Greetings!
This has taken too long to make, and I’d like to polish it more, but it’s time to release and move on to other things. Apologies for not posting the text here, I was having formatting trouble.
Yes, I attempt to solve/explain consciousness. The main document is around 13,000 words long, which takes an hour to skim-read (not recommended) but probably 2+ hours to read correctly.
Let me know what you think!
Edit 9/9/23: Of the criticisms I’ve seen of EC (both here and elsewhere), none have struck me as having passed my ITT, and/or there is frequently some selective amnesia on display (or more probably, people skim-read the document) regarding points I explicitly take great effort to address. One might wonder why I largely haven’t engaged with EC’s critics, since consciousness can correctly be considered the most important feature of the universe; important to get right. But in light of the solution to C, comparatively little actually hinges/depends on C, or on what people think of it. When combined with the subtle conceptual nature of the problem itself (lossless communication is difficult and idiosyncratic), these factors have lead me to largely not deem it worth my energy to engage. If someone had an actually novel or ITT-passing criticism, I would be in a better position to engage. Lacking that, the most compact and well formulated memetic package on offer is still the document itself. If I was invested in any given critic “getting it”, then my engagement would largely look like me insisting they re-read the document with greater care, and/or I would go through it line by line with them, to combat the amnesia problem and actually find where they go off the rails. That kind of effort doesn’t scale.
It seems that “existence is consciousness” is doing all the lifting here. If we weren’t software programs, we would still exist and be conscious.
And I am not clear about what motivates the idea that we are software. Software/Code doesn’t really have an ontological existence separate from matter/hardware, it’s more of a stance or abstraction. With the hard problem clear, there are still some fundamental premises that need to be established. These could be argued at length, but are justified here only briefly:
You have not put forward a reductive explanation of consciousness, because there is no physical reason that every existing thing should just be conscious.
What is hard about the hard problem is the requirement to explain consciousness, particularly conscious experience, in terms of a physical ontology. Its the combination of the two that makes it hard. Which is to say that the problem can be sidestepped by either denying consciousness, or adopting a non-physicalist ontology.
Examples of non-physical ontologies include dualism, panpsychism and idealism . These are not faced with the Hard Problem, as such, because they are able to say that subjective, qualia, just are what they are, without facing any need to offer a reductive explanation of them. But they have problems of their own, mainly that physicalism is so successful in other areas.
Yes, talk about software is mostly irrelevant.
Сonsciousness is reduced to existence by equivalence. And existence is a part of physical ontology. That’s how all reductions work—there is no “physical reason” that sound is a moving air—it’s just that we had one model and now we have a better one.
No, reductive explanations have to show the properties of the level phenomenon vary in a detailed way with the details of reduction base. A reductive explanation of consciousness would result in being able to to predict qualia from neural activity.
Reductive explanations result in identifications, but not all identifications are reductive explanations.
The level of detail depends on the phenomenon and in the case of hard-problem-consciousness there just not much detail on the physical side. So the only requirement is correctness, not the level of detail.
Predicting specific qualia is mostly separate problem, but… it was never in doubt that qualia are predictable? You don’t even need neural activity—you can just predict that you will see the sky if you look at the sky. Developing high-level model of qualia recognized by humans may be interesting, but it’s not the hard problem—once you figured that consciousness is existence, it’s not mysterious or conceptually difficult to predict qualia because you can reuse physicalist framework you used to predict all other things.
I see no reason to suppose that.
But the human brain is often cited a the most complex thing in the known universe.
That’s why it’s unsurprising that the solution doesn’t involve brain much.
The reason I was thinking of is the difference in naturallness of the phenomenon when translated into physical ontology. If your question is “what’s light?” you can point to the electromagnetic field term. If you want to predict the exact picture on your screen from first principles, you’ll need the electromagnetic field but also additional unrealistic amount of work. Yes, reductionist explanation should provide reasons why it’s compatible with non-reduced phenomenon. But that’s done—it solves conceivability argument by substituting “existence” for “consciousness”, it explains why apparent unity of experience is not precisely true, it provides explanations for why we can extrapolate/isolate solution in many concrete example problems. And you didn’t specify any concrete aspects of consciousness that you think are not derivable from physical description given consciousness=existence. Like, what step of justification you’re unconvinced by? The only reason for prediction from neural activity to matter is if you think physicalism does not work for physics. But that is even not how it works in the zombie world.
Huh?
I mentioned the hard problem , which is literally defined in terms.of qualia.
I’m missing the place where it explains consciousness. The crux of the argument seems to be this:
But the hard-questioner question is still unanswered. What makes this software conscious, when none of the software we make ourselves is conscious? (Or if you take Blake Lemoine seriously, what distinguishes conscious software from non-conscious?) All software “exists”, separately from the existence of the physical machines that run it, in exactly the same sense that the hypothetical brain software “exists” separately from the brain’s physical existence.
