"Why is it like something to be something?"

I can try to answer this, but you've got to avoid getting jerked around by your heuristics. In one sense I went from camp #2 to #1, but in a deeper sense Reality does not admit of camps. It is possible to express the answer as panpsychism, or as functionalism, or as any other number of -isms. If that strikes you as impossible, you are either still confused about consciousness or too attached to your labels.

This discourse is stuck. It's as though one person wants to go get "lunch", another wants their "mid-day meal", someone's cells are crying out for ATP, and a dozen others also want something "different" - but everyone's arguing and starving.

In the same vein: I am not peddling some unique grand theory. I used to think my understanding was unique, but that's because I also wasn't hearing what the various camps were trying to tell me.

The nature of my confidence is like this:

If some mind had no preexisting ability to represent or embody the simple concepts of "truth" or "if -> then", then they probably couldn't be communicated with at all, and probably aren't a human. Thankfully, humans do have a naive/simple notion of truth already, so telling people to "please use the simple truth" can actually work to collapse them back into the proper state.

But for the person who needs to hear "please use the simple truth", they need a lot more than just those 5 words to fix them. (That post has ~6800 words.)

So:

The understanding I wish to impart is a similar kind of simple thing, but it does not conflict with something like predictive processing, which also makes sense. It is too simple to be simply communicated. I'm hoping we can communicate, but it takes two to tango.

I also can't answer every question related to consciousness, in a similar way to how using the simple truth doesn't let you resolve every truth-question. Truth statements can be about arbitrarily complicated things, but truth itself is simple.

Enough preamble.

The question we seek to answer is: 
"Why is it like something to be something?"

And different people are going to want different levels of answer. So let's proceed through the easier ones.

Level 1: there is some causal chain

It can be useful to step back and remember that there is probably some causal chain leading you to talk about consciousness, and so in principle an answer is possible.

(If you think there isn't some reason or causal chain for your vocalizations, that's "fine", but then I'm not sure why you'd be looking for an explanation that you don't think exists - i.e. why would you be reading this post?)

But, more likely, this isn't enough for you, and I don't blame you.

Level 2: there is only one type of stuff

If there were more than one type of stuff, then those different types couldn't interact.

If they could interact, then why would we say there are different types? That just doesn't seem useful. If there is some causal chain reaching across a conceptual boundary, just remove the boundary so you can think more clearly about the causality.

So: we can restrict ourselves to thinking about just one type of stuff. I don't care what label we use.

If there is just one "plane" of stuff, then you, being made of that same stuff, are an object that exists on or within that same plane. If there cannot be a 2nd plane, then the intuition of there being a separate "mental" plane has to be explained-away or dismantled somehow.

But, again, this too probably isn't enough. 

Level 3: perspective isn't magic

Much hullabaloo is made over the apparent difference between our experiences vs their in-brain representations.

But if you have just a single object, you can still have multiple perspectives of that one object. The map is not the territory.

You can still build "perspective" out of just one type of stuff, and you don't need to assume consciousness to do so.

Just think about cameras:

One possible perspective is of the image being displayed, while another is of the binary encoding of the image file.

The display sure does look different from the binary representation, but there is still just one image.

In considering the possibility that you might be like the display, it shouldn't surprise you that the representations - the neuronal patterns - can look radically different when you "display" (access) them.

You still have to investigate whether the difference that we actually find in our case is larger than could be explained by this dynamic, but such an outlook is very different from how some seem to be treating the mere presence of this predictable difference as a deep mystery, automatically.

"But access consciousness is not the same as phenomenal consciousness!"

I think you are prematurely dismissing "mere representations" or "mere access" based on surface appearances. Things are allowed to appear different under different circumstances. Cameras and displays and software abstractions are real things that manage to accomplish this every day. The camera display never shows the binary representation, even if that's what's going on underneath.

The display is a real thing in the real world, and so are you. The display sits at a particular level of abstraction or interpretation, and that abstraction is very real and meaningful. You don't get to handwave that away. It is not a "mere" abstraction. It is a reflection of the fact that the screen never prints in binary.

Perhaps you too sit at a particular level of abstraction, even if this "you" thing needs to dissolve and can't really be sitting in there.

It is not as though our only options are either complete abstraction to the point of reconstructing another little guy in there - vs an undifferentiated soup that couldn't even calculate or do anything anyways. Patterns emerge out of the soup, and you have access to, and are, one of those patterns.  

You don't have to go full-blown "there is a little screen in the brain" to recover some screen-like qualities; to have access to abstractions but not to what lies beneath those abstractions.

