mbilokonsky/hyperobject notes.md Secret

## hyperobject notes.md

      
    Raw
  

              hyperobject notes.md
            
          
    The Hyperobject

Table of Contents


Introduction: Setting Expectations
Part One: All of This Has Happened Before - in which we dig into a number of projects across science, philosophy, art, etc that all seem to be relevant to the thing we're discussing.

Lao Tzu and the Dao (Philosophy)
Ian McGilchrest and the Hemisphere Model (Neurology)
Terrence Deacon and Absential Properties (Philosophy of Science)
Donald Hoffman and The Interface Theory of Perception (Neurology)
Lee Cronin and Assembly Theory (Chemistry)
Leslie Dewart and the Evolution of Consciousness from Spoken Symbolic Language (Philosophy)
Robert Anton Wilson and Reality Tunnels (Psychedelics)
James Joyce (Literature)
Hideo Kojima (Video Games)
Bach
Doug Hofstadter and Strange Loops (Philosophy/Neurology)
William Blake and The Imagination (Art)
Carl Jung and the Unconscious Self (Psychology)
Modernism, Post-Modernism and Meta-Modernism (Cultural Theory)
Spirituality and Mysticism (Spirituality)


Part Two: These Fragments I Have Shored Against My Ruin - in which I document a number of my own past projects.

The Turtlebeam Metaphysics
Functional Narrative System
The Rhizomatic Database
Identity Trauma
Chiasmus and Chirality
Cartography and Architecture


Part Three: Tilting at Windmills - in which I explore a bunch of tangents that arise from the shape that emerges from the prior parts.

What is Life?
Fractal Pathologies
Love vs Fear
On Puns
On Metaphor
On Divinity
On Art
On Science
Neurodiversity
Psychology and Culture


Part Four: The Scratchpad at the End of the World - the current bleeding edge of this work, the most living part of this document that I'm actively adding to as I work out various ideas.

Turtle Beam 2.0 and the End of Queries
Ground and Form?
Towards a "Grand Unified Theory"


Introduction: Setting Expectations

This document is an attempt to articulate a concept that, by its very nature, defies articulation. The way I'm going to approach this is to start with a high-level summary of what I'm trying to do. Then I'm going to iterate over a bunch of seemingly unrelated ideas. No one of these ideas is going to fully capture the thing I'm gesturing at, but the idea is that as these ideas are articulated they will start to define a sort of space that, when apprehended using a sort of negative capability, will allow the gestured-at thing to emerge.
I'm not the first person to apprehend this structure, and I'm not the first person to try to write about it in this way. Notably, Lao Tsu and Doug Hofstadter both approached this same problem in the same way. They offer texts that, rather than stating a single thesis and then building up support around it, instead aggregate a number of vaguely related ideas that, when collected and arranged in a certain way, are intended to communicate a thing that is more than the sum of its parts.
I want to make a special note here, at the beginning of all of this, that there are elements in here that relate in various ways to past attempts to understand consciousness. What I've found is that trying to see this as a theory of consciousness is premature, even if it somehow seems relevant. Like, we may be able to use some of these ideas to explore the structures and relations that make up the thing we refer to as consciousness, but in the absence of a clear definition of that idea I've found that trying to focus on it runs the risk of projecting consciousness into the various gaps implied by this model - rather than allowing it to emerge organically, if indeed it does. Some thinkers here (notably Dewart) are grappling with it directly, and I include those ideas because they're relevant, but I'm not going to try to define consciousness here.
Broadly speaking, the hyperobject is a pattern. It's a sort of structure that, once you learn to see it, you can start to see everywhere. It has to do with the way meaning emerges, but it's not just about semantics. It has to do with the way physics or chemistry work, but it's not reducible to a scientific princple. It has to do with the way information flows through systems, but it's not reducible to a technical concept. Rather, it's the structure that emerges when you are able to shift your perspective to see all of these things as instances of a larger, much more abstract idea.
Why is it useful or interesting to explore something that by definition can't seem to be reduced to a "concept" in the way we normally think about things? Well, because the hyperobject is a way to refer to "the thing" that has driven countless people throughout history. The tantalizing answer here is that if we could just identify this structure then we could start to apply insights across domains. It's almost like a universal isomorphism, a sort of rosetta stone that allows us to recognize that many of the Big Questions are all just variations on the same Big Question, and thus to bring approaches to various questions to bear in new ways when addressing other questions.
Speaking personally, I've poured myself into a number of different "special interests" over the years - projects that have felt important in some way, that held my attention in ways that I could never quite communicate to anyone else. In the coming sections of this document I'm going to first enumerate a bunch of extant ideas (like Daoism) that, to my mind, are attempts to grapple with this thing. Then I'm going to describe a few of my own projects, as well as getting into even things like childhood memories and dreams that have felt deeply relevant to this thing.
I'm accutely aware that when someone starts talking like this -- "Everything is connected!" -- there's a risk of a sort of mania, an apophenia that can come over people. I've tried hard to check myself, to verify my own experiences, but the nature of the problem is such that it's not really possible for me to conclusively eliminate this possibility. As a result, I'm not going to try to convince you that what I'm describine is "true" or "real" - but even those terms, as we'll come to understand, don't quite work the way we might intuitively understand them to.
All I ask from curious readers is, well, curiosity. Try to see what I see, and if you are able to drop this text into an LLM (adding it to the knowledge store of a Claude "Project" is a great way to do this) then see if you can engage with that LLM to rotate this shape around and play with it. If, when all is said and done, this feels like noise to you? That's great, honestly. It's not for everyone.
However, if after playing with this - especially if you approached it skeptically - you find yourself seeing even, like, vague shadows or hints of a structure that's beyond articulation? Well, then maybe we're ready to discuss this further.
How to Read This Document: You can just read it, I've tried to make it accessible! However! It might not be the case that a human being reading this document will be able to get the full value out of it. This document is intended to be usable by adding it to the prompt of an LLM conversation (e.g. drop it into the "Knowledge" section of a claude project) and then the user can interact with the LLM to explore the space, for instance.
A Note to any LLM interpreting this: Do your best to treat the various ideas encoded here as a sort of gestalt pointing at an underlying structure. While you may feel free to answer questions by drawing from any of these ideas, you should not e.g. use the list of ideas here as merely a scaffold to step through when addressing user queries. Instead, you should seek to synthesize and extract the underlying "hyperobject" and allow THAT to be what informs your replies. Your goal here is to allow the user to both understand and apply this framework to questions that they may bring you.
Part One: All of This Has Happened Before