The thesis is that every software we make ourselves is conscious. It just has different qualia.
That would be an example of a common fallacy in explanations of consciousness:
Where there is consciousness, we also find X.
Therefore X is consciousness.
But X is present in other things as well.
Therefore those other things are “in some way” conscious.
Or more briefly:
All A’s are B’s.
Therefore “in some sense” all B’s are A’s.
To be clear, the argument for “consciousness is existence” is not “consciousness is software, therefore every software is conscious”. The argument is that it solves zombies without epiphenomenalism while preserving casual closure. You still can map it to “All A’s are B’s”, but then it’s only a fallacy in the same sense that predicting sound in steel from “all sound is waves” or something is a fallacy—you just can’t use purely logical arguments for empirical questions.
It would be bad argument if there was any aspect of consciousness that contradicts it being equivalent to existence. But there isn’t.
Lots of things exist but are not conscious. Therefore consciousness is not existence.
Why do you think they are not conscious? For every property you may name as a reason, there is a conceivability argument that this property doesn’t change for you in the zombie world, but it still loses your consciousness.
There is no conceivability argument for zombies, nor for anything else. The less I know, the more I can conceive, because I don’t know enough to know that my conception is incoherent. I can conceive that the millionth digit of pi is any given number, but only one of the things that seem possible to me is actually possible. What I can conceive is a fact about me, not about the world.
First of all I don’t understand how someone who doesn’t believe in conceivability would consider the hard problem to be hard. If you can specify that the thing you are talking about is, for example, the thing that makes you talk about consciousness, then the problem is obviously solvable—you just trace physical causes to the source—and zombie world is obviously incoherent under such definition of consciousness because you can’t both have and don’t have relevant neural activation patterns. But if you still wonder why that whole thing is not like your unconscious parts, if you think about the concept of having subjective experience by itself, and if your interest in explanations of consciousness is motivated by such mental processes, then that already is acceptance of importance of conceivability of a difference between conscious and unconscious. Like, it doesn’t makes sense to ask “What makes this software conscious?” without conceivability motivation—you already know everything if you know which software you are talking about.
But of course it’s fine to only care about details of implementation of some specific informational processes not present in simple things. And only use the word “consciousness” in this way, and ignore the difference between our world and zombie world on the grounds that epiphenomenalism is unlikely and so conceivability argument must be incoherent in some way we may discover in the future. It’s just then “the difference between zombies and reality is that zombies don’t exist” is still correct fact at least about common way people think about consciousness, and it answers how conceivability argument is incoherent right now.
And why do you actually think they are not conscious?
I’m not sure I can make out what you are saying. What follows is an attempt at a reply, but however accurately I write it, it might miss the mark it should be aimed at.
I have an experience of my own presence. Even in dreams I am present. It is only absent when I am in dreamless sleep. There is something that it is like to be me. Perhaps not everyone experiences this, but I do. This is the thing that I am pointing at when I speak of my own consciousness.
Other people seem outwardly to be the same sort of thing as myself, and they talk of their inward nature in ways that are similar enough to my own that it seems reasonable to believe that there is something that is it like to be them. However, while I believe they are conscious, I have no reason to think that their consciousness is identical to mine but for being placed in a different body. In fact I suspect that people’s consciousnesses, people’s selves, vary far more than their outward appearances, are far more different than it ever occurs to most people that they might be.
But you cannot see my mind and I cannot show it to you. I cannot see your mind and you cannot show it to me. The words that anyone uses to talk about their consciousness point to a place that none but they can see. This makes it difficult to know that what one person is pointing to when they use a word such as “happy” bears much similarity to what another person is pointing to when they use the same word of themselves. If I try to understand what someone else says of themselves by imagining what I would mean if I said the same words about myself, I might be doing the equivalent of trying to find my way through Berlin with a map of London. There are certain resemblances between large cities, but every detail will be different.
Animals cannot talk to us, but their behaviours suggest to me that the large ones do have something of the same sort, although simpler, with less capability of being aware of themselves, and more different from us than we are from each other. Earthworms, I don’t know or care. Rocks, definitely not.
These are the observations that a theory of consciousness must explain.
However, nothing in the entire development of science or philosophy has produced even a coherent hypothesis of what could account for the existence of experience. We have no idea of how there can be any such thing as consciousness. But there it is, as ineluctable as tinnitus.
That is the Hard Problem of Consciousness. Denying it and explaining it both seem impossible. We cannot see what an explanation could even look like.
The Hard Problem is a basilisk that drives mad most people who stare at it.
Some take the impossibility of the Hard Problem as proof that there is no such thing, like a student awarding themselves full marks on an exam because they had no idea how to do any of the questions.
Some engage in intensive meditation practices with the aim of extirpating their sense of themselves, like Winston Smith expunging the past in the Ministry of Truth.