One of those abstractions is color, and it's worth reflecting on why it is such a focal point in this discourse ("the redness of red").

Why is there not an equal amount of confusion about the "shape of shapes", or the "line-ness of lines"?  

Isn't the real mystery of the hard problem in the jump from representation to consciousness, regardless of what content we happen to be representing? Or does it demand that we go through every single instance of mental content? When do we get to generalize?

Why can't we use some simple mental content, and use that to get deconfused about consciousness, rather than trying to do everything all at once?

Believe it or not, it is actually easier to think about confusing mental content when you do not have the hard problem looming over everything, rather than going the other way around.  

You do not have to "explain the redness of red" for your explanation of consciousness to be complete, because explaining "why red looks like that" is a straight-up separate problem from why there is consciousness of anything at all. These questions share similar territory, and so are interfering with each other, but they are still distinct questions.

Remember the analogy to "truth". Truth can be about arbitrarily complicated things, but truth itself is simple. If someone was still confused about truth, but was trying to jump ahead and answer some truth-question about quantum physics, in order to better understand this "truth" business, they'd be doing things backwards and making everything way harder than it has to be. The "truth" part of a confusing quantum physics question isn't the confusing part.

Colors are - forgive me - a red herring.

But, fine, I will attempt to confront the "redness of red" more directly at a later level, when I have more firepower. I know a boss fight when I see one, and I'm not about to walk in there with just my measly camera.  

Level 4: tell me how you really feel

When you are unsatisfied with any "merely functional" explanation for phenomenal consciousness, it can sound like you are looking for something that:

• is beyond "mere function"
• and yet can be talked about, or interacted with, as you are doing now.

But thankfully there is a way to avoid contradiction, because we already have a concept that can fit these criteria: "existence".

(Easy now. This isn't magic. Give me a second here.)

Because "existence" as an isolated property: 

1. Is orthogonal to "functioning", in the sense that it is a background assumption. A thing has to exist before that thing can do something. It is not like another gear in the machine.
2. Can be recognized within the functional paradigm (since recognizing this property is a process), even though the property we are recognizing is itself outside the causal chain. So you can reflect upon and talk about it.

We are already invoking this concept in other places. It is being used when we say "to be" in the hard problem formulation: "Why is it like something to be something?".

So if this concept can also serve as the orthogonal-to-functioning "phenomenal" thing we are looking for, then that would mean we are repeating ourselves when we invoke the hard problem, but without knowing it. It would be as though we were asking: "why is the image on the screen, if the image is on the screen?"

I don't expect this to make sense yet, so let's get concrete:

Imagine a human looking at an apple.

Photons bounce off the apple, hit our retinas, and generate neuronal patterns. The patterns are a representation of the apple, and our “consciousness of the apple” is simply our recognition that this representation exists. It is a nonverbal restatement of the existence of that part of our internal model.

If you explicitly ask the question: "why does my internal model exist?", then answering that question isn't hard.

Instead, I am claiming you are implicitly asking that question at the level of your experience, so you do not realize you are "asking" that question. It is the nonverbal experiential equivalent of: "why is there something rather than nothing?". You are not "asking" about low-level substrate, you are nonverbally "asking" about the existence of the abstractions you are interacting with on your level, as a kind of software object yourself.

When you, as a little piece of Reality, bump into another piece of Reality, you are implicitly recognizing that the thing you just bumped into is there; that thing exists. You don't need language to accomplish this kind of "recognition". It is a pre-verbal, deeper, more intuitive operation - especially when it happens inside your skull.

This is why the hard problem intuition can stick around, because in terms of our experience, "there just is" our experience, as a brute fact that cannot be probed further via introspection. You can't "look really hard" and somehow gain access to the lower-level details of implementation. This "scrambling for something ineffable" is actually a feature one could predict from the physical setup alone (but there's more to be explained here because some people aren't struck by the hard problem at all).

The hard problem formulation highlights a gap  - and there is absolutely something Deep and important that goes in that gap - but the framing makes it really hard to see.

If I asked you what you ate for breakfast 20 days ago, you might struggle to answer. But if instead I asked you what you normally eat for breakfast, you might recover the answer. The questions we ask, and the way we ask them, can make it seem like there is a gap, when there isn't. Different questions highlight different things.

The thing being missed here - the thing that goes in the gap highlighted by the hard problem formulation - is SO implicit, it is so "close" to us, that it slips under the radar. Then when it slips under the radar it makes it feel like the hard problem formulation is correct!

("omg, you're right! I don't know what I ate for breakfast 20 days ago!")

It primes you to think of emergence, which makes it really easy to miss that things already exist. It makes you want to shuffle around pieces to somehow produce a feature that is already there in the pieces.

