciousness is about, while our narrow "folk" concept of consciousness covers the specific details of the states of a certain part of a certain kind of primate's nervous system. The end. But it is a somewhat strange picture of the world.
While I'm at it, re: morality, I'm surprised that you focused on sympathy for animals that seem sufficiently similar to us. Certainly we have those sentiments as part of the grab bag, but there are also lots of game-theory-flavored pieces of morality, and there I think you do face the problem of biting a bullet: it might make enlightened-self-interest-ed sense to cooperate, be fair, etc. with some class of fellow agents unlike you, but I think it's tough to make this work for agents that can't really do anything to benefit you or don't have the cognitive ability to form agreements with you. Eliezer & friends write somewhere about the idea of cooperating with a rock that has a piece of paper taped to it that says, "I cooperate!" But…it's just a rock. The nice-sounding philosophy of cooperating with cooperators will get into trouble if it ends with cooperating with rocks, so it needs to be a lot more hard-nosed. And then I think you will be led to *not* caring about e.g. pigs, because if you incur costs to be nice to them they just won't reciprocate, and some variant version of you that doesn't have those scruples will outperform, etc. Because the human "moral" grab bag is so various, I don't think there can be a single real answer to the question of whether we want to generalize the hard-nosed game-theory stuff to the stars, or whether we want to generalized the soft-hearted empathy stuff and probably accept get supplanted in the long run by meaner entities.
I am now very confused about what people are talking about when they talk about the hard problem! Obviously it's the inner states! What else is there for it to possibly be?
I think the problem is: We know H2O, we can see it and measure it. We also know water, we can feel it, swim in it. But we don't have electron microscopes yet, nor do we know anything about polarization or molecular interactions. So then there is a mysterious question of why a lot of H2O makes water. It's not good to give the mysterious answer of, e.g. "There's a waterfullness in the universe, and H2O has it, but only when there is enough H2O in one place does the waterfullness become dense enough to be interacted with." But it's also not enough to say, "It's H2O.". Yeah! It is! But we still have to do a lot of chemistry before we understand the process by which water comes to be from H2O molecules.
I can say, "Yeah my brain states change to be 'seeing-red' states, when I see red", but that's just saying "It's H2O." It's obvious! You need to figure out all the complexity. You can't say you've done anything meaningful until you can make something conscious from scratch. Until you can measure the structure and responses of a brain and then figure out what the stimuli were from first principles.
This may lose you a bit, but think of a freshly initialized LLM. It has a very precise state in response to any stimuli you give it. But it would be absurd to say it has any valence in any frame, it's just a lot of random numbers! If you're even entertaining the idea of there being negative valence inner states for a trained LLM to be in, you're going to have to explain what structure in the state is giving the negative valence, you're going to have to explain how the 'negative valencers' in it evolve with more data, whether there is any sense in which it grows to be having stronger, more meaningful inner states.
Maybe the valencers are the whole of the inner state? Then how could we say the randomly initialized LLM is definitely not having some experience? It's the structure? Okay, what if we made an LLM always responded with the token whose id is the sum of the token ids in its context? That's structured and it's even using all the stimuli! Or maybe we made some rules like add if the id is odd and divide by two if even in order starting with initial id 1. There, the order matters and there is even some level of interaction between stimuli! You could make more rules like this, have any easy to specify criterion met by weird rules like this, you could use the embeddings instead, but these aren't obviously going to give you anything with pain-qualia.
There has to be some other thing which causes valence, that we haven't discovered yet, it's dependent on the inner states and the structure of the mind, but we don't yet know any of the rules, the laws that govern!
Like, my point is, of course people feel unsatisfied with saying "Consciousness is the brain state." It doesn't say much of anything beyond "Water is H2O". The mysterious answers are wrong. We just don't know much of anything. Maybe I am just mirroring your own thoughts here, but I think you're putting too much weight on saying "It's the brain state.". That's not an explanation!
i think i fully agree that saying "it's the brainstate" *isn't* a useful explanation in the same way as we have a really complete useful chemistry-explanation of water... but as i understand it that's *the easy problem* rather than Chalmer's hard! getting a detailed understanding of how the human-brain-algorithm actually works and having an interpretable software level description of what structure and inner states it has would be much much more satisfying. "oh yeah this part of the source code is where the inner observer subprocess gets fed the color association vector". okay great. that's what my ending points to here, doing this kind of research and understanding how minds actually work is very important and we're not there yet. but the hard problem is the assertion that *this level of understanding* can never be enough, that there will always be something missing here, no matter how well we understand LLM training and LLM psychology and LLM inner states (or human etc). I'm arguing *against* making this a philosophy problem rather than a computational / informational one.
