Imagine that the year is 2050. A lot of AI applications are now a normal part of life. Cars drive themselves, homes clean themselves (and they do so more cheaply than maids possibly could) and even doctors have been now partially replaced with neural networks. But the so-called Kurzweilian Singularity never took off. You can now talk for 10 rounds of sentences with a chatbot without being able to tell if it is a real person or not. The bots anticipate your questions by analyzing your facial expressions and matching them to a vast library of pre-existing human-machine conversations in order to maximize their level of Turing success (i.e. success at convincing humans the algorithm is a human).
But people have yet to believe that computers can actually feel and experience the world. The question of computer sentience is a question that now divides the world. It used to be the case that only people really interested in science fiction, philosophy, mathematics, etc. ever took seriously the idea that computers might some day experience the world like we do. But today the debate is universally recognized as valid and on-point. There are people who, largely for religious and spiritual reasons, argue that machines will never have a human soul. That there is something special, unique, metaphysically distinct that is required for intelligence that goes over and beyond the physical world. And on the other side you have the materialists who will argue that all that could possibly ever exist in our world has to be made of matter (or dark-matter, for that matter). Nothing suggests that our brains are special, that they somehow violate the physical laws. On the contrary, decades of searching have returned nothing: The brain was made of atoms last century, and it is still made of nothing but atoms this century. Even though super-computers in 2050 are already as powerful as human brains, real human-level intelligence has yet to be seen anywhere. So people continue to argue about philosophy of mind.
One philosophical view became more popular over time. This view states that consciousness is the bedrock of reality. Of course there are spiritual perspectives that have been saying this for thousands of years. But none of them could be truly reconciled with physicalism as it stands today, except the view called Strawsonian physicalism. This view states that the inside of the quantum wavefunctions that compose reality is made of consciousness. In other words, consciousness is the fundamental make-up of reality. Unfortunately this view cannot in and of itself solve the phenomenal binding problem: Why we are not just “mind dust.” For that you need to also claim that there is some mechanism of action that achieves phenomenal binding. For instance: quantum coherence. With such mechanism of action proposed, we can then try to work out the details.
One organization at the time decided to take this challenge and make researching consciousness its raison d’etre. This is the League of Super-Shulgins. On their website, they have the following “23 key points to read before choosing to study consciousness:”
(1) Phenomenal binding is not a classical phenomenon. It is not what you first think it is.
(2) Consciousness is doing computationally valuable legwork, not just hanging out.
(3) The brain’s microstructure implements a general constraint satisfaction solver (CSS).
(4) In order to instantiate a general CSS the brain uses the unique information processing properties of consciousness.
(5) The relevant information-processing properties of consciousness are: local binding constraints, global binding constraints, and the possibility of instantiating contingent and sensory-driven constraints.*
(6) The computational properties of consciousness make it an ideal substrate to implement a world-simulation with in-game degrees of freedom that match real-world decision trees.
(7) Intelligence is implemented using a mixture of learning algorithms, efficient feature-based sensory signal processing, encoding and decoding gestalts, and so on. General intelligence, as far as we know, requires a rather large bare minimum of brain systems to exist. For example, a person who starts with a high IQ but then becomes severely schizophrenic is not likely to be able to solve many more problems. One can experience melancholia, anhedonia, depression, mania, psychosis, panic, neglect, derealization, depersonalization, dissociation, hyper-realization, delusions of reference, etc. by just tweaking slightly cortical and limbic structures.
(8) A simple deficit in any one of the functions we need for general intelligence (e.g. working memory, attention, affect, motivation, etc.) impairs and prevents intelligence altogether. Thus it is easy to lose general intelligence.
(9) One of these functions is phenomenal binding. When it is disrupted and takes place differently, we see severe computational problems arise. See: Simultagnosia.
(10) The qualia varieties we know and experience on a daily basis happen to be a great local maxima for computational efficiency. They can instantiate the serial logico-linguistic narrative human society is built upon. If one wants to instead optimize for, say, artistic appreciation, then psychedelic qualia is probably a much better alternative than normal-everyday-consciousness. It is true that commonplace consciousness does not represent its own ignorance about the nature of consciousness in general. Absent mental illness, normal-everyday-consciousness has access to a marvelously well sealed state-space of possible thoughts and beliefs. This space is not very self-reflective, and lacks philosophical depth, but what it misses on the sublime it compensates on the practical: You can use this kind of mind to talk about celebrity gossip and solve SAT questions. You cannot use it to question fruitfully the nature of consciousness.