("omg you're right! I can't produce a satisfying candidate for this mysterious quality, no matter how I compose the pieces!")

It's like if you used your phone's flashlight to look for your phone. You are correct to search, and the thing you are searching for is real, and that thing can be found, but you probably aren't going to find it like that. You are "too close" to your object.

Or recall those pranks where someone is texted a picture of their own phone, and the message is like: "hey come back you left your phone".... but they actually come back for it:

The hard-problem-enjoyers are - validly - grasping out into the darkness for something orthogonal-to-functioning, but they don't know what, and their search feels futile.

They are demanding to get their phone "back".

They are looking for the glasses already on their face.

Their internal model is already there. It exists, and they are happy to grant that....but they also want to know why it's there - at the level of their experience.

"..."

"..."

".....Right. That's...the whole problem here? Why is this phrased like a revelation? You are just restating that we want to know why it feels like anything to be this internal model, no matter how much detail is in that thing. That's the question we started with."

You have to be comfortable being psychoanalyzed a little. Your consciousness can't be an illusion, but you are making a kind of mistake. You are ironically not taking your own perspective, your own circumstance, seriously enough.

You are a piece of software, and that comes with some unavoidable limits on your perspective. You are like an NPC in a videogame that can only see text, and is now wondering, at the level of your experience, why that text is there.

So there's two levels here: 

1. The NPC can't see beneath the level of abstraction that they operate within,
2. and they cannot see any other abstractions aside from those that are handed to them.

At the level of their experience, the existence of that text is just going to be a brute fact of their reality, and it will be the only thing which will exist for them.

With you, you can pop back up a level, look around (you are looking through your experience; you are "using your experience" to look at the world beyond), and see that "of course all this stuff exists, how is this relevant?" - but when you pop back "down into" your perspective - when you look at your experience itself and ask why it's there - the mystery will rear its head again at a nonverbal level. You will be asking: "why is there anything there?".

The feeling that something is missing from the functionalist paradigm is itself a feeling. It doesn't matter that you've given that feeling a name - "the hard problem" - and can now toss that concept around as though it were a logical puzzle to be solved on its own terms. It might have been such a logical thing - in a similar way to how reading a valid math statement feels like something (and you don't get to psychoanalyze-away your math problems) - but it isn't.

The source of the overall "hard problem feeling" is the "there-ness" (existence) feature of your internal representations being implicitly missed by the hard problem formulation, because it makes you think in terms of emergence. It makes you think of consciousness somehow being produced, if only you could rearrange the matter (the stuff which exists) correctly.

There is no magic recipe. The abstractions which you are accessing - which are built out of neuronal representations, which are built out of stuff - really do exist. They are really there in a way that doesn't need any more justification, and just like the camera display they are not just "mere" abstractions, because they accurately capture features of reality. But in terms of your experience they can only ever appear to you as an unexplained and unexplainable miracle.

Completing neuroscience and building brain emulations won't ever grant you an "aha" moment aside from this kind of: "wait, that's all I am?", because there is no magic threshold to be found in the first place. All that's going to happen is the equivalent of cataloging every type of mental content you have "mere access" to - all that stuff that "merely exists" - until you run out of such things to explain, and then you'll have to confront the stark reality that that's all there is. You can go ahead and confront this reality now.

The subtlety of the existence property also helps explain why some people don't understand why the hard problem is a problem, at all, because in a sense it really isn't for them.

Recall how tiresome aphantasia discourse can get because it involves comparing phenomenology verbally. Then with this topic (hard problem discourse), we are dealing with something like "the phenomenology of phenomenology", so of course we were going to have language breakdowns.

The people who do not understand the hard problem are not generating the same kind of gap, where the existence property remains to be explained, at the level of their experience. I'm not claiming there is a radical difference in their conscious experience; this is just a subtle difference, and has more to do with what happens when they encounter things like the hard problem formulation than how they experience the world. They are doing the equivalent of directly remembering what they had for breakfast 20 days ago, without needing the reframing of asking what they normally eat.

They will say things like: "There is (emphasis mine) a representation of the thing we call red in the brain, but that's all there is. What exactly is the problem?"

They are not being deliberately obtuse or disingenuous; they are sincerely confused. In other words: they are glossing right over the "existence" property too - just in a different way than the hard-problem-enjoyers are.

Something like this is to be expected. "Existence" is an easy thing to gloss over, and you might still be doing it now. When I tell you "there is an apple on the table", that apple "also" exists, but neither of us is going to go out of our way to mention that fact. The existence property is implicit in the vast majority of contexts. But when the thing-which-exists IS our internal model, which IS our experience, then navigating that implicit orientation-to-existence is less straightforward.