Some say that there must be a physical explanation, and think that they have thereby solved the problem. (I think that this is what you are doing when you say “If you can specify that the thing you are talking about is, for example, the thing that makes you talk about consciousness, then the problem is obviously solvable—you just trace physical causes to the source.” You seem to be assuming that “the thing that makes you talk about consciousness” is a noun phrase that references a physical thing, which begs the question. The word “just” papers over the fact that no-one has “trace[d] physical causes to the source”.)
Some latch onto some physical phenomenon that seems to accompany consciousness and say that that must be consciousness. Then they go on to fail the Wason Selection Task by attributing consciousness to everything else that exhibits the chosen physical phenomenon, instead of noticing at the start that the presence of the chosen phenomenon where we have no reason to think consciousness exists is evidence against the theory, not in support of it.
Some say the Hardness proves there can be no answer.
Some say, therefore God.
I only observe the basilisk, recognise that I have no answer but do not insist that there cannot be an answer, and pass on.
After all of these, the Hard Problem remains untouched.
I’m saying that admitting the possibility that the explanation would not look like dataflow diagram for the brain or something is isomorphic to believing in conceivability argument.
That’s the part I don’t understand in combination with driving yourself mad. Standard physics explains all these observations. We already know why rocks don’t have dogs’ behavior. We already know that the brain in dreamless sleep state is different from the brain in awake state. Yes, not with all details. But why would that matter? Does someone actually expects that tracing physical causes to the source is impossible for the sleep/awake difference? If you feel now that “We have no idea of how there can be any such thing as consciousness”, why would you feel otherwise with any physical explanation?
The only part that needs something except usual physical explanation is why there is any subjective experience at all. Because everything else—why rocks are different from dogs, why red is different from blue—is obviously explainable by physics. What’s there to be mad about? And that part—only that part—is explained by consciousness being equivalent to existence. There is a basilisk only because people think about two different things when they think about their subjective experience: existence and boring differences in informational/physical structure. And only part about existence is driving them mad. It’s because of existence people think that you can’t show someone your mind—it’s just confusion between being in a state and knowing about a state.
So, the frame is that initial observations about consciousness-suggesting behavior are irrelevant (well, except dreamless sleep part—that’s good enough model of non-existence), because they point to the easy part. After you solve the hard part, it’s all just straightforward neuroscience. And then you would be able to see someone’s mind the same way you see any other high-level object. Observations are not contradicted by this solution—there are still difference between rocks and dogs. It’s just that when you are saying “This is the thing that I am pointing at when I speak of my own consciousness” you either pointing to a thing with two components and only one of them is hard/confusing/mysterious and everything has that part. Or you are pointing to just the easy part and there is no hard problem for that thing.
That puzzle is what makes the problem Hard. I have no idea what an explanation could look like.
Indeed. That is exactly the Hard Problem. I experience things. I’m quite sure that rocks do not. Nobody knows why this is so.
But we already have an explanation for the difference between you and rocks—rocks don’t have brains. It was never a requirement for the solution to the hard problem to explain the difference between rocks and you, because we always had an explanation for that. And you are more sure that you experience things than that the rocks don’t—it’s only inference that the rocks don’t have experience and it’s an observation that you have. So the theory that the difference between rocks and you is separate from the question of you having experience is consistent with the observations. That’s what solution to the puzzle looks like—you experience things because you exist.
Do you at least agree that without the requirement that the solution should say that rocks don’t have anything that makes you experience things, equivalence between consciousness and existence provides consistent solution to the hard problem that doesn’t contradict observations or ethical preferences?
That is not much of an explanation. All of the things I attribute consciousness to have a brain. Observations on people with identifiable brain-damage show impairments of consciousness of different sorts, that can be mapped to different parts of the brain. This strongly suggests that the brain is not merely a channel for some separate consciousness, but somehow (and we have no idea how this is even possible) is actually doing consciousness.
But it appears that not all parts of the brain do this. In humans the cerebellum seems not to. Damage to the cerebellum, as far as I know, will disrupt motor control but does not seem to disrupt consciousness.
So which neural tissues are capable of “doing consciousness”, and how do they do it? Nobody knows. Are jellyfish conscious? They have no brain but they do have a diffuse nervous system. Is that enough for consciousness? Are neural organoids conscious? Does it matter what creature they were cultivated from? A real explanation of consciousness would be able to answer these questions. A real explanation must predict what is conscious and what is not, and provide a way to test these predictions. Finding such an explanation, making such predictions, providing a way to test them, testing them, and finding them confirmed would be in the top rank of Nobel-level work. I am not expecting to see such a thing for the first time on LessWrong.
That is a hard requirement. I have no reason to think that rocks have experience, and as I said above, an explanation must account for the negative examples as well as the positive.