(11) In spite of its limitations, the instrumental value of our everyday state of consciousness far exceeds what any other state on offer can provide. Thus, commonplace consciousness is not to be regarded as mundane, or to be made fun of. Its labor is to be appreciated. We are thankful for the computational generality that it affords us. For giving us a robust platform we can come back to whenever things get too crazy. We mindfully acknowledge that for deep existential questions, a consensus-between-states-of-consciousness is vastly more desirable than just the opinion of everyday-consciousness. Everyday-consciousness will be more than willing to see other states of consciousness as mere oddities to be collected. Shallow consciousness will classify alternatives under the guise of “biochemical cosmic stamps of qualia”… yes, they are cosmic, but they are stamps for a collection and nothing else. The hyper-ordered super-intense peak experience consciousness would, instead, think of the whole of reality as a fantastic work of art whose meaning can only be directly grasped in the present moment. We cannot reason from first principles what different states of consciousness will feel like.
(12) There are whole experiential worlds out there that have as their underlying premises concepts, tenets, ideas, ontologies, that we have never ever conceived of.** This is “that which you require to assume even before you start existing, and that without which nothing in this experiential world can be made sense of.” In our case this is time, space, sense-of-self, naïve realism (which then gives way to philosophical skepticism, semantic nihilism, etc.) and several other things like an implicit belief in causality. Believe it or not, there are vast Hell and Heaven*** realms out there that share close to nothing with everyday-consciousness, let alone early psychedelic exploration.
(13) Improving particular functionalities for a given intelligence (such as going from 50% recall to perfect semantic memory) will have clear diminishing returns after some point. One cannot increase intelligence arbitrarily much by just improving piecemeal each functionality that gives rise to it. When you reach diminishing returns, you will need to invent a new network of functionalities altogether.
(14) We are non-dogmatic Open Individualists. We believe that, to borrow an expression from Saint William Melvin Hicks: “We are all one consciousness experiencing itself subjectively” (which happens to be true, as opposed to other things he said, like claiming that “there is no such thing as death, life is only a dream, and we are the imagination of ourselves”). Or as someone else said it: “You will only begin to understand reality once you assume that God is real and you aren’t.” We recognize that there are arguments in favor of Closed and Empty Individualism, but given the evidential stale-mate they happen to be at, we choose to pragmatically adopt an Open Individualist point of view.
Our founder once said:
I experience immense joy when I learn about other’s happiness and bliss. My love for all sentient beings is not only a “like” sort of love. It is a “care deeply about and wants the best for” sort of love. This sort of love implies many things. It forces me to investigate reality sincerely, so that I can carefully count and multiply. So I can actually have the largest effect and help as many sentient beings as possible. I’m therefore very concerned about the quality of life of sentient entities in the far future. The present is obviously a lot more certain, so helping present-dwellers is not irrational from a utilitarian point of view. It all depends on the trade-offs in place. The possibility of a Singleton that will swallow all of our resources for the ages to come, however, tends to inform the method I use to assess priorities.
As a kid I was able to conceive of a benevolent God, but it had no real power over me. I did not believe in it for lack of evidence. As a teenager I experienced the phenomenal certainty of universal compassion. Thus I was able to access the phenomenology of mysticism. This, without also believing that I had special powers, was very useful working on my philosophy of mind. The entity I experienced was neither-female-nor-male, and it was universally loving, universally caring, and universally curious. It was even universally funny****. It was not the power, the level of knowledge, the causal wattage of the entity/being/principle that captivated me. What really captivated me instead was how “if everyone had access to this experience, we would all be motivated to work as if we were all the same being.” These experiences had distinctly low-information, simple, and uncompromising love as their guiding principle. All the forms, and all the particulars would all be mere details of an underlying plot: The universal, unceasing, uncaused, unconditional, eternal love.