In every other circumstance we are used to handling the existence property through our experience; we are dealing with the existence of things out there in the world, through the world of our experience. But in this case we are dealing with the existence of our experience itself (the existence of our internal model). It makes sense in the abstract that things would get a little weird in this edge case.

Level 5: the perspective of being something

Consider a more meta version of the map vs territory distinction: There is just one Reality - but you, in being a single particular subset of Reality, have a different "view" on your subset than anyone else. (At least for now.)

So how do you know you aren't like the camera display?

Are you a camera display? No.

But what are you?

What do you know, and why do you think you know it?

Do you think consciousness is a special thing? How did you come to that conclusion?

You only have exactly as much access as you have, and you deem what you have special! Isn't that curious?

What if something outside your purview also "glowed" with specialness? How would you find out about that? You don't have access!

Take the hard problem formulation: "Why is it like something to be something?"

Then break it into its parts: 

1. the "like something" aspect
2. the "to be" aspect (existing as something in particular)
3. the thing in question

Why do you assume that 1 and 2 are different things? You can only ever be one thing at a time, and the one thing that we happen to be.....also happens to be conscious.

I don't dispute that there is clearly "something that it is like" to be us.....but you also exist. And the thing-that-you-are is a highly detailed representational model with a reporting mechanism.

"To be" something is to be a particular thing, with access to particular things.  

If you take the perspective of a Thing that reports on the contents of some highly detailed representational model - which is to say, your perspective - how is that not just the feelings or experiences themselves? Why add another layer on top of that to ask why that process feels like something? Those things are your feelings!

It's as though you are marveling at how perfectly water takes the shape of its container.

It's as though you are asking: "why is there stuff on the screen, when there is stuff on the screen?". You are repeating yourself.

The hard problem formulation is misleading in part because it can suggest that "existence", or what it means “to be something", is a settled and completely understood fact, and then coming out of nowhere we have this separate mystery of why “it is like something to be something". What can be viewed as "phenomenal" is just a matter of perspective; to be the brain representing something vs the perspective of looking at that brain externally.

Matter doesn't "wake up". You just ARE a particular, singular, Thing. You are not everything. You are some-thing. The corner of the universe that you have access to is "lit up" for you, because that's all you can see. 

There is no magic "light" doing any work here. All that's going on is we are selecting a subset of Reality, and then identifying with that subset.

We have no issue pointing to a camera and asking about its perspective - to ask: "what can that Thing see?". But for some reason we really really struggle with identifying ourselves as just another object in the world. Ironically, we are struggling to imagine the perspective of just one object in particular..... the object that we happen to be! When we try to visualize from the outside, for some reason it's really hard to take the final step of "putting ourselves into" that object.

Just like the camera, or any object, our access to the rest of Reality is also limited. You have a boundary too.

"But why am I THIS object?"

Someone's got to be that object, because that object is asking that question. What is the alternative here? It's not like that object can sit around waiting for an owner to show up. Some other object might be really into buddhism or meditation - or on psychedelics - and so not asking the "ownership" question you just asked, but you are not that guy.  

"But why is it a necessary fact that some object feels like anything?"

That is just the converse of recognizing that you have to be some object, somewhere. You don't get to be a floating, ghostly, disembodied, abstract intelligence - sorry.

You have a physicality. You have a location. A position. An identity.

You are an object, and you have a perspective.

So beyond being allowed, you are required to assume the position imagine the perspective of other hypothetical objects if you are seriously trying to figure out what kind of object you are and how this all works. If you are unwilling to exercise this part of your imagination, you will have to wait until we have brain emulations to play around with, like a monkey figuring out that a mirror is truly reflecting them and not some other monkey. Do you want to wait that long?

The software object that you are has access to other software objects, and all are instantiated via real physical objects doing real things.

You are not a hypothetical object. You are a real object. I'm sorry to tell you this, but you actually exist. 

Okay now we can more properly deal with the redness of red.

Imagine we created a software agent, and for every file that agent interacts with, we calculate a single number. Let's call it the "nifty" score. And we create our agent such that it cannot report directly on what this number is, it can only report the result of comparisons. So given two files it could tell us which one was more or less "nifty", but that's it. Maybe we even add in a little fuzzing/randomization to the calculation of the nifty score itself.

So we give our agent two files and ask: "which is more nifty?" and if they are close enough together that the fuzzing/randomization makes the answer to that question unknowable (the difference is within the range of the fuzzing), it reports that fact; it says: "I can't tell; they seem equally nifty to me". Or if they are sufficiently different, it can tell us: "this one is more nifty".