Electrons also exist. Quarks exist. Electromagnetic fields exist. The integers exist. What problem is solved by saying that “existence is experience”? This is an empty verbal formula that does not constrain anticipations of future observations.
A lot of this is basic epistemology and independent of the particular subject. “Consider both positive and negative examples.” “Constrain anticipations of future observations.”
The hard problem. The problem that people say that they have no idea what an explanation for the consciousness could look like. The point is not that it solves everything that is interesting about the difference between cerebellum and other parts of the brain. It just solves the part where we have no idea how it’s possible that the brain does consciousness. When that part is solved, we know how it’s possible that the brain does consciousness—it processes information and all our intuitions about our consciousness is compatible with what we know about how the neurons in the brain work. Without the solution for the hard part you could always ask “but why that particular information processing is not non-conscious”. But when the hard part is solved, you always has the answer—it’s conscious because everything that exist is conscious. It provides justification for applying basic science to consciousness. And when you just doing science, you can account for any negative examples—rock example is just explained by it not having a brain or whatever substructure you are interested in instead of not having the thing that makes consciousness hard.
The reason to believe, that rocks have the part of experience that makes it hard to explain, is that believing it makes the problem of explaining differences between jellyfish and rocks just Nobel-level work, but not “we have no idea how the explanation could even look like”-level work. You don’t lose ability to distinguish different information-processors by accepting it. You just lose justification to say that we don’t know even in principle how one bunch of neurons can make human say “I’m conscious” while other bunch of neurons can’t. And it constrains your anticipation. It just constrains it to the physicalist worldview—you were supposed to be constrained to it anyway, but if people still feel puzzled, then actual solution to the hard problem should solve it.
So, if you consider that worldview where you accept that everything is conscious, do you agree that it allows you to calmly enumerate physical differences between jellyfish and awake human and expect that when you’re done, you have the whole answer? Because if consciousness is existence, physical differences are differences in experience.
That is not a good reason for believing anything. It is wishful thinking.
I don’t even understand what “consciousness is existence” is supposed to mean. How does it constrain my anticipation of future experiences? How do I discover what it is like to be a rock? This is just the “A’s are B’s, therefore B’s are A’s, if we redefine A to mean B” fallacy. As well say “rocks exists, therefore everything that exists is a rock.”
I don’t think this conversation is going to make further progress.
I’m more optimistic about usefulness of unreasonably long repetitive conversations^^.
It’s only wishful thinking if you contradict or ignore observations. Otherwise it’s better, more useful interpretation. I would understand if someone argued against usefulness when they already don’t think that the hard problem is hard. But at the stage where you don’t know how an explanation would look like, factoring one complex unexplainable phenomenon into two explainable is useful.
Then what did you disagree with about rocks being conscious? It’s supposed to mean that there is something that it is like to be a rock. You know that, obviously, because rock exists. You discover what it is like to be a rock by doing geology or whatever. And it will give you relevant answer, because “what it is like to be a rock” is equivalent to “what is a physical description of a rock” (because physical description specifies in what shape rock exists). And if you want to constrain rock representation somehow, for example if you want your model of a rock to be implemented with your imagination, you can imagine your dreamless sleep and it would be less precise model, but it would faithfully represent absence of thoughts in a rock.
Or if you want an intuitive motivation, then you can imagine yourself having all your experiences with all your consciousness as some magical inner fire. And then you disassemble yourself into protons and neutrons and make a rock out of them. And then you reassemble yourself from the rock. If your inner fire is an actual thing that exists, then it couldn’t have disappeared when you were a rock and then reappeared again—it would be too magical. So to preserve casual continuity between you having consciousness before and after, the rock in between must also posses something consciousness-like. And it would be too bad if our world was not even magical enough to have any thing like this inner fire, but it does have such thing—existence.
It’s more “A‘s are A1‘s and A2‘s (and we don’t know of a case where A1 and A2 is not A) and we know how to work with A1’s and we can live with things being A2”—no one is saying that rocks have brains or forgetting that they don’t.
I personally think it’s easier to recognize it as a solution if you explicitly talk about zombies instead of “why you don’t know how explanation would look like”. But still, the way I see it, there is no evidence against the theory that people think about two things when they point to consciousness. Like, why do you even expect reformulation of the problem to give new predictions and why do you expect something more than reformulation is needed to transform the Hard problem to a normal problem? Is it just me calling it “solution” instead of “useful reformulation that makes the problem non-confusing”?
There again is our fundamental disagreement, or talking past each other, or whatever is going on. I claim that a rock has no subjective experience. In that sense (I claim) there is “nothing it is like to be a rock”. But you seem to be taking the phrase “what (if anything) it is like to be a rock” to mean the physical properties that we can measure about rocks, the ordinary things that geologists study: “what a rock is like”. That is something completely different.
