Why Consciousness is Causal

The original title of this post was “Is Consciousness Causal?”  The last few days I’ve had some incredible insights.  Here is my original post from October 4, which I had not yet had a chance to post:

I had a few insights yesterday, while at the Models of Consciousness2024 conference, at which I presented my talk, “Artificial Intelligence Cannot Be Conscious.”

Start with a physical system in state S1 at time t1.  The laws of physics dictate that that state will cause or determine the next physical state S2 at time t2, subject only to new information (quantum collapse events).  In other words, starting at S1, there WILL be a following state S2, and what determines that is S1 plus any new information.  You cannot “destroy” or “erase” a physical state – all information remains embedded in the current physical state of the system.  Information never disappears from the universe.  Time always moves forward (because time travel isimpossible) and the physical world is physically irreversible.  Every physical state evolves in time to a next physical state, either with or without new information, that embeds all information from the earlier state.

It seems like conscious states are similar.  I have shown, or am in the process of showing (Youtube video), that every conscious state inherently embeds every prior conscious state.  For example, to get conscious state C2 at time t2, what you have to do is start with conscious state C1 at earlier time t1 and let it naturally evolve in time, along with the addition of new sensory stimuli (such as from our five senses).  That is, it seems like C2 is determined – i.e., caused – by C1 and any additional sensations. 

The alternative is that conscious states are NOT causal, in which case they happen to arise purely by accident (e.g., as an “emergent” characteristic of an underlying physical state).  Of course, this notion already produces its own problems, particularly that of how consciousness could have arisen by Darwinian natural selection.[1]  Still, if consciousness is not causal, then when S1 “accidentally” produces C1, and then the physical world changes to S2 (through time evolution from the laws of physics plus new quantum information), why should it be the case that state S2 “accidentally” produces conscious state C2?  After all, C2 feels (i.e., is experienced) as if C1 was experienced just prior, immediately followed by the experience of new physical sensations.  In other words, C2 feels like a natural and continuous flow from C1.  Why should that be the case?  In other words, conscious states are experienced as if they were causal and continuous: C1 (plus some sensations) seems to create C2, which (in combination of more sensations) seems to create C3, and so forth.  If conscious states actually are not causal, then we need a good explanation for this incredible coincidence. 

But another problem, perhaps related to the one above, is why physical states cannot produce more than one instance of a conscious state anywhere in spacetime (which is my landmark proof here).  For instance, say that physical state S6 produces conscious state C6, and S7 produces C7, and S8 produces C8.  I have already proven that conscious states can’t repeat, which means that it is impossible for next physical state S9 to produce, for example, (already experienced) conscious state C7.  If conscious states are not causal, then what could explain this?

Notice, by the way, that the underlying physical state determines (at least in part) what physical sensations a person experiences, which means that whether or not consciousness is causal, a given conscious state depends at least in part on the underlying physical state (e.g., any new quantum information).  But if conscious states ARE causal, then just as conscious state C2 depends on physical state S1, S2 must also depend on C1. 

For example, imagine that I am driving and coming to an intersection, where I see a stop light turn red:

·         If consciousness is NOT causal, then the universe does not “care” how I experience the color red – that is, my experience is entirely inconsequential.  My next conscious state will depend only the next physical state, which will be independent of my experience of the color red.

·         If consciousness IS causal, then the way the universe physical evolves depends at least in part on how I consciously experience the color red, which means my next conscious state also depends on that experience.  (I have neglected talking about free will in this post, but obviously if consciousness is not causal then there is no free will, but if it is causal, then there is a possible opening for free will, somehow related to conscious experience, to affect the underlying physical state.)

Another thing… if consciousness is causal, then there does not seem to be any way to stop it or destroy it.  You can’t destroy or reverse the information in a given physical state… every physical state produces a next physical state.  Similarly, if consciousness is causal, then the same is true of consciousness: every conscious state produces a next conscious state.  I have already shown that you can’t reverse a flow of consciousness and, analogously to the underlying physical system, if we start with a conscious state C1, all we or the universe can do is add new information (e.g., visual or auditory stimuli), which will inherently produce a next conscious state C2.  There is nothing we can do to end or stop that stream of consciousness, just as there is nothing we can do to stop the flow of time or the time evolution of the physical state of the universe.