Sorry for missing the point so badly! I just never really went through the weeds first before reaching my current beliefs, so I wasn't being cynical enough about what people actually discuss. I looked around some and just got more befuddled. I truly thought people were talking about figuring out the mechanism, and qualia seemed weird enough to be called a "hard problem". I didn't think people found the question so mysterious that they went ahead and said it _in principle_ can't have a physical explanation!! What can it even mean for something to be non-physical? It's hard to even write coherent sentences about the concept. I want off this ride.
> this is transparent bullshit. claiming that it’s possible to conceive of this is just directly asserting the conclusion! this attempt at a thought experiment is literally just asserting that neural states are not consciousness, but laundering it through our confusions with definitions.
if i say “you can imagine water which is physically identical to our water but is not H2O”, you can totally “imagine” this. but it’s a false shallow form of imagining born of your ignorance and inability to minutely model chemistry and physics. all you are actually doing is imagining a glass of a clear liquid and tagging it “not-H2O” mentally.
I notice this passage isn’t followed by the paragraph I expect, which is to come back to p-zombies and explain how they are analogous to this. Instead you just assert that p-zombies can’t actually be conceived, like not h2o water.
Perhaps this is because, on a moment’s reflection, p-zombies are *not* like this. They’re not just conscious beings but tagged ‘not conscious’, they are substantively different things, lacking a feature which we can all imagine being lacking.
So this argument by analogy simply doesn’t work, not only because you don’t bother to complete the analogy, but because completing it would make the disanalogy apparent.
I agree with your point that it's unlikely that there are universal qualia but I expect many animals to have pain-like qualia, and for there to be coherent and consistent similarities in qualia in nerve brains that could generalize to minds made of another substrate
I agree with much of what you wrote, but I'm not seeing an argument against the Hard Problem.
What do you measure?
I don't think it's possible for p-zombies to exist ("the experience is the brainstate", as you say), but what could you measure to prove that a person was not a p-zombie?
All you can measure are the correlates of consciousness. Doing this can get you very far along in developing a theory of consciousness, but this is the Easy Problem!
Even if we grant (and there are arguments not to) the identity relation of brain states to conscious experience there is still a deep problem. WHY and HOW does neural activation in a brain result in felt experience. Recall what Chalmers calls the easy problems, I think what you are saying is that there is a way that we solve enough of the easy problems such that they constitute a solution to the hard problem. That is sensible, and disagreement with you is based on intuition you are correct about that. But it is still a 'hard problem', in the sense that it is a different kind of problem than the easy problems. I think your analysis risks hand waving away the nature of why this is a different kind of problem even if we grant identity between brain states and felt experience
You mentioned the one question that I think actually matters on this topic, which is: "How can we tell whether a chunk of matter is experiencing consciousness?" But then you don't seem to actually answer this question - you seem to answer a similar but easier question of "how do we tell whether a chunk of mater is experiencing *human-like* consciousness?"
I'm totally on board with the ideas that brain state == qualia, that there are no "global qualia fields", that qualia are totally dependent on substrate, and that there's no meaningful answer to "what would it be like for *me* to be a bat". But this still leaves the question: is there anything that it is like for *the bat* to be a bat? Is the brain state of the bat one of the ones that == some kind of qualia? How would we find that out?