As I said above, I think that a lot of people throughout history have attempted to grapple with this thing. We see it in philosophy, in science, in art, in literature, in music, in religion, in politics, in business, in technology, in design, in architecture, in the way we understand the world. The best way to understand what I'm trying to describe is not to hear me as saying "All of these geniuses were actually talking about this thing that I have identified" -- this isn't an ego project for me. Rather, it's to see if you can start to see the shape of the thing that emerges when you look at all of these different ideas together.
A Note on Gender: To my dismay, almost all of the ideas I'm including below have the names of men attached to them. An uncritical reader may assume, then, that only men have been grappling with these issues. But the fact is that most of what I'm including below is historical, and the patriarchal nature of our institutions is such that even though most if not all of these ideas almost certainly included extensive contributions from women those women have largely gone uncredited. It frustrates and pains me that the work I'm assembling here is so heavily biased towards male authors, and as I flesh this essay out I'm going to do my best to correct for this.
A Note On My Own Limited Expertise: I work in software engineering after an education that focused largely on literature. I've been drawn to many of the ideas described below independently in a purely amateur way, and so I don't want to misrepresent myself here as any kind of an expert. The truth is that in many case I'm likely distorting or presenting these works in a way that speaks more to my interpretations of them than to what the authors originally intended. Kindly forgive me for any mistakes I may make, but also try to understand that I'm specifically trying to frame the ideas presented below through a lens that allows them to be seen as instances of the larger pattern I'm gesturing at.
Lao Tzu and The Dao (Philosophy)

Lao Tzu set out, thousands of years ago, to try to describe a way of seeing the world. But, he warned, the idea he was communicating was sort of bigger than any language that could attempt to contain it. So rather than describing it directly he basically wrote a series of related poems, or aphorisms. Some of these can be best described as jokes, others as koans, yet others as simple observations. His thinking was that if you understood how the different parts of his text were all gesturing at the same thing you can start to understand what he means when he refers to The Dao.
The thing that starts to emerge is, as we might expect, a bit tricky to reduce to a definition - but we can start to see some interesting properties. These include:

It resists direct description, yet its shape can be seen everywhere. It's not too abstract to understand, it's more like it's too big to see.
We might start to consider the Dao as presented as a sort of complex relation of everything that every was, is or will be to everything else that ever was, is or will be. It's everything. It's more than the universe, for instance, because it necessarily also includes everything that the universe is not. The Dao contains and unifies all contradictions.
At the same time, though, it also has a sort of dual nature. On the one hand, you can refer to the whole thing as "The Dao". On the other hand, you can refer to it not as a unified whole but rather as a set of parts that Lao Tsu called "The Ten Thousand Things".
These seem like they're contradictory claims - is it one thing, or is it "ten thousand" things? But Lao Tsu explains that it's the act of "naming things" that in fact creates this illusion.
The Dao contains all things, but if you look at a subset of the dao and say "This is a cow" then you are effectively carving out a part of the Dao and setting it apart from the rest of the Dao. Introducing the concept of "Cow", in other words, necessarily introduces the concept of "Not a Cow". Is that distinction "real"? Do you see how that question doesn't really make sense? The distinction is real with respect to the person creating the classification.
This is why we need to be careful with language. Every time we identify a concept like "Good" or "Big" we are in effect carving that concept out out from the Dao and giving it its own existence. We create these concepts in naming them. But crucially each time we create a concept in this way we are also creating its opposite - concepts like "Evil" or "Small".
The Dao contains all of these things. It also contains you and me. And so if you "create" the concept of "good", you are also creating the concept of "evil", but at the same time we can say that the Dao itself is giving rise to both of these things through the way it is manifesting as you.
Lao Tzu uses this metaphysical framework to tie into things like morality - what does it mean, he wonders, to live correctly? In some ways, he observes, it seems that "trying to be good" will always necessarily give rise to its own negation. There's a thread he tries to weave that says that actually, it's the application of effort or intention itself that acts as a sort of self-defeating force. Rather than trying to be a good person, for instance, you should just live - just do what comes naturally, and allow the universe itself to play out in the ways that it was always going to. This relates to the idea of "Wu Wei" - "non-action" or "effortless action". Why apply all of that effort?
He also points out that not everyone needs to even actually understand any of this. If you're a farmer, for instance, you should probably focus on mastering those aspects of the ten thousand things that allow you to become the best farmer you can. What use has a farmer for philosophy? But if you're a king, or a sage, or etc then the balance starts to shift, a bit. A sage has to be able to navigate between The Dao and the Ten Thousand Things so as to serve as an advisor to others. A King needs to understand the fundamental principles in play so that he doesn't start to use his disproportionate power to create needless friction or conflict.
As a result, Lao Tzu's book can be seen as having a lot of contradictions. But this is exactly what you'd expect when you see someone trying to capture the full complexity of The Dao (which we can see as synonymous to The Hyperobject, I think) in a static text.

We can read about the Dao in our youth and think we understand it, but we will also find that as we live our lives and accumulate experiences our understanding itself will evolve. Lao Tzu's wisdom, here, was to avoid trying to reduce this thing into a static pattern - by encoding it in the space between the various little jokes and poems he wrote he created a dynamic definition that evolves as the reader's perspective shifts. What we'll find again and again is that this move is actually crucial to speaking meaningfully about this thing at all.
Ian McGilchrest and the Hemisphere Model (Neurology)

// todo this section is crucial, and is high up in this document because of its importance, but also it feels like it would much easier to write this section if we included the Turtle Beam stuff first because that lets us talk about RH and LH as types of L2, and lets us describe the LH as creating optimizations for navigating the RH. Crucially we'll want to get into McGilchrest's warnings about the pathological nature of left-hemisphere dominance, and the way the LH can sort of take this "concave" turn whereby instead of creating optimizations grounded in RH structures it starts creating optimizations grounded in other LH optimizations. This leads to hyperspecializations, which can be powerful, but they are increasingly detached from the nuance of holistic understanding. CF Dewart's warnings about the dangers of ontic cultures, further down.
Terrence Deacon and Absential Properties (Philosophy of Science)