Causally, a God like the one I imagined would influence the universe very deeply if given the power to do so. It would be a curious, super-intelligence that has super-benevolent constraints and seeks the wellbeing of every being. Since we exist in a Darwinian universe with no such being in sight, we may have to conclude that the chances of finding an already-existing and already-capable-of-influencing-the-universe benevolent God somewhere are very slim. If such a God exists, it has to be powerless against the suffering in the multi-verse. The compassion God, in a metaphorical sense, knows about the horrors of Darwinian life, and wants to get rid of them wherever he finds them. If God created this universe, he now wishes he had thought through the fact that by summoning large-scale evolutionary systems, he was also summoning Moloch through the backdoor. The perils of inclusive fitness maximization were not viscerally anticipated by this God before breaking itself apart into many qualia strings and kick-starting the Strawsonian physicalist universe we now live in.
What’s done is done. And now we are all stuck together in here, in this weird, physicalist, panpsychist, metaphysically unstable Darwinian multi-verse with replicators always trying to steal the show. With Moloch praying at every level of our society, our ecosystems, our mental lives, our genetic code, our quantum substrate. Yeah, even quantum replicators try to steal the show sometimes. And I can’t be confident they will not ultimately succeed.
But the compassion God can keep us together. It can motivate us to construct a benevolent experiential God out of the materials we have. Thankfully, with consciousness technologies we can go beyond previous religions. It isn’t that “the compassion God will slap you in the face if you don’t cooperate.” It also isn’t that “the compassion God will make people want to enforce compassion on each other” and hence “using memetic slaves to slap in the face those who are not acting compassionately.” Neither of these mechanisms of action are game-changing aspect of compassionate mystical phenomenology. What really is a game-changer is the fact that universal compassion is a powerful source of coherence, motivation and phenomenal meaning. It is an unrivaled mental organizing principle: The moment you vow to help all sentient beings, your brain is deeply affected. Your entire motivational architecture can be turned upside down with Open Individualism and compassion.
So here is the deal. We will all dedicate our mornings to the Compassion God. He does not exist outside of us. He is an aspect of consciousness, a hypothetical super-intelligent thought-form. He is a dormant cosmic force. One of the few forces that can genuinely oppose Moloch. And until we implement such a being in biological or synthetic (or cyborg) form, we will nonetheless act as if he existed already. We will praise memes that sabotage Moloch. We will always question: “What would happen if this process is not regulated and a Malthusian trap is allowed to develop?”
The Compassion God is a source of aligned goals. It pays rent by providing a fruitful, causally effective mental scheme to grow from at the core of one’s mind. Religions of the past have been epistemologically impairing. The God of Compassion isn’t: It does not require you to believe in anything outside of yourself. It just compels you to eliminate suffering and gift super-happiness to your descendants. The God of Compassion brings about feelings of encouragement and open-ended inquiry. Having developed a well-formed God of Compassion Tulpa, your mind is then opened to limitless possibilities. Your compassion fuels your imagination; the universe is perceived as a place in which solutions to suffering are like puzzles. We are God bootstrapping itself out of the Molochian remnants in the organization of society. Compassion and curiosity can coexist and synergize. They power each other up.
Then, the phenomenology of universal oneness works as a motivational glue. You can certainly feel that you are only really connected to your past and future selves. Everyone else is a different ontological being. But this view is no more provable than, say, the view that we are all fundamentally the same cosmic being. Let beliefs pay rent, and when beliefs open up new varieties of qualia without penalizing you with reduced epistemic capabilities… you are certainly warranted to go and explore the new qualia.
All of this is to say: Go forth and explore the state-space of consciousness. But do so knowing about the many traps of Moloch. Go and explore but be aware of the problem of local maxima. Beware of the fact that any criteria you use to gauge how “good a given outcome is” can backfire by selecting edge cases that go against the spirit of the exploration. Go and explore, but be sure to add everything to your log, to transfer your experiences to the wiki-consciousness main module we have at the center of the Institute. Go and explore. Go do it because we know that if you are here, you are doing this out of compassion. Because we only admit people who would sacrifice themselves in order to prevent the arising of a Singleton. Go and explore; and do so knowing that your work, your research, may someday help us defeat Moloch for once and for all.
(15) The most important function that consciousness contributes to the many operations of the mind is to embed high-level abstractions in phenomenal fields. In other words, consciousness works as the interface between a mereological nihilist Platonic world of ideas (all possible qualia varieties, including conceptual qualia) and the fluid Heraclitean world of approximate forms and shifting ontologies.