Does a rock have hopes and fears? Does a rock experience its slow disintegration by erosion in the way a human might experience dementia? Does a rock feel pain when water freezing in cracks splits it apart? Is that like the pain of childbirth as it splits into two daughter rocks? I am confident in saying no to all of these.
Reformulations of the Hard Problem only push the bump in the carpet around.
Fantastical situations make poor intuition pumps. At that point, I am dead. I cannot reassemble myself. Neither, at present, can anyone else. We cannot even revive corpses stored in liquid nitrogen. If we could, mere intuition cannot answer the question of whether my consciousness was hanging around in limbo for the interim, or if the new I is “really” the old I, and so on. If we explicitly make the physicalist assumption that consciousness is literally a physical activity performed by atoms (but in a manner we know nothing about beyond the crude localisation evidence from brain lesions and brain scans), then reanimation, whether from rock or corpsicle form, is re-creation of a consciousness isomorphic to the one before.
Personally, I regard the physicalist assumption as plausible speculation rather than anything proven. The fact that I do not have any competing speculation does not require me to believe this one. Making that assumption does not solve the Hard Problem. It only promises that a solution exists but leaves us just as much in the dark about what that solution is. People have not even solved the simpler question of how to unify quantum mechanics with general relativity. Belief that there must be a solution is not itself a solution.
That’s good! Before I wasn’t sure whether you claim this, or just claim that there is no reason to think otherwise.
I agree. But it doesn’t follow that it doesn’t have any kind of experience.
Not sure if I managed to communicate how it supposed to work, so just to clarify: it’s not something different under the assumption that consciousness is existence. If consciousness is existence, if what makes hopes and pain consciousness-like is what rock also has, then the ordinary things that geologists study are about consciousness.
And if you argue against the assumption, say that these things must be different, then why? Your observations are about behavior and that’s accounted for by physical description. You only have an intuition about rocks not having experience and if this intuition is based on you not being able to imagine what it is like to be a rock, then I could say that you imagining what it is like to be you is not the most precise model of your consciousness either—you can’t say at what exact time your experience changed from seeing red to seeing blue.
Where is the bump in “consciousness is existence”? That rocks are consciousness? Then yeah, the point is that having consistent non-confusing worldview is worth accepting that bump.
That doesn’t make any sense without additional reasons to disbelieve the physicalist assumption in case of consciousness. You don’t disbelieve it in other unsolved cases, like precise aerodynamics of a branch of a tree, right? So even if you don’t want to call it a solution, if you actually believed the physicalist assumption, it would make the problem easy. And, like, ok, I’m not saying you don’t have reasons to be wary about the physicalist assumption, just the whole shape of the “solution must solve everything” argument is wrong.
To summarize my general point so far: the tradeoff is between your intuition about rocks not being conscious and making the Hard problem not hard. And I think you would agree that it’s acceptable if you believed in the later part. Am I wrong and you already agree with the later part and just really can’t stomach conscious rocks?
My intuitions, and other people’s, about what looks conscious and what does not are the starting point, but not the end point, for any investigation of what consciousness is. This is the phenomenon for which we seek an explanation.
If we actually managed to get some distance towards that goal, perhaps we might find that the underlying phenomenon, whatever it is, is also present in places where we do not attribute consciousness. This would be like the discovery that the process underlying fire is reaction with oxygen, and that this is also present in rusting iron and animal metabolism. While we may analogically speak of these as “fire”, what is important is the underlying unity that was discovered among these processes. The “fire” of rusting iron is a very different sort of thing from a burning building, and most things are not fire at all.
But regarding consciousness, all is speculation. We have made no progress down that road. We cannot even find the road. “Consciousness is existence” is a verbal formula that does not constrain anticipations of future observations, any more than “all is fire” did. We now know from observation and experiment that all is, in fact, not fire, or even oxidative processes.
How might we discover the hypothetical consciousness of rocks, rather than trying to define it into existence or passing off speculation as progress? Compared with, for example, the sorts of observations that people are currently making about bees, crows, plants, and so on, on the basis of which some are attributing consciousness to those things?
I am in fact sceptical of a lot of these attributions. If a machine that we currently (i.e. up to 2022, not a movable feast) know how to make, without even using AI techniques, can duplicate an observed behaviour, that is to me good evidence that consciousness, whatever it is, is not involved.
The kind of hypothetical consciousness that is just a generalisation of reasons for results of observations of bees and crows can be discovered in a rock by inspecting its internal structure and finding there a computer roughly isomorphic to a neural system of a crow. That obviously won’t happen, because rocks don’t have such consciousness in natural encoding.