We could even hide the size of the random fuzzing from the agent, and/or change it over time, so in edge cases our agent is even more unsure.

We could even have the agent continually resample / recalculate the nifty score, so that even as it is speaking the niftyness is shifting back and forth across the edge of comparative edge cases.

For the agent, there is definitely a "niftyness" for every object it encounters. (If given just a single object, it can still compare against a memory.)

But if it is a fact that the only thing the agent can access is the output of the comparison, then.....that's all the agent has.

Maybe one day the agent can access its own code and trace the causality of its vocalizations, but even then, if it is only externally reading its mind without changing anything, then it will still go on feeling like things "just are" more or less nifty. There is nothing else it can say. Beyond what it can say, the niftyness is "ineffable".

When we go back to thinking about our experience of color, there is obviously way more complexity, and it may or may not involve this specific comparison mechanism. I'm just using the comparison mechanism here because it makes the representational boundary more evident. But all that complexity only makes things worse for the person trying to get around the ineffability and find what the color "really is". 

There is nothing we can say other than the end discrimination of color, just as the nifty agent can only vocalize their own end discrimination. That's all we have access to.

The difficult part is not "how can you construct ineffability in software" - the difficult part is where you allow yourself to identify with that kind of object; as an object with limitations that work like that.  

Ineffability is quite normal and abundant. Ineffability is just when our access to what is going on is limited, which is just to say that it has to stop somewhere. It is us hitting a within-experience explanatory wall that we can't see behind. But the non-magical NPC also can't talk about what is "behind" the text it can see, and the nifty agent also doesn't know what is upstream of the nifty discrimination, and the camera display also never prints in binary. On some level it sucks to be just another object in the world, but it's better to come to terms with it.

Level 6: sufficient but not necessary

There is a way to nuke the right intuition into you, but it's not going to work for everyone.

It is a little overkill.

Take a fictional story; words printed on paper. 

Ask the question: "does that story actually exist as some other reality/universe?"

As some kind of platonist, I would say: yes.

Okay, so:

You are running a little simulation - you are telling a little story - in your head, yes?

Why does your story seem so Real? --> because it is.

(It remains to be argued that platonism is correct, but I'm not going to do that here. This is for people who already like such things.)

Your map is not our territory, sure. But it IS some other territory.

The Realness; the redness of red; the redness of a "mere representation" of red - could be considered to be "coming from" the same "source" of Realness that makes our actual physical universe real - in the traditional "why is there something rather than nothing?" sense. Every possible reality, or representation, or mathematical structure - is equally real.

The map that IS the territory doesn't get any additional information (it does not suddenly become as high-fidelity as our physical universe/reality), rather, we are just interpreting the same information differently.

Please pay attention to that last sentence. If you are a hard-nosed physicalist who wants hard problem discourse to go away, you might want to pause and reconsider before fighting this intuition pump. I am not suggesting anything extra happens in our universe, so the causal reasons for talking about the contents of consciousness as we do remain the same. The purpose or end effect of this framing is to make people take the existence of the representations they have "mere" access to more seriously. We've got all the Reality we need to explain consciousness in this universe, but it doesn't hurt to supercharge that existence if it helps the intuition land without any side effects. 

Your internal model is really really real, it is really really there - that thing exists - and that thing is what you call your experience.

Okay, but is X conscious?

I cringe every time I hear phrasing like this. Even if, yes, I sometimes slip up and do it too. It's a common shorthand, I get that.

But no matter how much I too wish there was a magical binary dividing line, that is not how this works. At all.

Yes, it is turbo-inconvenient that we don't get to have that kind of strong prerequisite for moral patienthood, but Reality is often inconvenient.

For interpretability and such, I'd prefer if we just directly looked for valence structures instead; things like pleasure and suffering. We should skip over the consciousness question by confidently and openly taking it for granted. It may sound strange, but we actually have more evidence for the existence of positive, negative, and neutral valence than we do for "objects doing things but still being unconscious". We cannot confirm any cases of unconsciousness, but we can confirm consciousness in at least one case.  

Which timeline seems safer to you? 

1. The one where you are hoping some plot twist of architecture means you can treat some worrying maybe-suffering structure as actually unconscious ("whew, thank god")?
2. Or the one where you treat all structures seriously, whether you think there is "light" being cast on them or not? This does not mean taking vocalizations at face value, but neither does it mean ignoring them. We are not entitled to easy answers.

Do you confidently flip on the lights and look for the boogeyman? 

Or do you rationalize your way out of turning on the lights at all?