Either consciousness is causal or it is not. 

Here’s a related thought.  Physical states embed their entire histories.  State S3 tells us everything about state S2, as well as what new information appeared since then.  State S2 tells us everything about state S1, plus all new information, and so on back.  It certainly seems (but have I properly shown/proven?) that conscious states are perfectly analogous.  I could not experience my current conscious state C3 without first experiencing state C2 and any intervening physical sensations, and I could not experience C2 without first experiencing C1, and so on back. 

In a very real sense, physical state S3 is the indestructible physical manifestation of its entire history of events or facts (as I discuss here).  It can’t be destroyed because the universe only produces new information; it does not destroy information.  Analogously, isn’t my current conscious state C3 the evidential manifestation of my entire conscious history (including all my physical stimuli and how I consciously experienced/perceived them)?  If my consciousness could somehow be destroyed, then where in the physical universe would that evidence/history of consciousness reside? 

If that were possible, then consciousness is not only non-causal, but it arises as a cosmic accident and does not affect the physical world.  Of course, that seems ludicrous to me.  But the only other alternative is that consciousness is causal, which means that my current conscious state C3 just IS the universe’s evidence of my conscious history, in which case it is as indestructible as the universe’s physical state.  It would also imply, as I already believe, that the way I consciously experience the physical world actually affects it (whether or not free will is involved).

This isn’t a proof, but I am noticing a close analogy:

·         Physical states are irreversible.  (I have proven that) Conscious states are irreversible.

·         Physical states are history-dependent and embed their entire histories.  The same is true of conscious states.

·         Physical states are causal.  It seems like conscious states are causal.

·         Physical states cannot be destroyed because information (facts about the universe) cannot be destroyed.  My conscious experiences (e.g., experiencing the redness of a traffic light, or feeling sadness or whatever) are indestructible facts about the universe, and those facts must be embedded in the universe.  It seems like those facts are embedded in my conscious state, so it seems like my conscious state cannot be destroyed for the same reason.  The only rebuttal is if conscious states accidentally, non-causally, and inexplicably arise from conscious states. 

************************************

Then, yesterday, I had some more insights and was able to prove – yes, logically prove – that consciousness is indeed causal.  The following is not well organized or edited, but the insight was so important and poignant that I wanted to just post what I’d written:

My conscious awareness/identity is nonlocal, which means that it is not contained inside (or entirely produced by) my body/brain.  But the physical sensations I receive (via my body) are certainly localized, and if my consciousness is causal (and/or I have free will), then the effects are also localized.  Couple thoughts:

·         When I have a certain conscious experience (e.g., I feel happy), that is a fact that gets embedded in the history of the universe. 

·         Let’s say someone wants to kill me… i.e., permanently end my consciousness.  Let’s say he tries to destroy my body (although the body itself does not have an identity… it has lots of parts that are put together… so “destroying” would simply mean adding more information by way of, e.g., slicing up the body parts).  Certainly the body affects my conscious experience, but if we already know that conscious identity is NOT localized in the body (or anywhere), then why would adding local information to a physical system somehow destroy conscious identity?

It seems to me right now that consciousness is causal… but causation also implies a time delay.  When I experience sensations at state C2, that state may affect subsequent physical states S3, S4, etc.  This seems to imply:

·         Not every aspect of my conscious state needs to be physically embedded, and those that are not (which are still facts about the universe) must still be permanently embedded in the universe.

·         That “leading edge” of C2 can’t yet be physically embedded, but because it’s a fact about the universe (my experience of feeling happy at C2 is a fact), it cannot be destroyed.