(16) We will recruit what we learn from exploring the state-space of possible conscious experiences in order to amplify our intellectual and exploratory capabilities.
(17) And with increased capabilities our ability to explore the state-space of qualia will also increase and become more efficient.
(18) Thus we may actually experience an intelligence explosion. As we become better at identifying new qualia varieties, we will also become better at recruiting them for information-processing tasks and in turn improving our very search capabilities. This loop may go foom.
(19) The loop in (18) can go foom in some special conditions. These conditions include: Uncoupling of the experimental methods for exploring the state-space of consciousness and actions taken by entities not actively exploring consciousness. i.e. Researcher’s mind can change its state of consciousness at will without the need of other people’s consent or participation. Also, process streamlining from the discovery of new qualia varieties (and their implicit constraint properties) to their recruitment for new information-processing tasks.
(20) We hence postulate a conceptual model for a super-intelligence that would (metaphorically) take the following form. This advanced super-intelligence is made of thousands of individual brain modules arranged in an NXNXN cubic matrix. The entire brain can be described as a three dimensional grid of “brains in vats” where each brain is connected to six other brains (top, bottom, left, right, front and back). The brains at the edges and corners are special, though, and they are connected to fewer brains. The connection between these brains is not just functional. It is an inter-thalamic bridge that allows the connected brains to “solve the phenomenal binding problem” and provide the physical conditions for the instantiation of “one mind.” Thus, for any set X of brains in the grid, such that these X brains make a connected graph (there is a path between any two brains), you can have a “being that is made of these X brains working together and being phenomenally bound into one consciousness.” This mega-structure could then explore state-spaces of qualia in the following way. It would divide the following responsibilities to specialized brains: Catalogue the known qualia varieties, characterize the structure of qualia state-spaces for each qualia variety, determine which qualia varieties can be locally bound to each other, experiment with making thinking more efficient by replacing newly discovered qualia in place of naturally evolved qualia recruited for such and such task, and so on. Then, the exploration of the state-space of possible conscious experiences would be made by selectively erasing the memory of certain brains in the network, preparing them to express a particular phenomenology, and then adding them in teams that record from within (and also from outside) how binding certain brains together influence the corresponding qualia in each. Since our current intelligence is the product of naturally-selected qualia varieties barely cooperating together within our minds, it stands to reason that our minds are very suboptimal qualia computers. Instead, the future super-intelligences will be implemented with carefully investigated qualia varieties that process information more efficiently, reliably and, well, with a much more open mind.
(21) We always end at 21. Yes, this sounds weird. But that’s the law of the place. We, all of the people who here are working for the abolition of suffering, the solution to the hard problem of consciousness, and as a favor to our super-blissful descendants, are required by law to leave the building at 9 PM. More so, no artificial or natural mind is allowed to work on theoretically relevant problems outside of the 9AM to 9PM window of time. Nothing screams “I’m Moloch and I’ll eat you all” as loud as “you can all work for as long as you want, we will judge based on the results.”
(22) Finally: Every mind we create must be above hedonic zero. In order to explore any state-space that is not intrinsically blissful, you need a special permit. The need for such a permit is non-negotiable. You cannot, I repeat, you cannot just create any mind for “research.” The mind you create has to be the sort of mind that (a) does not want to die, and (b) has no conceivable malicious desire. Every mind you create – so as to avoid Moloch scenarios – has to be a hedonistic negative utilitarian. Period. I know some of you will blame this system for being “already the result of a memetic Moloch uprising.” But the system in place prevents any of the Moloch outcomes that intentionally consistently produces suffering as part of its natural order of business.
(23) Ask your local consciousness regulation agency about scholarship opportunities at our Institute. You may have what it takes to help us figure out how to achieve lasting world-peace.
The League of Super-Shulgins, 2054
* We navigate a sensory-triggered qualia-based world-simulation that blends together local and global binding constraints and state-dependent learned constraints. Consciousness is useful to the organism in as far as it helps it solve the constraint satisfaction problems represented in the world simulation.