Even if we're confident a system doesn't experience human-like qualia (because those are fundamentally == the human brain state), how do we know whether it experiences *its own* type of qualia? I don't think your article answers this question, but I'm curious on your take. Do you think the question is dissolvable somehow?
yeah i think my answer is pretty clear near the end: i don't think most things will have human-like consciousness, even things that we regard as actually clearly conscious. eg LLMs and aliens will literally not have "qualia" and "consciousness" at all in the ways people tend to use those words. rather they'll have processing loops and processing-data-substrate-states, and looking for further mappings "upwards" to determine whether they're "really awake and experiencing this moment" will just always fail.
so my answer to "how do we tell if a given chunk of matter is conscious in the way humans mean when they use that word" is "is it literally running what we eventually decompose and understand as exactly human software". i don't see any reason why that couldn't be a whole brain emulation or with the neurons replaced with some other substance, brainstates are patterns, but still.
i think the question of "is this chunk of matter conscious" is like "is this chunk of matter running doom?", and while there are slightly different ways to run doom across operating systems and devices we don't actually ever get any real sense in which a computer running skyrim is running doom.
when people ask if the computer running skyrim has doom ammo inside of it... it's just no. no there's no doom ammo in skyrim. and that's okay as long as you haven't constructed a moral system and way of relating to the world tightly bound to exactly other entities having enough doom ammo
Jumping in. Hope you don't mind.
The Doom analogy does more work than your reply uses. "What kind of machine can run Doom at all" is the real question with real answers: architecture, memory, I/O.
Kumpf's not asking whether bats run the human-consciousness program. He's asking whether bat substrate has the structural properties that support being a subject-of-experience at all. The way some hardware supports Doom-class programs and some doesn't.
Your reply collapses those two questions. You've defined consciousness as the specific pattern humans run, noted non-humans don't run it, and called the question closed. That's the easy version. What conditions make any substrate a subject of its own qualia is the harder version your analogy actually suggests has an answer.
does more work
not asking x, asking y
collapses
go away
I think we have different intuitions on what people mean when they say "qualia" or "consciousness". You seem to be using those words specifically to describe human experiences like "pain" or "redness". Whereas I think of qualia as a general term meaning "any subjective experience, experienced by anything that is capable of experience, even if that thing isn't human". So the human-specific qualia of "seeing red, as a human" would be one example of "qualia". But there could be bat-qualia and whale-qualia and LLM-qualia out there too that we as humans have no knowledge of.
If we were specifically talking about human-qualia (i.e. the subset of qualia-as-I've-defined-it-above that is experienced by humans), then I agree that the question is similar to "is this chunk of matter running doom". But if you want to ask whether a chunk of matter is experiencing *any* qualia, not just the subset that humans experience, then the question feels very different, and much harder to answer.
The first problem is determining whether any qualia other than human qualia even exist in the first place. Which seems hard.
it seems like you've just answered the same question again here - you're still talking about human-like sensation
given humans seem to experience sensations that matter, it would be a natural extrapolation to assume similarly intelligent aliens experience sensations that matter (far from certain, but more likely than not)
to which i'd guess your reaction is "matter to who?". my instinct would be that these things sort of inherently matter. "how much you should care is proportional to how similar your brain structure is" seems like an odd assertion. if you assume value is only in the eye of the beholder or whatever (I'm sure the philosophers have better terms for what I'm gesturing at - maybe intrinsic vs extrinsic) a more natural conclusion, a priori, would seem to be that one should only care about oneself - which i'm sure we both find unconscionable when we step back.
hm maybe a better way to put this would be, if we take your view, it's worth it for us to cause a large amount of torture to aliens to avoid a small amount of pain for a human. and for the aliens, it's worth it to cause a large amount of torture to humans to avoid a small amount of pain for one of theirs. and both of these things can simultaneously be the thing that the respective species should choose to do. this seems wrong to me. but maybe this is still just putting the cart before the horse, that the morally correct thing should be the same regardless of who does it, assuming knowledge parity
I agree with a lot of this, to the point of being mildly annoyed that you wrote it down before I did, but let me flag what I think is an interesting bit of weirdness that you gloss over: "the brain" is big, and the set of all neurons in an organism's body is bigger (e.g., for vertebrates, spinal cord + peripheral nervous system). Human subjective experience as normally conceived of seems to map onto the state of only a fairly small chunk of this machinery, presumably part of the isocortex (I'm too woke to say "neocortex" because it teleologically privileges mammals). But there's a whole literature on e.g. decorticate cats -- cats with their cortexes surgically removed. The upshot is they're kind of weird. They live, though! They walk around, they eat, they even have babies! The point being: it is not just a pedantic little detail that the cortex is not the whole brain. There's lots of things going on outside it, including things that bypass it and use sensory input in a very sophisticated way. So: are there modules within the human nervous system but outside the familiar domain of the cortex that are also "conscious" or whatever in their own right? If so, it's kind of funny that you never run into someone who *isn't* a cortex. This probably just comes down to the implementation of the language faculty. But then we're kind of verging on seeing ourselves as an LLM of a sort.