In my proof, I assumed locality in a thought experiment to show that two spacelike separated instances of C1 that evolved to C2 and different C2’ results in a contradiction, because the local C2 could not also experience C2’ (which is outside of its light cone).  Thus I proved that consciousness is nonlocal.  But doesn’t that imply that my (nonlocal) identity could experience stimuli from spacelike separated sources/bodies?  For example, if my earthly body was destroyed, there’s no problem with the thought that I might still perceive stimuli from a different body (even light-years away).  Or what about perceiving information from far away?  Either: conscious experience is not subject to special relativity; or (the information in) any conscious experiences that get embedded locally must still somehow be delayed by special relativity; or something like that.

Assume consciousness is non-causal... then:

·         My experiences are inconsequential.

·         The evidence of (the facts about) all my feelings/experiences are already embedded in the physical state of the universe.

I showed that conscious identity is nonlocal, and I know (but can’t prove) that I have an identity.

Clearly there is something physically fundamental about conscious identity.  The reason that S7àC7, S8àC8, S9àC7 isn’t possible is because:

·         I have been assuming that C arises from an underlying physical state, but I think C is part of that state…

·         If S7 and S9 can both create the same C7, then that implies that some subset of S creates C, but I’ve already shown that’s not true.  Therefore my consciousness literally depends nonlocally on the entire universe, including facts/events the information of which are spacelike separated from me!

·         But wait… imagine the result of some quantum outcome could produce S9 or S9’… if they would produce the same C9 then, again, C would not depend on that part of S, so C would depend on a subset of S, which is false.  But that means (assuming that S creates C) that every single event/fact in the universe must produce/correlate to a different C state.  But that’s ludicrous.  Yes, at my current state in life, there are lots of possible consciously distinct states that I could have experienced, but that’s much much MUCH smaller than the amount of information in the universe.  (This is what Aaronson doesn’t realize… the idea of consciously distinct states is a fantastic insight.)

·         Therefore, the assumption (that S creates C) is false.

Assume C is non-causal.  Then C is caused or arises (inexplicably, emergently, accidentally) entirely from S – that is, there is nothing else involved in creating a conscious state/experience other than an underlying physical state.  I proved that C cannot arise from any local region of S, which means that it must arise nonlocally.  But I showed above (… because the amount of info in the universe – i.e., the number of possible states it could have been in – vastly exceeds the number of possible distinct conscious states that I can experience…) that C cannot nonlocally depend on S.  So if C can neither locally nor nonlocally arise from S (which are the only two possibilities), then we have a contradiction, so the assumption (that C is non-causal) is false.

Wow!

Therefore, the way I consciously experience the world indeed affects not only my next conscious experience but also the physical world!  (Be careful here… I don’t want to say that C is nonphysical… simply that when I refer to physical state S, I am referring to the world as physics currently describes it, which does not include a “consciousness particle” or whatever.)

So that means… I am having conscious experience C1… the physical world changes (e.g., new quantum information, although I don’t think new information is necessary… even deterministic physical evolution could, I think, result in physical stimuli that I consciously register as “new”)… I get stimuli from my senses… C1 and those new sensations cause conscious experience C2… but C is part of the causal physical world, so C2 (at least sometimes) causes changes in S3, etc.

But if I am now considering C to be part of the causal physical world (comprising C+S), then maybe not all conscious experiences have to immediately (or ever?) manifest themselves in S:

·         Indeed, if C is causal, then (as I pointed out previously) there are at least some conscious experiences on the “leading edge” that can’t yet be physically embedded in S.

·         C cannot permanently end because there is always (at least) that “leading edge” of conscious experience, which is a fact about the universe, that cannot yet be embedded into the physical structure of the universe.

·         What does this tell me about free will?

·         Are there some conscious experiences (like thinking inside a sensory deprivation tank) that I can (measurably?) experience long before they affect and/or correlate to S?

I wonder if that “leading edge” (those conscious experiences that haven’t yet, or can’t have yet, embedded in S) has anything to do with my experience of “now.”