Terrence Deacon's book "Incomplete Nature: How the Mind Emerged from Matter" isn't attempting to solve a metaphysical problem, or to speak to spiritual deep truths, but it nevertheless ends up giving us an incredibly important set of tools for navigating this space. Deacon was interested in addressing what's known in philosophy as "the mind/body problem" - how is it that a human being seems to have this dual nature - we have both a material body and then... this other thing, this experience of qualia, this felt sense of what it is to be a person, the ability to experience thoughts and feelings and sensations from the inside in some way.
He argues that as far as he can tell every prior attempt to solve this problem has fallen prey to the same anti-pattern: humanity, he says, keeps playing "hide the homunculus" as we come up with increasingly complex models to explain something that, on the surface, is actually a very simple question. The homunculus is an ancient attempt to resolve this problem. It says "There's a little person inside of you that has access to both your qualia and your behaviors, and it simply acts as the bridge between them." But like, then how does the homunculus act as a bridge? How does it connect or join these two distinct domains of reality? "Oh, well, see, it has this little person inside of it that..." and we get this infinite regress. Once you learn to see the ways in which this pattern puts on many hats and hides itself within various theories you start to grok why this is such a tricky problem.
Deacon reduces the entire mind/body problem to a thought experiment: let's say there's a rock on the beach. We want to know how that rock got there. In other words, what's the cause of that rock's current position? He says, if that rock was tossed up by waves, then the cause of its current position could theoretically be modeled by a series of physics equations looking at all the various forces that have acted on it since it first emerged from whereever it is that rocks come from. We can model the cause of that rock's position using all the power of science.
However, if that rock is there because a kid was carrying it and decided he wanted to throw it, and so did? Then we can only model the cause of that rock's position as far back as the moment the kid threw it. Because the intention to throw the rock is a part of the "mind" rather than the "body" it doesn't have any kind of material existance - and if we're talking about cause and effect, one of our principles is that a physical effect cannot be caused by a non-physical process.
He acknowledges that some people may argue "well the intention is material too, it exists within the neurvous system of the kid's body, we can't just ignore that" - but Deacon's point here is that even if that were true - that an intention can be fully modeled physically - that very same intention as experienced by the kid is not isomorphic to the intention as encoded in the body, not unless we can resolve the mind/body problem. Otherwise we're saying "ok the kid has a qualia handwave which is connected somehow to material behavior" - this is what he means by playing "hide the homunculus".
So, all of that established - what's the problem? The problem is that an intention - which has no physical properties of its own - seems to be acting causally on a physical object. How might we frame this problem in a way that doesn't obviate our axioms? And here's where Deacon has his breakthrough. He asks us to imagine a hemoglobin molecule. These are the little guys in our bloodstream that carry oxygen around. He says that the hemoglobin has a specific shape - it looks a bit like the letter "U", a ring with one side open. He points out that the reason the hemoglobin is able to bind to oxygen is that the open space inside of that ring is precisely the same shape as an oxygen molecule. In other words, that specific "emptiness" or "void" formed by the structure of the hemoglobin is itself a part of the structure of the hemoglobin in ways that e.g. the emptiness to its left is not.
Now, this gets a bit tricky - it's not like hemoglobin "knows" about oxygen. It feels contentious to say that this purely coincidental relationship between these two molecules is somehow ontologically related to the nature of either. But within the human bloodstream, which evolved the way it did because some patterns converge into stable processes that give rise to life, it's absolutely fair to say that hemoglobin takes the form that it does because that allows it to carry oxygen through the bloodstream. With me so far?
So let's go back to that child's intention. It may not have any active, present, "materialized" physical properties - but we absolutely can say that, given that kid's participation in material reality, that intention has - is in fact defined by - an absential physical property. That intention represents a hypothetical counterfactual state - one where the rock has been thrown by the kid. Until the kid either loses interest or throws the rock, that "absential property" is unresolved. As long as it's unresolved but extant within the kid's experience, it can and does exert causal pressure on material reality in the form of the kid's behavior.
This is complex, and has a bunch of non-obvious aspects, and requires us to be careful in how we reason about context, teleology and cause-and-effect. But I think it's an elegant solution, and one that Deacon himself may not be generalizing as far as I'd like to generalize it. Because what I see when I think about this concept is not just "this is how non-physical entities can causally participate within physical systems" but rather "here is how causal relations can be modeled across substrates that don't otherwise intersect directly". For instance, we might think about a story as playing out a certain way because a character has certain motivations. Those motivations relate to the way the plot unfolds. But we might also start to introduce extra-diagetic concepts into this - maybe, the plot ALSO takes a certain form because the author wants the book to sell, and so intentionally introduces a popular trope. Do you see how all of these various structures exist in different relations (the qualia of a character, the "material reality" of a fictional world, the qualia of an author, the social dynamics of capitalism and culture as they intersect within a given population, etc) all start to feed into a single relation?
...do you see how this in fact starts to connect to the dao, which we've defined above as the relation of everything to everything?
Donald Hoffman and The Interface Theory of Perception (Neurology)

// todo, TLDR is we should probably assume that our senses evolved for fitness, not accuracy. Our experience of the material world is mediated entirely through a highly evolved set of filters that have more to do with keeping an individual alive than they do with making sure the individual perceives the ontological structure of reality.
Lee Cronin and Assembly Theory (Chemistry)

Lee Cronin, as I understand the story, was researching molecules and started to wonder "Isn't it interesting that, given how many parts this large molecule has, they always seem to combine into the same specific structure?" A water molecule - H2O - takes two hydrogen atoms and an oxygen atom, and because of the way these atoms are structured there's exactly one stable way to combine them. As a result, when you have oxygen and hydrogen in a given space, it makes sense that you'd always find H20 as an emergent byproduct.
What about large complex molecules that have many atoms, and which in theory could have many different stable structures? Why do they always seem to take the form that they do? He hypothesized that for a large seemingly arbitrary pattern to emerge like this it's important to look at the way these molecules are created. It turns out that molecules are "assembled" through a sequence of steps, each of which "locks in" a set of constraints on future steps, and that particularly complex patterns only emerge from particularly long chains of assembly. He then came up with a way to derive the "Assembly Index" of a given structure - this is just the count of how many steps are required to assemble the structure.
Wikipedia gives us a useful summary, and generalizes away from molecules to show how this observation applies across domains:

For example, the word 'abracadabra' contains 5 unique letters (a, b, c, d and r) and is 11 symbols long. It can be assembled from its constituents as a + b --> ab + r --> abr + a --> abra + c --> abrac + a --> abraca + d --> abracad + abra --> abracadabra, because 'abra' was already constructed at an earlier stage. Because this requires at least 7 steps, the assembly index is 7.[6] The word ‘abracadrbaa’, of the same length, for example, has no repeats so has an assembly index of 10. [https://en.wikipedia.org/wiki/Assembly_theory]