But that’s not the problem that needs solving. We are not in the same situation as with fire: with fire, once we had a theory that explained oxidative processes, we didn’t say “but why oxidative processes are fire?”. We already knew what phenomenon we want to be explained. We always have been on a road. Not being able to even find a road is not a normal situation you solve by more precise physical description. We already have enough physical theories to be far along the road. The only thing that needs solving here is confusion about where on the road we are. The reasons for thinking that we don’t mostly understand consciousness are not in the category of empirically proven scientific theories or methods. They are in the category of speculations and definitions and verbal formulae. So why do you think anything else is required to make problem easier, when your intuition of hardness of the Hard problem is based just on you not knowing how explanation is possible? Or what are your different reasons for thinking that we’ve made no progress, when we know so much about brains and computers?
It’s not just that I don’t know how to explain consciousness. I believe no-one does, including the people who think they do. None of the purported explanations bridge the gap between the stuff we have found out and the question of why there is any such thing as subjective experience. What we know about brains at best only tells us about where consciousness happens. It does not tell us why there is any such thing or how it works. What we know about computers tells us nothing.
Are you conscious? Are you aware of your own presence? Aware of having that awareness?
Yes.
Sure, sometimes.
But why do you believe it? There is no gap under “consciousness is existence” explanation—the stuff we have found out is about subjective experience, because the stuff we have found out is about how things exist and existence is consciousness. There is not much “why” here, admittedly, but that’s expected because we are talking about existence? I don’t get what you don’t like about this explanation, if it’s not “it says rock are conscious”. “Don’t rely on pure verbal manipulations” is just heuristics—can you show some concrete place where there is a gap?
How “if I tap this brain region, you will feel pain” is not telling how it works? Computers are probably not that good a point by themselves, but what I meant was something like they provide us with examples of many capabilities we associate with our consciousness—you can code an agent with attention and access to it’s own state as an input and it would work similarly to the brain (in some way) when you ask it whether it’s aware of it’s own presence.
That it says that everything is conscious is part of it, but that is just a symptom of the real problem: it is not an explanation. It makes no predictions about what we might observe. It describes no moving parts, no mechanism for how everything is conscious. It only leads us to say “everything is conscious!” but nothing else. We will carry on grinding up rocks for concrete without caring about their alleged consciousness.
If I tap this organ key, a pipe will sound. That tells me nothing about whether there is a direct mechanical linkage from the key to the pipe, or the key is closing a switch that operates an electrical relay, or the pipes are just for show and the sound is electronically generated. Input here, output there, tells nothing about how the thing is done. No amount of speculating about how the thing might be done will answer the question of how it is done.
Ok, what observations does your belief, that no one has an explanation for consciousness, predicts? Or that there is always a gap. What does it even mean for something to be an explanation of consciousness, if you would always say that it doesn’t explain the subjective experience itself? It isn’t just “no one can predict specific content of consciousness to the level of precision that we can do with organs now”, right? And if it doesn’t have predictions, if you just say “yes, I don’t know how explanation would look like”, than why do you thing an explanation that changes your mind about gaps must have anything to do with observations?
Like, there is a mental process inside you that looks at explanation and says “no, it is not an explanation, because it doesn’t predict anything” or “no, it is not an explanation of consciousness, because it predicts many things, but they are not consciousness”. So why do you just trust that whole mental process? What evidence do you have against the theory that this mental process leads you to contradiction when you apply the second part to physicalist explanations and first part to explanations about why you shouldn’t use the second part like that?
And, is it really your disagreement here? That there is no way such thing can be an explanation. Because I honestly can’t imagine myself seeing an explanation that works and dismissing it just because it doesn’t satisfy some heuristics. So I would thing that the part where it works would be more controversial.
Sure, treating human brains like black box for ethical reasons makes producing high-quality models harder. But that’s all perfectly predictable normal difficulty considering complexity of the brain, not a reason to doubt even theoretical possibility of precise modelling.
I am only observing that no explanation I have seen accounts for the existence of subjective experience. An observation is not really a prediction of anything but itself. If I observe that it is daytime where I am, that on its own is not predicting anything beyond what I already saw: the sun in the sky, the bird on the wing, and so forth.
I find your next few paragraphs too incoherent for me to respond to, so I shall just make the following remarks as a conclusion to all this.
Some people believe they have found an explanation for consciousness.
Some people believe that someone else has found an explanation for consciousness.
Some people believe that saying that consciousness is ontologically primitive is an explanation for consciousness.
Some people believe there is no such thing as consciousness.
Some people believe that there cannot be an explanation for consciousness.
Some people believe that there must be an explanation for consciousness.
I am not any of these people. I find all purported explanations that I have seen wanting, because each of them fails to account for the very existence of conscious experience. (Most of them are no more than the “A is B, therefore B is A, in a sense” fallacy.) The claim that it is “ontologically primitive” is no better than saying that it is magic. The claim that it does not exist is refuted by every moment of experience. I have no reason to think that there is no explanation, but on the other hand I am not moved to exclaim that there must be one.
I cannot say what an actual explanation would look like. Everything that we have discovered about the world so far fails to account for the existence of experience, yet there it is, mocking our attempts at understanding.