I am in state C1.  It will cause next state C2, but that experience will also depend on physical sensations, and those will depend on S1, which was caused by S0 and any new information.  That new information could be quantum, but it could also be new information caused by my experience at C0.  (This also implies that my experiences and their effects – including my decisions if I have free will – on the physical world are not physically predictable; if they were, then S would evolve independently of C and C would not be causal.)

 

Let’s look more closely at the “leading edge,” which is my current conscious state (although my current conscious state also embeds my entire conscious history…)…

Let’s first look only at consciously distinguishable stimulus frames… say there are N of them (a billion?)… My next state C2 is created by C1 as well as (my experience/perception of?) the next stimulus frame, of which there were N possibilities.  (Although how many were actual possibilities?  Those stimulus frames were caused by S1 and any new information, and realistically they could have produced only a small fraction of N.)  Let’s imagine that the situation is such that S2, with certainty, produces only 1 possible stimulus frame. 

I guess what I’m trying to figure out is: how many possible (different) conscious experiences could I have at C2, and (since C2 is causal) how much information it could insert into physical state S2?

The simplest example of this would be… a psychologist showing someone pictures and then asking the patient about their emotional reaction.  Let’s assume for simplicity that there’s only one possible consciously distinguishable stimulus frame for the patient (looking at a photo).  (I need to be careful… even if only one frame is physically possible, C doesn’t “know” this… from C’s perspective there could be lots of possibilities… OK maybe <<N, but maybe the patient doesn’t know what the next photo will be… there are lots of possibilities, and each could then elicit lots of different conscious states/experiences.)

Now I understand that (because C is causal), the patient’s conscious experience must (at least sometimes) provide new information to S that is fundamentally unpredictable (by the psychologist or even anyone who “perfectly” knows S).

I am trying to figure out how much new information could be “contained” in a conscious state… or how much of that does or could causally embed in a physical state.  Let’s say that a certain conscious experience could embed Z bits to S… maybe it actually embeds something less than Z bits… maybe it doesn’t do it all at once… maybe it slowly adds that information… but then how could a state in the distant past, which doesn’t exist anymore, add information?[2]  Anyway, Z might even be as low as 1… maybe just a yes or no situation/decision.  (And maybe some conscious states would have Z=0, but at least sometimes Z would have to be at least 1.)

By the way, it seems that… if the total information Z that a conscious state could embed into S is ever more than what it actually embeds, then the conscious state itself will forever carry along that remaining information, causing the person to be forever physically unpredictable.

The way the patient (in state C1) will react to the photo (to create C2) will depend on C1 (and the stimulus frame… the photo… of course).  As I explained in the paragraph above, the patient is probably thinking about something (or having some conscious experience) that is not (perfectly) predictable.  But even if all Z bits from a person’s every conscious experience get embedded in S, there is still a lag (because of causality).  So, because Z bits from C1 could not yet have embedded in S1 (if they ever will at all), the patient’s reaction will depend on C1, which was physically unpredictable.  Said colloquially, there is no possible way for the psychologist to know (exactly) what the patient was thinking about (as the only way to know exactly is to BE the patient) when he shows him the photo, so the patient’s reaction to (and subsequent response about) the photo, which depends at least in part on what he was already thinking, provides new information.  My question is… how MUCH information?

I’m starting to wonder now if maybe C1 is, or at least carries with it, all of the information it could have embedded (but didn’t?) into S.  That would certainly help to explain why the “now” is so clear while distant memories seem to fade.

Hmmm… so given a particular conscious state AND a given stimulus frame, how many possible bits Z could be added to S?  I have no idea.

I just realized that I that my proof that consciousness is nonlocal is really a proof that consciousness is not physically local… because that proof depended on the assumption that C supervenes on S… that C is non-causal.  I now don’t know whether C is local, but I know it’s causal, which is much more powerful!

---

[1] If consciousness is not causal, then there can be no evolutionary pressures to select for either consciousness or a belief in free will, both of which seem to be essentially universal among humans.

[2] Well… maybe it causes an event in S, but it takes time for that to come back and make a difference to the consciously distinguishable stimulus frames of the patient, psychologist, etc.