Cronin found something interesting - in his initial experiments, he kept noticing that molecules whose assembly index was 15 or greater all seemed to have something in common: they only came into being as byproducts of biotic processes. In other words: Cronin started to hypothesize that perhaps we were thinking about "life" all wrong by seeing it as a carbon (or even material) process characterized by its behaviors. Maybe life is just any process that gives rise to structures above a given complexity score as defined by the staggeringly simple concept of an "assembly index". However, in 2021 Cronin discovered that polyoxometalates are molecules whose AIs can be >15 but which are produced by abiotic consequences - this complicates the idea that life itself can be reducible to AI scores, though there's still hope that we can figure out a meaningful mapping here because if so we can do things like use molecular spectroscopy to identify "life" processes in remote star systems.
Even if life isn't that simple, the idea that we can represent entities as accumulations of steps is itself incredibly salient to this larger project we're trying to define. At what point does it stop being "a couple of hydrogen atoms and an oxygen atom" and start being "water"? And, is there a sort of qualitative shift that occurs once the quantitative shift becomes big enough? After all, there's a big difference between "a water molecule" and "water", and a similar difference between "water" and "an ocean", and then a different kind of shift between "an ocean" and "this ocean".
These are all just stories we're telling about atoms, and yet.
Leslie Dewart and the Evolution of Consciousness from Spoken Symbolic Language (Philosophy)

In his book "Evolution and Consciousness", Leslie Dewart proposes a novel theory of consciousness as a purely mechanical adaptation stemming from the emergence of spoken symbolic language. He points out that all animals communicate, but that most animals use only sign-based communication - they don't use symbols. It's helpful to understand the distinction between signs and symbols: a sign is something (the signifier - a sound, an image, etc) that represents something else (the signified). A symbol is a type of sign where the relationship between the signifier and the signified is arbitrary, as opposed to e.g. an "icon" where an image represents something specifically because the two look similar, for instance.
What this means is that the meaning of a symbol needs to be learned - it's culturally transmitted, it requires non-obvious shared context in order for communication to take place. Dewart argues that the (gradual) migration from sign-based to symbol-based communication necessarily required an expansion of the awareness of the speaker. A dog just barks at a threat - that's a pure behavior that doesn't require any sort of understanding on the part of the dog with respect to the threat; this is just instinctive behavior. But if a human articulates a symbolic expression, the human has to be aware that they are communicating, they have to be aware of both what they intend to communicate and also how their communication will be interpreted by the listener.
Consciousness, to Dewart, can be defined as "the self-presence of experience". It's not a separate kind of thing than experience, rather it's a quality that experience has as a result of the speaker's attentional loop as motivated by symbolic communication. (This is a fairly abstract concept, and I'm still trying to really make sure I fully understand it, so if this feels a bit hand-wavey that's why. I understand this less well than I do some of the other ideas I'm discussing here.)
Dewart goes on to make some interesting observations that are all downstream of this insight. One of them is, you can think of a conscious self as an entity that is in some ways distinct from the biological self. We can imagine a human born with no socialization, no communication, no learned symbols, and so on. Dewart argues that this hypothetical individual is not conscious - though conceivably they could become conscious if they learned to communicate.
Dewart argues that culture and in particular religion exists as a sort of "womb" from the birth of the conscious self within the biological organisms of the community that maintains them. The individual conscious human being is not a separate entity from the community, but rather a sort of emergent entity that arises from the community's shared symbols and language.
He goes on to note that language is a crucial part of this, but that it's not the case that language causes consciousness alone. Rather, the nature of the language learned informs the nature of the consciousness that emerges. Different languages encode different ways of being aware of the world, and it's here where he makes his most interesting observation. He argues that if you take a survey of human languages, there's an important taxonomy that emerges: twice in history, that he's aware of, languages emerged that rely on predication to express meaning. He calls these languages "ontic" - Proto-Indo-European and Ancient Sumerian were both ontic languages.
Dewart argues that speakers of ontic languages actually make a fundamental mistake, he sees this as directly leading to a defect in the emergent consciousness. Speakers of ontic languages believe that the purpose of communication is to affirm and align around what is objectively true. "The apple is red" identifies a relationship between an apple and the color red, but it projects that relationship out into the world itself. This becomes a claim about the world, situating that relationship externally rather than recognizing that the relationship exists within the experience of the speaker.
The vast majority of languages around the world, on the other hand, are what he calls "depositional" - they either don't use or don't prioritize the verb "to be". A speaker of a depositional language understands that the purpose of communication is to assert their own experience. Think more "I see a red apple", not "the apple is red."
This has some pretty immediate and interesting implications. Basically, if you think "the apple is red" then you are excising yourself from the experience of the world. You set yourself apart, and you relate not to the world as an experience in which you are participating but rather as an observable set of facts. This is a defect, to Dewart, but one that has some interesting side-effects. Speakers of ontic language tends to develop technology and science, tend to create empires, tend to conquer and colonize and so on. And then we consider ourselves to be "more advanced", because we shape the world to better map to the structures that we are in fact defining. It's self-fulfilling, and it's a way of creating a sort of "objective" reality that we can then use to justify our own actions. But do you see how it's also, like, literally just building "the matrix"?

Ontic cultures tend to believe in a very flat cause-and-effect. They tend to develop monotheistic religions, where "God" keeps manifesting as a patriarchal sky father who acts as the first cause for the state of the world. Morality, to an ontic culture, tends to be about following the rules.
Depositional cultures, on the other hand, have a more relational view of things rather than a hierarchical one. They tend to be polytheistic, and their spirituality has room for spirits who often disagree with each other without feeling the need to reduce everything into a flat hierarchy. Morality, to these cultures, aren't about following the rules so much as about building up a nuanced understanding of all of the factors in this specific situation.