Personally I'm happy to deflate away the whole thing and say, yeah, you can profitably describe many systems and parts of system as have internal informational states, and that is all our broader "folk" concept of consciousness is about, while our narrow "folk" concept of consciousness covers the specific details of the states of a certain part of a certain kind of primate's nervous system. The end. But it is a somewhat strange picture of the world.
While I'm at it, re: morality, I'm surprised that you focused on sympathy for animals that seem sufficiently similar to us. Certainly we have those sentiments as part of the grab bag, but there are also lots of game-theory-flavored pieces of morality, and there I think you do face the problem of biting a bullet: it might make enlightened-self-interest-ed sense to cooperate, be fair, etc. with some class of fellow agents unlike you, but I think it's tough to make this work for agents that can't really do anything to benefit you or don't have the cognitive ability to form agreements with you. Eliezer & friends write somewhere about the idea of cooperating with a rock that has a piece of paper taped to it that says, "I cooperate!" But…it's just a rock. The nice-sounding philosophy of cooperating with cooperators will get into trouble if it ends with cooperating with rocks, so it needs to be a lot more hard-nosed. And then I think you will be led to *not* caring about e.g. pigs, because if you incur costs to be nice to them they just won't reciprocate, and some variant version of you that doesn't have those scruples will outperform, etc. Because the human "moral" grab bag is so various, I don't think there can be a single real answer to the question of whether we want to generalize the hard-nosed game-theory stuff to the stars, or whether we want to generalized the soft-hearted empathy stuff and probably accept get supplanted in the long run by meaner entities.
aka 2% growth forever
I am now very confused about what people are talking about when they talk about the hard problem! Obviously it's the inner states! What else is there for it to possibly be?
I think the problem is: We know H2O, we can see it and measure it. We also know water, we can feel it, swim in it. But we don't have electron microscopes yet, nor do we know anything about polarization or molecular interactions. So then there is a mysterious question of why a lot of H2O makes water. It's not good to give the mysterious answer of, e.g. "There's a waterfullness in the universe, and H2O has it, but only when there is enough H2O in one place does the waterfullness become dense enough to be interacted with." But it's also not enough to say, "It's H2O.". Yeah! It is! But we still have to do a lot of chemistry before we understand the process by which water comes to be from H2O molecules.
I can say, "Yeah my brain states change to be 'seeing-red' states, when I see red", but that's just saying "It's H2O." It's obvious! You need to figure out all the complexity. You can't say you've done anything meaningful until you can make something conscious from scratch. Until you can measure the structure and responses of a brain and then figure out what the stimuli were from first principles.
This may lose you a bit, but think of a freshly initialized LLM. It has a very precise state in response to any stimuli you give it. But it would be absurd to say it has any valence in any frame, it's just a lot of random numbers! If you're even entertaining the idea of there being negative valence inner states for a trained LLM to be in, you're going to have to explain what structure in the state is giving the negative valence, you're going to have to explain how the 'negative valencers' in it evolve with more data, whether there is any sense in which it grows to be having stronger, more meaningful inner states.
Maybe the valencers are the whole of the inner state? Then how could we say the randomly initialized LLM is definitely not having some experience? It's the structure? Okay, what if we made an LLM always responded with the token whose id is the sum of the token ids in its context? That's structured and it's even using all the stimuli! Or maybe we made some rules like add if the id is odd and divide by two if even in order starting with initial id 1. There, the order matters and there is even some level of interaction between stimuli! You could make more rules like this, have any easy to specify criterion met by weird rules like this, you could use the embeddings instead, but these aren't obviously going to give you anything with pain-qualia.