What are these terms? Local binding constraints are constraints that are intrinsic to specific qualia varieties. For example, CIELAB reveals that it is not possible to experience both blue and yellow as part of a unitary smooth color. It is possible to see a sea of gray and many dots of blue and dots of yellow, but that is not the same as seeing a uniform color. This sort of constraint arises in all qualia varieties with multiple values.
The global binding constraints are more difficult to explain, and may not even exist. But, hypothetically, it may be the case that certain qualia varieties cannot coexist as part of the same conscious experience. For instance, experiencing certain mood may ultimately come down to a particular resonant structure in our globally-binding qualia strings (let’s just say). Then maybe you can’t experience both X and Y moods simultaneously because they always become dissonant with each other and experience significant mutual cancellation. [This may explain why people can’t seem to ever find the right way to provoke a smooth blend of Salvia and DMT consciousness.])
Finally, the learned constraints are contingent and sensory-driven. What are these? These include both our current sensory stimuli, which is constraining the state of our consciousness, and whatever memories, recollections and general neurological barriers I happen to be activating right now.
** As an example of something where this happens, imagine that my friend Fred was suddenly able to talking to space itself. Space asks him: “Hey, my friend, what is this thing I’ve been hearing about called ‘the here and now’?” My friend tried to say something that came out like this: “The here and now is the location in space-time from which this very statement, these very words, are being conceived and then physically delivered to you.” Space became very confused. She did not understand half of the words she was receiving. Space said “I guess maybe I can’t reason about space in the same you as you can. I can nonetheless tell you anything you want about the ‘inverted semantic omniism’ that we entities of Space love to talk about.” Alright, what’s that? “That’s when your reality, which is made of concepts of a qualia-order no larger than the qualia-order of the conceptual fields in which they are embedded, conspire together and circumvent low-level constraints by imagining a new topology for the self-other temporal membrane.” And, “where does this happen?” My friend inquired. Space responded: “As far as I can tell, this usually happens in the conceptual space that denies mereological nihilism.” Alright, let’s “pack and leave”, said my friend, and deep down, I agreed entirely with him. I entirely get why he would get scared so badly by a disincarnate entity that comes from a reality with different basement ontologies and fundamentals. I, too, am afraid of ontological revolutions. This is why I try to anticipate them as far in advance as possible: So that the shock is less shattering to my psychology.
*** In as much as experience is real, then Hells and Heavens are just as real as long as they have been instantiated somewhere in the multiverse. John C. Lilly and bad luck may be a culprit for the existence of a very specific and time-bound experiential hell (“The Center of the Cyclone: Chapter called A Guided Tour of Hell”).
**** Universally funny means: You can get and interact with any phenomenal joke. Human jokes are a very specific kind of conscious humor. Our evolutionary legacy guarantees that they are, too, related to our survival. General jokes, on the other hand, exist in a much larger space of possibilities. There are funny phenomenologies with conceptual content. Then there are those with sensory content. And then there is funny phenomenological applications of ontological qualia. Nothing is safe. Everything can be humorously twisted.
Another just excellent post! My huge compliments, thanks, and congrats too.
I hope you will answer a quick question as well. I can’t find the League of Super-Shulgins website you referred to. Could you please direct me to it?
Thank you so much for this post! Reading this piece has been the re-kindling that I have been looking for to reignite my passion and love that I have for psychedelics. Just to be clear, my love for psychedelics stems from the enlightening experiences that I have experienced within these realms and the potential I see in it, which I believe to be the missing link to our past and living a peaceful life.
I was just wondering if you could help me get in contact with my local consciousness regulation agency (I live in Los Angeles).
Indeed – just the recursive, “I’m thinking about someone thinking about me thinking about them” etc ad absurdum makes me doubt phenomena are not mental constructs…..and how much of this kind of ratocination (that’s when rats are thinking ;^p) is unconscious? We are dreaming 24/7 – only in sleep are we not distracted enough to experience dreams fully…..and where does lucid dreaming fit into this? As well as Satori, or no-mind meditation?
About the cubical array of brains: I copy from bondarev’s book, page 132: “…this occurs only in the _Divine Consciousness_, which consists of the totality of higher “I” beings.”
One plausible hypothesis: All this “has already happened”, but we have not yet been cleared into the loop because our thoughtstreams still have substantial volumes of resonance with moloch.