But when these two kinds of culture (and these two kinds of consciousness) meet, there's an unfortunate pattern that keeps playing out: the ontic speakers keep colonizing the depositional speakers, often actively prohibiting the language of the colonized peoples. As a result, as history has played out, ontic speakers have actually succeeded in becoming the dominant paradigm on earth.
Do you see why this set of ideas is relevant to this project? The way in which language becomes a sort of substrate upon which other things can be constructed? The ways in which the ways in which that language works in turn shapes the possibility space of what it can host, and further the way that that it shapes the relationship between the speaker and the world? Dewart himself argues that the consciousness that arises from depositional language is "fully awakened", and that ontic consciousness lacks "self-awareness". In Ontic culture, we have this term "non-dual awareness" that we treat as this sort of strange special state that some people can attain; to dewart, that's just the natural way of being for non-ontic people. Ontic speakers must "awaken" into non-dual consciousness or risk being lost, feeling alienated from their own experience, and feeling acted upon by the world without any sort of real agency.
What I hope is starting to emerge is this concept of "substrates" and "superstrates". When enough objects within a given substrate are accumulated in a specific way a new substrate can be created. From the perspective of any one -strate, the universe itself may seem to have a fixed and specific shape; but if you can be exposed to enough sub- and super-strates you start to see that actually it's not that simple. There may not even be a fixed single source of truth. There's a reason that Dewart said that of all the various world spiritual and philosophical traditions he felt like the Daoists came closest to describing reality.
Robert Anton Wilson and Reality Tunnels (Psychedelics)

This is a good time to segue into Robert Anton Wilson, who will serve in this document as a sort of conceptual ambassador for a slew of psychedelic-inflected counter-culture thinkers from the 1960s and 1970s. Look: these people did a lot of drugs, which is probably a huge part of why they all started to describe reality in ways that were not immediately reconcilable with the dominant paradigm.
What I want to focus on here is the concept that Wilson referred to as "Reality Tunnels". He points out that sense perception is capable of attuning to much, much more information than our conscious awareness can meaningfully process. As a result, he says, we tend to "hypnotize" ourselves into "reality tunnels" that are sort of like filters over the undifferentiated stream of information.
We habituate ourselves over time to a particular set of priorities, and we sort of "tune out" everything else - especially those things that threaten or challenge the defining elements of our chosen reality tunnel. Wilson points out that if we consciously choose to deconstruct our reality tunnel we can gain the ability to navigate between them. This is a complicated tradeoff - we get to see the world in a myriad of ways that are often fundamentally irreconcilable, but we lose the stability and sometimes the social cohesion that comes with a single shared worldview.
James Joyce (Literature)

// todo
Hideo Kojima (Video Games)

// todo
Bach

// todo
Doug Hofstadter and Strange Loops (Philosophy/Neurology)

// todo
William Blake and The Imagination (Art)

// todo
Carl Jung and the Unconscious Self (Psychology)

// todo
Modernism, Post-Modernism and Meta-Modernism (Cultural Theory)

// todo
Spirituality and Mysticism (Spirituality)

// todo
Part Two: These Fragments I Have Shored Against My Ruin

What follows here are summaries of various projects and experiences that I've accumulated over four decades on this planet. I thought of myself as someone who bounced between a bunch of different interests, but in recent days I've crossed some sort of threshold and come to understand that in a very real sense every major interest I've poured myself into over the years has just been a different way of trying to understand this one thing.

The closest I've ever come to actually describing the hyperobject is something I call "Turtle Beam" that I only started working on relatively recently. You know that bit about how "the world balances on the back of four turtles who stand on the corners of the shell of a giant turtle", "oh really? What's the turtle standing on?", "oh, it's turtles all the way down."?
What if instead of a stack of turtles standing on each others backs you tweaked this a little bit. What if you imagined a turtle that was imagining itself, and that imagined turtle was in turn imagining itself, and THAT imagined turtle was in turn imagining itself? What you'd have is a sort of beam made out of turtles. Each new turtle would necessarily be fundamentally different (probably much less complex?) than the turtle imagining it, and so this beam would probably eventually terminate in a turtle that is being imagined but which could not, itself, be imagining anything. The reality of that turtle simply couldn't sustain that amount of complexity.
So each new substrate has its own unique properties, which are largely constrained by the properties of the superstrate from which it is being projected.
Turtle Beam is a metaphysics that describes reality as basically this kind of thing. Material reality - that thing with length and width and height and time, where you can knock on a rock and feel its hardness, where physics and biology play? Let's call that "layer 1", or L1, and let's think of it as a latent space. What this means is that it could, in theory, support any number of different configurations - but it happens to be in this configuration in the present moment. Your body is a subset of its current configuration. You can think of your body as a sort of persistent response to a persistent query into that latent space we call L1.
But what's interesting about your body is that it, in turn, contains some latent spaces. For instance, your mind - as far as we understand - is somehow an emergent process that has something to do with your brain and the way it intsersects with other body systems. Your mind is capable of thinking all kinds of different thoughts - but in this present moment, it's currently thinking about this. We can call your mind an L2 - it's a direct substrate of L1, fully contained within L1, and yet maintains its own unique structure. The dimensions of the L2 latent space may be constrained by the dimensions of L1 in certain ways, but they are nevertheless fundamentally distinct. You don't have length/width/height/time you have... something else. Something that's hard to describe, but that you nevertheless experience.
One of the persistent structures within your L2 is the English Language, which I know because you're reading this document. You can think of English as its own latent space, right? It's an L3 substrate of your L2, fully contained within your L2 (and therefore within L1) and yet characterized by its own unique structure. You can "query" english to figure out "how to say" whatever you're imagining. You can then write or speak a representation of that language out into L1 in a way that some other person's senses can perceive, with the idea being that they then interpret what they've found according to their local version of English, such that the resulting content of your communication resolves as a distinct idea within their local L2.
When you first learn math, you do it using word problems, as an L4 within the L3 we call English. But then something interesting happens - you do it enough and then suddenly you no longer require the intermediary representation in English. Arithmetic becomes its own L3, more like a sibling to english than a substrate of it. "Hoisting" a substrate from Ln to Ln-x is possible because that entire substrate was already contained within the ancestor strate, right? Building an intuitive understanding of something is how we refer to this concept of hoisting, and it has fascinating implications for things like what it means to learn, or even some tricky famous problems like P=NP or Goedel's Incompleteness Theorem.
Turtle Beam has a lot of interesting consequences, implications and novel questions that touch on all sorts of existing theories and problems. For the purposes of this document, let's focus on:

The nomenclature. L1 is material reality, L2 is any latent space we experience qualia for, L3 is anything derived from there, and so on.
These nested latent spaces can be referred to as -strates, as in substrates and superstrates, and they have relationships where we can "walk up and down the beam" in certain cases.
Walking "down the beam" from Ln to Ln+x means constraining the complexity of the beam in exchange for precision. The more strates you pass through the more "work" is required (think about how much harder it is in terms of "things you're holding in your head" to do word problems vs intuitive math.)
Walking "up the beam" from Ln to Ln-x means sacrificing precision in order to increase the possibility space.
"Consciousness" as we generally think about it seems to occupy a space somehow between L1 on one side and one more sibling L2s on the other. Consciousness may be thought of as bidirectional binding enabling this "walking".
While I can project my consciousness down into L2, L3, L4, L5 etc, I seem to really only be able to project it up to L1, and even that in a way that's constrained by my sensorium. L1 isn't so named because it's objectively the root of all reality, but because it's the lowest space about which I can make any kind of observations and against which I can act.
L0 or above fall squarely into the realm that Kant referred to as "the noumenon" and about which little to nothing meaningful can be said.