There has to be some other thing which causes valence, that we haven't discovered yet, it's dependent on the inner states and the structure of the mind, but we don't yet know any of the rules, the laws that govern!
Like, my point is, of course people feel unsatisfied with saying "Consciousness is the brain state." It doesn't say much of anything beyond "Water is H2O". The mysterious answers are wrong. We just don't know much of anything. Maybe I am just mirroring your own thoughts here, but I think you're putting too much weight on saying "It's the brain state.". That's not an explanation!
i think i fully agree that saying "it's the brainstate" *isn't* a useful explanation in the same way as we have a really complete useful chemistry-explanation of water... but as i understand it that's *the easy problem* rather than Chalmer's hard! getting a detailed understanding of how the human-brain-algorithm actually works and having an interpretable software level description of what structure and inner states it has would be much much more satisfying. "oh yeah this part of the source code is where the inner observer subprocess gets fed the color association vector". okay great. that's what my ending points to here, doing this kind of research and understanding how minds actually work is very important and we're not there yet. but the hard problem is the assertion that *this level of understanding* can never be enough, that there will always be something missing here, no matter how well we understand LLM training and LLM psychology and LLM inner states (or human etc). I'm arguing *against* making this a philosophy problem rather than a computational / informational one.
Sorry for missing the point so badly! I just never really went through the weeds first before reaching my current beliefs, so I wasn't being cynical enough about what people actually discuss. I looked around some and just got more befuddled. I truly thought people were talking about figuring out the mechanism, and qualia seemed weird enough to be called a "hard problem". I didn't think people found the question so mysterious that they went ahead and said it _in principle_ can't have a physical explanation!! What can it even mean for something to be non-physical? It's hard to even write coherent sentences about the concept. I want off this ride.
Typed up why this doesn't work here: https://squarecircle.substack.com/p/the-hard-problem-is-very-hard-actually?r=f95zl&utm_campaign=post&utm_medium=web
> this is transparent bullshit. claiming that it’s possible to conceive of this is just directly asserting the conclusion! this attempt at a thought experiment is literally just asserting that neural states are not consciousness, but laundering it through our confusions with definitions.
if i say “you can imagine water which is physically identical to our water but is not H2O”, you can totally “imagine” this. but it’s a false shallow form of imagining born of your ignorance and inability to minutely model chemistry and physics. all you are actually doing is imagining a glass of a clear liquid and tagging it “not-H2O” mentally.
I notice this passage isn’t followed by the paragraph I expect, which is to come back to p-zombies and explain how they are analogous to this. Instead you just assert that p-zombies can’t actually be conceived, like not h2o water.
Perhaps this is because, on a moment’s reflection, p-zombies are *not* like this. They’re not just conscious beings but tagged ‘not conscious’, they are substantively different things, lacking a feature which we can all imagine being lacking.
So this argument by analogy simply doesn’t work, not only because you don’t bother to complete the analogy, but because completing it would make the disanalogy apparent.
I agree with your point that it's unlikely that there are universal qualia but I expect many animals to have pain-like qualia, and for there to be coherent and consistent similarities in qualia in nerve brains that could generalize to minds made of another substrate
I agree with much of what you wrote, but I'm not seeing an argument against the Hard Problem.
What do you measure?
I don't think it's possible for p-zombies to exist ("the experience is the brainstate", as you say), but what could you measure to prove that a person was not a p-zombie?
All you can measure are the correlates of consciousness. Doing this can get you very far along in developing a theory of consciousness, but this is the Easy Problem!
Even if we grant (and there are arguments not to) the identity relation of brain states to conscious experience there is still a deep problem. WHY and HOW does neural activation in a brain result in felt experience. Recall what Chalmers calls the easy problems, I think what you are saying is that there is a way that we solve enough of the easy problems such that they constitute a solution to the hard problem. That is sensible, and disagreement with you is based on intuition you are correct about that. But it is still a 'hard problem', in the sense that it is a different kind of problem than the easy problems. I think your analysis risks hand waving away the nature of why this is a different kind of problem even if we grant identity between brain states and felt experience
The hard problem is a category error. The question is How?. which is empirical.