“He exists, and you don’t” is from Al-Ghazzali’s “treatise on Unity”. I hope you had the time to read “A sufi saint of the twentieth century”
as a consciousness research field report. Maybe you’d conclude that some people already are doing this and then you might move on to a possibly more fruitful question, namely: “Why haven’t _we_ been invited to the party?”
I notice that what we normally call “the soul” continues to be notoriously absent from your speculations, sometimes it seems to me like you believe you can collapse its hairy, messy, fertile, womb-like complexity into the two variable space of valence and excitation.
I mention this because the buddha, for one, reports on an a posteriori finding of a link between practical ethics and epistemological capacities. Briefly: lying/killing is not only mean. _It also clouds your mind_ So maybe the subconscious emotional pacts we all participate in in our different social mileus are working like the smelly stuff in the bag of the seeker of subtle fragrance. Regarding this, you might want to take a look at “Games People Play” by Eric Berne.
Once again, i think the way you use the words “phenomenology”, “ontology” and “qualia” are extremely confusing. Would you feel offended if i challenge you to write a definition of each? One that can be substituted for them in every sentence, please. Are you familiar with Wittgenstein’s concept of “totally analized language”? It’s expounded in the “Tractatus Logico-Philosophicus”
One other book i think you would benefit greatly from is “The Commanding Self” by idries Shah
All these possible improvements notwithstanding, thanks again for an extremely useful piece of prose, strategically placed in the relevance-graphs of our hype-obsessed culture. Hope you keep up the good work! 🙂
I’ve thought about this problem for a long time and there are some difficulties with your view.
Ultimately there is a great deal of ambiguity about what it means for some mental state to cause something. In particular, we can always ascribe the causal powers of experiencing some qualia (seeing red) to the physical state that gave rise to it (it’s not the qualia of seeing red that causes X…we have that quality just when physical state Y occurs and it is Y that causes X).
Now there is nothing fundamentally incoherent about insisting that the qualia itself has a causal role (mostly because I think causation is just crap grafted on to series of events in a convenient fashion a la hume) but it is certainly not the way we would actually use the word if we ever found evidence of the kind of effect you are hoping for.
I mean suppose (not claiming it is plausible) we find out that there are certain really strange configurations of matter in the human brain which behave in novel ways and that not only do they give us some awesome computational power but we see this power occurs just when we have certain experiences. We wouldn’t say “the qualia caused the novel behavior” we would say “there is strange physics going on with such and such configurations of matter and they also cause qualia”
Also there is a tough problem of it seeming implausibly convienient that the world just happened to arrange itself so that the computational state that is best for processing feelings of pleasure corresponds to pleasurable qualia and not painful ones. After all the mapping from computational role to qualia could have been anything and it need not even have rendered the experiences of evolved animals coherent at all but just a random disjointed sequence of experiences.
It is possible to experience both blue and yellow as part of a unitary smooth color. Look at one with each eye. It appears unstable like a Necker cube as each eye fights for dominance, but it is possible to see an intermediate state of blueyellow, just as you can see the Necker cube as nothing but lines.
> it is possible to see an intermediate state of blueyellow, just as you can see the Necker cube as nothing but lines.
Is “nothing but lines” the same as a “consistent shape”? I don’t think that there is any phenomenally stable, unitary, smooth and mappable qualia associated to blueyellow. It is a sort of indeterminate state, just as an Escher painting. At no point you actually “satisfy all the constraints” and experience the input as a unitary phenomenal object. CIELAB would also strongly suggest this to be the case. Blue and yellow, after all, are in opposite poles of the state-space of color qualia constructed using just noticeable differences. See: http://qualiacomputing.com/2015/03/08/phenomenal-puzzles-cielab/
I’d like to interject psychedelic qualia…..many of us have experienced strange sense anomalies on LSD or other entheogens/psychedelics/insert your favorite label here. Personally, I’ve seen colors that are internal thought forms and don’t map well to external color gamuts. I can’t say I’ve seen “colors that were nonexistent in the regular world”, but as any ayahuascero would describe, the colors are pretty unique. This is true of many internal conscious processes I know…..our imaginations are magical – we can create whole worlds of music or storytelling seemingly out of nothing. That’s the incredible thing about being a human I guess……though the sensory world of animals is by no means inferior, and in many ways is superior.