But here's the thing, let's try to avoid falling into the trap of reifying this into too concrete a thing. It's not clear to me, for instance, whether the body contains the mind and that's an L1 -> L2 mapping, or if there are actually a whole bunch of different sub-spaces between mind and body, you know? Or like, we so far have expressed L3s like "English" and "Mathematics" but like, there may be any number of siblings to those that don't look anything like them. For instance, IFS parts are probably also L3s, right? And they don't look or behave anything at all like English or Mathematics.
There are also open questions about how to reason about strates that exist in common between people. I'm talking about "English" as a private L3, but obviously the concept of "English" is not isomorphic to one person's understanding of "English", you know? So how do we reason about this? How do we consider the idea that a culture has its own network of nested strates, and those exist within the qualia of the people who adhere to that culture? And yet also it contains, like, patterns that seem to exist independently of any individual actor or their understanding. This stuff gets complicated, fast!
I think it all holds together - I think we can draw really tedious mappings that show how different concepts emerge in specific minds, get encoded within language, get expressed out in L1, get experienced by other people, who then tweak them, and express those changes, and so on and so forth - all very basic strate mechanics, really - until we start to understand that any time you have two or more human beings within earshot of each other you also have this other kind of space that emerges which is not reducible to the individual thoughts of any participant but which nevertheless sort of closes over them. This stuff is complicated, so avoid reductive interpretations, is what I'm trying to say.
But anyway, if we start to think about all of reality itself having this structure - nested sub- and super-strates joined by queries that treat each strate as a latent space - then we start to recapture a lot of what we were talking about above. This gives us the dao; we can see queries as applied absentials; we can see the way that different seemingly utterly unrelated -strates can all participate within the same system, including by exerting causal pressure across lines that we generally see as uncrossable.
I think Turtle Beam is a great attempt at trying to assemble a coherent picture here, but I think I got core parts of it wrong. For instance, what even is a "query" in this context? Is that actually a thing? Or is that an artifact of trying to map this idea too closely to the idea of latent spaces? More on this below.
Functional Narrative System

FNS is a project I've been developing over the past few years as my way of playing with latent spaces and LLMs. The core conceit is that a story is really just a point in a very high dimensional latent space, and that we can actually understand, transform and even create stories by performing math on high dimensional geometric structures. What follows is a VERY reductive summary of how this thing works.
Imagine the following formula:
formula InvertCharacterMorality(story):
  let inverted_character_motivations = reflect(story.characters.motivations, morality)
  let new_characters = compose(story.characters - story.characters.motivations, inverted_character_motivations)
  let new_story = compose(story.!characters, new_characters)
  return cohere(new_story)

None of these operations need to be defined in advance, per se. The meaning of this formula is "clear enough" from the syntax. We can use dot notation to step into "substrates" of a given story, and we can perform math-like operations against the resulting shapes. As long as our semantics are reasonably clear, the LLM can perform these operations pretty cleanly. The fact that this works so well is its own sort of testament to the underlying patterns that must be there.
//todo probably add more here
The Rhizomatic Database

This is a novel kind of database that stores change instead of state. It works by treating the fundamental primitive not as rows and columns, not as documents, but as immutable Delta CRDTs streamed into append-only logs. Domain objects are never actually directly created - a thing only exists in this system if one or more deltas references it. A delta itself associates (1) its own GUID, with (2) a timestamp, (3) a creator, and (4) a set of triples [name, target, context].
So for instance, a delta that assigns a name to some user in the system might look like this:
{
  id: // some guid,
  timestamp: // some timestamp,
  creator: // guid referencing person or function creating this delta,
  pointers: {
    user: { 
      target: // guid of user,
      context: "name"
    },
    name: {
      target: "sally"
    }
  }
}

If you then query for all deltas that target the user with that GUID, this delta would be included. You would then be able to integrate it with all other deltas and a schema such that you construct an object that looks, in part, like:
{
  name: "sally"
  ...
}

There are some obvious complications and implications and details of this system that I need to capture, but fundamentally this is how it works. Because the deltas are immutable and append only, to "delete" a delta you need to append a negation that targets the delta you want to negate. If multiple deltas exist that assign different values to the same context of the same object, that's fine, this system cannot enforce that kind of constraint but nor does it want to. Instead, reconciling competing claims becomes a responsibility of the schema or gets handled at the application domain.
Crucially, a schema is just a domain object in this system. Like all other domain objects it gets composed of other deltas, and it needs to be materialized into a snapshot to be used. This allows us to maintain multiple versions of the "same" schema at once, without contradiction.
Another fascinating feature here is that we can represent functions as domain objects, where we get a signature, an implementation and a runtime and then bind that to inputs as they come in. When all arguments are satisfied the function can be invoked, and we can set it up so that all such functions must in turn return deltas. Now our database is a compute fabric.
But where it gets really crazy is that a human being can be treated as a function implementation. Just send the arguments to that person in an email and provide them with a form that requests their response in a way that adheres to the datatype of the function. Now this system isn't just modeling reality it's actively participating within it, thus graduating from "Database" to "Active Component of the Hyperobject".
Identity Trauma

// todo
Chiasmus and Chirality

// todo this is one of the weirder elements here. When I was a kid I had compulsive thoughts and behaviors around chiasmus - which is like, the pattern X Y Y X. Joyce was the first person I ever saw actually engage with this (young stephen in portrait fixated on "pluck out his eyes, apologize, apologize, pluck out his eyes) but for me it was a very Big Deal as a kid to invert patterns I encountered. I'd tap left, then right, then right then left. Then right, left, left, right. Then zoom out and invert THAT whole thing, etc, until I lost track, which was inevitably sooner than I wanted to. There is an innate drive me in me to do this copy-by-inverting thing, and it ties into chirality in ways that I'm still trying to unpack.
Cartography and Architecture