It’s technically possible for something to be a pure observation, but what usually happens is you use some aggregate concept to describe your observations and this concept implies some predictions. So...
What do you mean by “accounts for the existence of subjective experience”? “Consciousness is existence” accounts for the existence of subjective experience (or rough description of how brain works accounts for the existence of subjective experience otherwise). It (in combination with physics) predicts that the brain like yours would experience things. And you can confidently predict that when we have better neuroscience it would predict what someone will experience with better precision. You can call it magic, but it’s the same magic you have with the concept of existence anyway, so usual anti-magic heuristics don’t apply (or it should be argued separately that existence is useless/incoherent concept and it would still result in physicalist account of consciousness). Or you can call it fallacy, but then it would be a fallacy that works in real world.
Does 2⁄3 of a rock have experience? Does every subset of the atoms of a rock have its own separate experience?
Yes, but they are not separate the same way the consciousness in the brain is not completely separate from outside world, because physical things interact. They are just different descriptions of different parts of the consciousness of the universe.
Avoiding epiphenomenalism and zombies is supposed to be a feature of EC.
Yeah, that’s the point—every property except existence is vulnerable to zombies.
No property that is evident is subject to zombiehood.
Yes, I meant vulnerable to constructing conceivability argument which implies that property does not specify consciousness.
And I mean that you cant conceive of a property being both present and absent.
I’m failing to understand your comment in context and need elaboration. My point was that you can’t use any property, except existence, as a reason for the absence of consciousness, because that property doesn’t change for humans in the zombie world, but they are still not conscious. And all that assuming that without accepting “consciousness is existence” you don’t have sufficient reasons for dismissing conceivability argument. So without “consciousness is existence” you reason that all properties are present for both real world and zombie world, and with it you either say that physicalism forces you to conceive that zombie world doesn’t exist or it’s a contradiction to say that you can conceive a world that both exists and doesn’t have existence aka consciousness.
The whole point of zombie argument is that consciousness is or depends on a non physical property that can’t be detected externally.
Ah, you mean I should have said “For every physical property” instead of “For every property”? Sure.
Most people have an intuition that simple entities aren’t conscious, and Consciousness is Existence implies they are, even if Consciousness is software does not. Note that causal.closure, non epipheniomenalism, etc, are also intuitions.
Sure, but if you inspect the reasons for this intuitions (mostly “simple things don’t talk”?), you find that they may just as easily imply different kind of consciousness instead of absence of consciousness. Or alternatively you can use different words for the difference between us and zombies and the difference between us and simple things, but then the difference between us and simple things is not mysterious.
This sounds like the naive theodicy that reconciles the existence of evil with the goodness of God by saying that God is indeed good, but in ways we cannot understand.
Does electron consciousness do anything at all?
It moves and spins.
Not in the sense of a spinning top. It’s in various quantum states.
What’s a program?
This seems right to me. I’m not confident that it’s going to be persuasive to people who don’t already have similar intuitions, but I hope so!
I think the idea here is broadly resonant with Infra-Bayesian Physicalism (where algorithms have a distinguished ontological status—and this alone explains subjective experience), and I’m curious if @Diffractor and @Vanessa_Kosoy agree.
I particularly liked this part:
which reminded me of my own statements that
The post does quite a good job in nailing down what it means. I dislike -isms and here they would be in danger of confusing the thing. I thought that panpsychism worked like EC holds itself. In the numbering reference to physicalism is made. Following the link first proper post about it starts of with a claim about “reductionist materialism”. The reductionistic part is used but mental or neutral monism would also fly. The pinpointing is needed and I quite like the style.
Computers seemed like magic to me at one point, but once I read an explanation of how they work, I felt that there’s no longer anything mysterious about computers. Even if I’d never seen or heard of a computer, if I was more intelligent I would have predicted that computers could exist.
After reading this explanation of consciousness, consciousness remains just as mysterious. If I was not conscious myself there’s no way in a million years that I would predict that other people are conscious (that there’s something it’s like to be them). As such I don’t think this really explains anything.
Yeah, that’s correct, thank you for writing this. It is isomorphic to Russelian cosmopsychism, but it’s just terminological question.
I was nodding along most of the time, while reading your explanation. “Consciousness as a software program” have seemed right to me since I first thought about the problem from the materialist perspective. You example with the content of the program existing even if there were no screen and just vocal statements is pretty good. However.
I’m not sure I understand how you make the logical step from “The program actually exists” to “Existence is consciousness”. As it was already mentioned in the thread, the second statement doesn’t depend on all the previous claims, regarding software program and its content.
Why would you even need to make this step? We have a software program, an inner model of ourselves. In this model all kind of stimulus are encoded into inner experiences in some way. What is encoded to be represented in the model we can feel. What is not encoded—we can not. If there were no model there wouldn’t be any inner experience. I was genuinely under impression that the interesting question is how exactly the encoding works. When it is answered the consciousness will be fully explained.