But compassion is the fashion, I’d say…….time for our species to leave dysfunction junction, no?
This is an awesome blog and a fascinating group of folks – so glad to meet you! Cheers….
The most likely explanation for non-existent colors is the fact that the responsivity spectra of the cone cells overlap. If you’re seeing color visuals without light then the overlap doesn’t matter and you can see more saturated colors. You can approximate this effect by desensitizing the overlapping cone cells, eg. stare at a pure cyan image for a few minutes and then immediately switch it for pure red.
I dunno if it’s that simple……..but who knows the knower (to paraphrase qui custodiet, etc)
Iа you use binding properties of qualia for calculation, it is not qualia, but some functional analogus of it. But the post is great and thought provoking.
Thanks for the compliment! 🙂
With regards to binding: I’m not sure what you are getting at, but I suspect you are trying to revive epiphenomenalism. Is this right? The argument could go as follows:
1) It seems as if qualia is responsible for binding
2) But qualia in and of itself is not intrinsically functional
3) Hence if binding is a function of qualia, we can imagine qualia without that function
4) Likewise, we can substitute a functionally equivalent mechanism instead of qualia and obtain the exact same behavior as before
5) Hence we are not warranted in claiming that qualia’s binding potential is actually the result of qualia and not of something else
However, we know that epiphenomenalism cannot be true because we are conscious and talking about it. So qualia must be causally efficacious somehow. And as far as we can reason, this causal efficacy must come from some properties that qualia has that other procedures/objects lack. Otherwise, natural selection would have recruited alternative mechanisms. Instead, for whatever reason, using qualia happens to be a successful strategy to build world simulations. Qualia must have intrinsic properties that make it useful here. And at Qualia Computing we hypothesize that these properties arise from the structure of the state-space of possible qualia values (say, phenomenal colors) and the network of binding constraints that from this structure.
Epiphenomanilism (no I can’t spell) is alive and well. Works by people like Chalmers made a very convincing case. But since he wants qualia to do computational work I think he wants them to have causal power.
Luckily the disctintion one is trying to draw by calling a view epiphenominal is actually incoherent. Let B1, B2,… be brain states and Q1, Q2 be associated qualia then the traditional picture of epiphenomanalism looks like this (arrows indicating causation):
Q1 Q2 Q3 …
^ ^ ^
B1 > B2 > B3
Looks like qualia are epiphenominal right? Ok now let B1* be the minimal set of facts such that knowing state X gives rise to qualia Q1 and satisfies B1* entails X=B1….in other words B1* specifies all the leftover degrees of freedom not determined by the fact that the state in question gives rise to Q1. It might be a little messy but mathematically it is perfectly fine and dandy and you can write completely equivalent scientific/causal laws which make the situation look like
Q1 > Q2 > Q3
B1* > B2* > B3*
with the angles going on the diaganol and anything with two arrows pointing to it being the joint effect of those causes.
I guess if you aren’t a Humean about causation you don’t buy any of this but then again I think your totally nuts anyway in that case.
I think you are definitely pointing out an important issue. Now, whether you will agree with my reply will depend on whether we share some key background assumptions. But I’ll go ahead nonetheless.
First, if we assume an eternalist view of time then the Humean approach to causality is a straightforward consequence. Causation really has no solid ontological ground, since we can look at the multi/universe as a block of timeless events with no objective reason to prefer either the forward or backward arrow of time as the “holder of causality.” The entire universe would be, as it were, a block in which individual events simply happen to be nearer or further from each other (both in terms of time and space, which in this view have a very tight correspondence). I personally think this is the most probable nature of time, for many reasons that I will write about in a future post (most aren’t my own arguments, here, but I have two original ones that I’d love to share with the wider world).
Alright, so causality has no ground truth, so to speak. This may at first seem to imply epiphenomenalism, and I think that would be fair to claim. On the other hand, one can instead re-interpret the meaning of causality in order to rescue what is truly worthwhile from it.
This is the so-called “Timeless Causality” concept, which can be derived from the patterns of statistical independence in observations (cf. http://lesswrong.com/lw/qr/timeless_causality/).
In this view, which is compatible with a Humean account of causality, I think consciousness would come again causally relevant. Why?