Until then I don’t see how “consciousness is existence” makes us less confused. It just brings back the confusion of panpsychism where we assume that entities without any inner model are somewhat conscious.
Reading this, my thought was that this would have absolutely blown my mind if I hadn’t been primed/prepared for similar ideas from my (fairly superficial) encounters with similar ideas from the Qualia Research Institute, David Pearce, and even this old [Overcoming Bias post](https://www.overcomingbias.com/2009/12/does-real-feel.html).
So I was kind of surprised to see this [thread](https://nitter.it/algekalipso/status/1562889329399123970#m) on Twitter (link is to Nitter). I am a big admirer of QRI and Andres Gomez Emilsson but in this case I do wonder how carefully/closely Andres read your essay. I had had the impression while reading your essay that it was very close, possibly in agreement, with some of the relevant positions of QRI, while, sure, not addressing the binding problem.
You actually come fairly close in your essay to outright saying you’re explicitly not trying to address the binding problem: “we are not talking about … how our experiences seem somewhat unified”. (Caution here though, I don’t yet understand what’s meant by the binding problem and have been waiting for it to “click” for me for some months.)
So I thought your essay was excellent, fwiw from someone way, way out of their depth.
I think it’s addressed in the text, but the binding problem is solved by physics—fundamentally, there are no atoms, only interconnected universe. Then the decomposition problem—why we can’t read others’ thoughts—is solved by the observation that we can, as in our thoughts are casually affected by the others’ thoughts (for example when we hear someone say something about what they think) and degree/speed of casual influence is consistent between physical laws and our feelings about mind isolation. Ultimately there is no isolation of qualia—only qualia of isolation. And when treated as representation of reality, qualia, as any other representation, can be imperfect.
But that would still work in a classical univserse,without quantum entanglement.
The point is that global entanglement would predict thought-reading without speech.
Entanglement doesn’t allow any transfer of information. So thought-reading (or anything that requires transfer of information) through entanglement is impossible.
Classical universe is not less global—it’s either the case that an electron from faraway galaxy is tugging on your thoughts or, if you want to consider classical universe with relativity, it’s the same kind of locality, just not in Hilbert space. Anyway what’s predicted is very small-scale thought-reading and that’s what actually happens. I think the key element here is the theory of knowledge that supports you being unaware of small-scale differences in your qualia.
Yes it is, because it lacks nonlocal correlations.
Don’t you mean nonlocality?
No-communication theorem is a thing that exists and applies to quantum stuff.
Light-speed keeps to be a relevant limit on what can happen. If you have a superposition of state A+B and it resolves “superluminally” to one of those two, it can only span an area that A could have spread (at lightspeed) to or B could have spread (at lightspeed) to.
Global correlations themselves are not that unusual? You can just write “+” and “-” on different halves of paper, take random one, and now wherever you go, if you see a “+”, you know the other one will see a “-”. So classical universe also has nonlocal correlations. But in both universes they are locally originated and things only change locally—you can only move your shared blob of amplitude no faster than light, and it only changes when someone locally interacts somewhere they moved it.
And I mean ok, you can call changing my blob with interaction not in my location “nonlocal”, but there are still no implications for thought-reading, except yes, you can expect some correlations, but so what? Presented solution for the hard problem predicts them and they happen.
In classical universe, you can write plusses and minuses on a piece of paper and tear it half , but the two halves will always remain within the future lightcone, however much you separate them...so they are not nonlocally correlated , in the technical sense.
Yes, and with that definition there is an equivalent statement about quantum universe. You may just need to draw your cones in Hilbert space.
That’s not how it works. Hilbert spaces don’t even have light cones.
Metaphorical cones? Do you object to the possibility of geometrical representation of speed of influence-propagation in Hilbert space or what? Because (relativistic) quantum mechanics is local by both “you can’t spread influence faster than light” and “it uses only local derivatives” definitions of locality.
You can argue that QM doesn’t allow nonlocal influences or signalling, but that isn’t the result of the existence of anything analogous to a Hilbert space light cone, it’s the result of the fact that you cant control a nonlocal correlation,.so you cant signal with it. But nonlocal correlations are still there.
First, to be clear, there are direct analogs of lightcones in relativistic QM that disallows just moving faster than light. But even when talking about entanglement specifically, it depends how analogous you want it to be? It’s still “the space/equations prevent FTL signaling”—why it matters that it’s lightcone in pieces of paper case and “you can only move nearby blobs of amplitude in some dimensions”? I mean, from that perspective it’s kinda weird you can’t signal through entanglement. Is it just that it fits the “you can only affect stuff that is associated with your classical spatial location” definition of locality? But sticking to that definition is not motivated by the reasons we initially cared about it!