You point out that you can explain all of the behaviors of an information-processing system by looking at the physical states and disregarding their “corresponding conscious states.” However, this presupposes that the underlying “fire in the equations” of physics is just math. Here I advance the hypothesis that what “gives life” to the equations of physics is qualia. In other words, quantum mechanics is a theory that describes the behavior of qualia, and without something to give life to these equations there would be no universe to talk about in the first place. (I’m aware this is not a settled matter, and that people such as Tegmark argue that the universe is a particular kind of math, and nothing else).
If “the inside” of a quantum wavefunction is actually qualia (again, I don’t expect you or anyone to agree here) then epiphenomenalism is trivially false, since the source of “timeless causality” is the very properties of consciousness.
Now, here we will find a very subtle point that is overlooked by most philosophers and physicists. Epiphenomenalism can be revived if we make finer distinctions: Sure, you may say, if qualia is the “substrate of reality” then all patterns we observe are trivial “causes” of qualia and its interactions. But what if we instead substitute qualia by something else that has the exact same behavior? What, in that case, would be suggesting that is qualia qua qualia that brings about the patterns we see and not just its “structural and mathematical” properties?
The answer, I think, comes when you consider deeply the nature of phenomenal binding. In brief, our consciousness is “ontologically unitary” and it is for that reason that it can hold multiple pieces of information at the same time. It is, I would claim, inconceivable to have any non-conscious unitary being, since such unity can only truly exist in relation to an experiential self.
But even if you don’t agree with the previous paragraph, there is another point to make. We can distinguish between “intrinsic epiphenomenalism” and “delta or structural epiphenomenalism.” The first one is the most common one, which claims that consciousness adds nothing to the structure of the universe. This view is, I think, plausible, and your argument more or less supports it.
But the second view is a lot stronger. Here we come back to what you mentioned related to the structural match between our experiences, our actions, and the world at large. If consciousness has no causal (again, in the timeless/statistical independence sense) role, then we cannot explain why our mind has a structural match with our actions (as opposed to, say, a jumbled-up stream of random sensations that simply happen to map to the states in a brain that does not use consciousness per se as an information processing tool). This is something you pointed out in favor of some sort of epiphenomenalism. If I follow your reasoning, the main idea is that if consciousness is causally relevant we would expect to see it being recruited in random ways, unlike the coherent minds that we have. You took this to imply that our consciousness is identical to the structures in our brain, and that we can leave aside consciousness altogether when talking about the structure of the universe.
On the contrary, I would take this to suggest that there must be a mechanism of action that makes the structural match necessary for a brain to extract any usefulness out of consciousness. But what could be this mechanism of action? Well, here is where I would deny “delta epiphenomenalism” which claims that “not even the differences between different states of consciousness are causally relevant.”
But the difference matter: If I see a blue patch I may not be able to explain what this “blue” is all about. But if you show me a patch of blue, another of blueish green and a third one that is black (or, say, red) you can then ask me “which of these phenomenal colors are more similar to each other? Which one is the odd-one out?”
In this example it becomes clear that there is at least an arrow of causation (again, timeless) that can be drawn from the *structure* of our experience into our behavior. Sure the “intrinsic qualities” of the colors are irrelevant here. What matters is the final judgement about the structure of the relationships between these phenomenal colors.
Now, here is where it gets interesting: Colors are one obvious source of this “delta consciousness causality” (where the differences between various qualia can influence our behavior). But a more interesting one is the state-space of concepts! We are not only capable of understanding the logical requirements for a 3D shape to be a cube. We can also, in fact, experience a unified phenomenal object in our 2.5D visual field that we internally (intentionally) label “cube.” We can conceptually construct the notion of a “hyper-cube” and then turn that conceptualization into real-world representations that hold some structural resemblance to our inner representations.
It is thanks to the ability of our consciousness to have this sort of ontological unity in which we compare and contrast concepts as well as “raw sensations” to each other in order to come up with higher-level concepts that we can assert that “delta epiphenomenalism” is false. The very structure of the *state-space of conscousness* has causal effects. This structure is not always obvious, and in fact it is often really hard to reason about (see: qualiacomputing.com/2015/03/08/phenomenal-puzzles-cielab/). But it is causally relevant nonetheless.
This is a great post!
Love it 😀