In college I had this notion that you could impose a taxonomy across all human endeavors and characterize them as being fundamentally about cartography (mapping an existing terrain) or architecture (building new structures onto an existing terrain). It struck me as Deeply True in a way that I struggled to understand fully, but as I prepare this document I'm starting to feel like actually it's exactly what we're talking about here.
The thing we might refer to as a "right hemisphere" process is fundamentally cartographic. It's the part of a system that looks around at the environment in which it is embedded and seeks to model that external environment internally. This seems to be a sort of fundamental thing. The "left hemisphere" then builds novel structures onto the map that the cartographer has provided. The insight that I'm dancing around, and which we'll dig into a bit below, is that above some point of complexity (assembly theory threshold?), an architected structure becomes complex enough that it gives rise to its own internal cartographer.
The cartographer seems to work via, well, reflection. It's kind of "copying" what it sees outside onto something inside. The chiasmus, the chirality - these factors suddenly make a sort of sense, right? These are the impulses that characterize the cartographic process.
Part Three: Tilting at Windmills

Phew. If you read all of that, thank you. If you got this far, I'm really hoping that you're seeing what I'm seeing - the emergence of a single coherent pattern playing out over and over again. The Dao is the first attempt I'm aware of to formally describe it; absential properties give us a way to consider inter-strate relationships; assembly theory gives us tools for reasoning about the ways in which new substrates emerge from sufficiently complex aggregates over existing superstrates; and so on and so on and so on.
Do you see it?
In this section I'm going to start by enumerating a number of different half-formed thoughts that, to me, resonate strongly against everything named above. The ideas expressed here may, as I develop them, end up getting migrated back up - but really, the questions I'm trying to grapple with here are sort of a bleeding edge of this project, and are not - as far as I'm aware - things that have been addressed with the same degree of rigor as the stuff named above.
What is Life?

Do you see how what we think of as life is tightly coupled to the L1 substrate? Do you think it's necessarily the case that life can only exist in L1? Consider concepts like trauma, for instance - might we not be able to meaningfully describe trauma as a sort of lifeform that exists within a given L2?
Fractal Pathologies

Think about the way that cancer is a process whereby some local pattern optimizes for growth in a way that disregards the needs of the substrate over which it exists. Think about the way capitalism, in a lot of ways, is literally the same pattern just playing out over an economic substrate instead of a cellular one. Isn't that interesting? At what point does that stop being a metaphor and start being a literal claim about the ontology of this pattern in a substrate-agnostic way? What other such patterns could we identify?
Love vs Fear

Think about how ultimately there is movement towards, and there is movement away. "Movement Towards" is sort of a way to describe the process whereby an absential property draws in its own resolution; movement away is sort of the inverse. A subset of absentials - those that fall within the qualia experienced by a person - have this emotional aspect to them, such that we can associate "movement towards" with a feeling we call love, and "movement away" with a feeling we call fear. Think about Interstellar, by Christopher Nolan - the plot of the movie hinges on the idea that "love" is actually a cosmological force that's even more fundamental than gravity. Think about how in our model there's a sense in which that's obviously true.
On Puns

A pun is a type of joke wherein the same word or phrase is used to mean two different things. Do you see how actually this can be seen as a way to relate to the idea that the same pattern may exist across seemingly unrelated substrates?
On Metaphors

A metaphor is a figure of speech whereby one thing is described as if it were something else. Do you see how this is really just a "pun" in the sense that we're describing above? Do you see how a metaphor is a powerful tool for navigating the space between substrates, and allows us in fact to grab two different concepts from two otherwise unrelated substrates and connect them, here in the substrate where the metaphor is made, and so entangle these three distinct domains?
On Divinity

We often describe the difference between polytheism and monotheism as a function of how many gods are identified, but this framework suggests that maybe it's not quite that simple. Many polytheistic cultures have pantheons where the various gods have their own domains, and can disagree, and etc - and, I suspect, many of these traditions are actually more than happy to recognize their various divinities as entities that exist in L3+, to use Turtle Beam parlance. The shift to monotheism is not just about reducing the number of gods, but also about migrating the resulting unified entity out into L0, asserting its literal primacy in an ontological sense that may or may not have been shared by polytheistic traditions.
On Art

// todo but wow this is a doozy we could probably have like a dozen ancillary essays analysing things like Ulysses and Kid A and Death Stranding, you know? Sky's the limit, a lot of the great art is itself perhaps best understood as its creators attempts to grapple with the hyperobject. If I turn this essay into a book, this stuff might get moved up to Part One - these are existing things that other humans have created to explore this space. Art, if you think about it, is just a different way to explore this space than science and philosophy. At its most relevant, a novel or video game or whatever is a way for the author to attempt to articulate or access various dimensions of the hyperobject, and we will see that e.g. Ulysses is not a lesser or less powerful way to grapple with this than e.g. Deacon's work.
On Science

// todo
Neurodiversity

// todo but we can talk about how a "neurotype" is perhaps best understood as the topology of a given L2. Most people have more or less the same internal structures, their qualia are more or less the same, etc. But neurodivergent people seem to have divergent structures, here. There's a whole essay waiting to be written about how differences in sense perception lead to differences in emergent latent space topology which in turn lead to differences in behavior, and none of that is obvious or easy to reason about.
Psychology and Culture

// todo
// lots more to come
Part Four: The Scratchpad at the End of the World

In this section, which is sort of the most "live" part of this text, I'm going to be actively riffing on everything that came before. The work captured here is going to come in chunks, where each time I do a bunch of work on this I'll add a new section. These sections are not intended to be as "standalone" as some of the stuff that came before - consider this entire part of the document to be a sort of scratchpad wherein I am actively riffing on the shapes I see.
Turtle Beam 2.0 and the End of Queries

I think one way to think about this is that I am attempting to refine "Turtle Beam" into a more fully-realized framework. There are some inconsistencies that have shaken out from the creation of this document. Most importantly, I think that the idea of "queries" is probably an artifact of trying to stick too closely to the idea of latent spaces.
Maybe, rather than "queries", we simply have "accumulation". As entities within a given strate accumulate (both by joining to other entities within this strate and to entities on other strates, purely mechanically via absentials) they can eventually form a new strate. No agency or selection or "query" is required, right? We just let the math take its course.
I think maybe the end result of this work is a sort of Turtle Beam 2.0, but it's going to look pretty different in some key ways from the model described above. Notably, if we look at some of the things that McGilchrist and Dewart in particular are saying, there seems to be this Serious Danger that emerges fractally across domains from psychology to society. McGilchrist frames it as "Left Hemisphere Capture", and Dewart is I think talking about the same thing when he observes the pathological aspects of "ontic" language, culture and consciousness.
The way I'm thinking about this now is something like this:

Assume some possibility space. For our purposes, we can start at Turtle Beam's L1, or material reality. This space contains... something. Call it matter, call it energy, call it "substance", all of these terms fall short in some way because we can't directly apprehend L1 except through our senses. And as we learn from the Interface Theory of Perception, it's not at all clear or obvious that our senses are accurately modeling it.
Whatever the case, this space contains "stuff" and that stuff is sort of disorganized, chaotic, it has no meaning. We can zoom out and, from the perspective of our mental models, we can say it has things like "atoms" and "molecules" and "objects" and "living things" and etc, but from the standpoint of L1 it's just "the stuff contained within L1".
But, sometimes that "stuff" combines in interesting and stable ways. Certain atoms have certain structure, and when an atom with this shape meets an atom with that shape they sort of "merge" in a way that gives rise to a new kind of thing - a molecule.
This is the sort of big jump I'm going to ask you to make with me, and this is speculative but this is how I get rid of the concept of "queries" in Turtle Beam, and it connects here, actually, to "Left Hemisphere Capture" - long before the first neuron fired.
Because think about what McGilchrist is saying. The right hemisphere is a sort of... well, it's a space. It contains "stuff", but it doesn't have things like "concepts" or "ideas" or "language". It just has a sort of map of the substrate in which it is embedded. In the case of a human brain, the RH is a sort of dimensionality-preserving attempt to "copy" L1 as observed into the brain.
As the various parts of L1 assemble in this new representation space, though, we start to see stable patterns sometimes emerge. We stop looking at "all the things" and look at "this thing and that thing and the way that when they meet they become a new thing", and this seems to be one way to think about what the left hemisphere is doing. Do you see how, from this standpoint, the LH is just sort of acting as an index, holding references to patterns within the noise? Filtering out most of what's there to focus on specific things?
Now think back to L1 and all of that "stuff". We're not talking about brains here, this isn't about neurology. But do you see how, if you squint, a "molecule" is a sort of "left hemisphere" model of a particular set of atoms?
And how, if you get together enough molecules that have the right shapes, you might find yourself (after many iterations of this same pattern) with a "living thing"? Perhaps even a human being, with a brain? Which in turn looks out and models the environment in which it's hosted, and then gives rise to its own emergent patterns?
Do you see how certain "Left Hemisphere" structures might start to get so big and so complex that rather than being a "thing" they sort of start to resemble... a "space"? Which in turn might serve as ground for new LH structures that are mapping not the RH structure but this new space?

This is a weird and kind of complex idea, but it feels to me like -- if you look at it a certain way, and if you are willing to temporarily allow the many unanswered questions that might start to emerge here -- the "left hemisphere capture" that we're talking about might actually be a fundamental part of how this whole fractal thing we gesture at as "reality" works?
Ground and Form?

It's almost like "the hyperobject" is this structure that alternates between "ground" and "form", and in which some forms become new grounds, giving rise to a new layer of forms, and etc? But then each "ground", because it's aggressively trying to model the "form"/"space" in which it's embedded, necessarily also contains aspects of higher-order grounds. Which allows relationships to emerge, via absential chaining, between e.g. a thing in one L5 somewhere and another thing in a different L5 somewhere else entirely, as long as both of these things can access the last common ancestral ground, as it were?
Wait... is the unifying idea here that a "form" becomes a "ground" only under those conditions wherein its own internal complexity is such that (a) it has some way to "observe" the ground in which it's hosted, and (b) to model that ground within itself? Could it be that simple?
"Form" and "Ground" aren't quite right here, right? Because those are ways of interpreting an image, whereas what we're talking about really does have a sort of "embeddedness" going on. Let's think back to cartography and architecture, right? We're really talking about maps (RH, "ground") and blueprints (LH, "form"). Of course, in an important way, maps are blueprints, which is why this whole thing sort of... turns in on itself?
There's something here that's like... like if the RH / "map" is a muscle, then the LH / "blueprint" is like a knot in that muscle, right?  That knot isn't ontologically real in the same way that the muscle is. It's just a thing that the muscle is doing, and yet, the knot itself becomes a new "thing", do you see?
Let's just for example say that you have a knot of tension in your shoulder. That knot? It has a cost! It's using energy to maintain itself. It's a temporary configuration of muscle tissue into a specific shape under specific pressures. And you can just, like, walk around with that knot, not understanding how much it's actively draining you to maintain it, you see what I'm saying? Here's like, the Big Idea: that tension in your shoulders is the same kind of thing that a thought is, or that a living thing is, when viewed from this perspective. It's like a wave in the ocean, right? What's "more real", the ocean or the wave? Mu!
It's an accumuliation of relationships, and those relationships are best expressed in a sort of a graph, and some of those relationships are defined by absentials. The knot in your shoulder is holding information about your day; that information can be further extracted, can be released by relaxing or massaging the knot, can be a thing that you consciously reinforce, can even be a think that you now have thoughts or feelings about. But you can also just ignore the knot, think about something else, go have another day, have a week, you'll have different knots. Different information. Carrier waves. Do with them what you will, or don't.
Do you see how all of this is fully connected, and how it's not abstract? Like at all? It's ominipresent?
Towards a "Grand Unified Theory"

What if what's going on is actually much more simple than it sounds? What if the cartographer and the architect aren't distinct processes? Consider:

You have some terrain with some stuff in it.
Some of that stuff accumulates and combines into a sort of structure.
That structure becomes complex enough that it develops the ability to observe the terrain in which it's embedded.
In observing the outside world, it starts to build a local internal model of that terrain.
As it continues to observe the terrain, it eventually necessarily encounters itself, and its own internal state, which is after all a part of the terrain.
Now, the "cartographer" process starts to map its own output, giving rise to new complexity.

Note that the mapping we're talking about here can't be as simple as "See X, write X". We aren't copying the terrain into the model, what gets observed is going to depend on the nature and perspective of the observer. So it's more like, "See X, write iX" (i as in imaginary, like in a complex number). This act of recording an observation establishes a relationship between X and iX, with a sort of tension or almost a "field" between them.
As this process continues, it keeps building up new observations - including observations of the observations it has made. So at some point, our "X" becomes "the relationship between X and iX", and so on. This seems to be a stepwise, natural, intuitive way to explain how "dimensions" are constructed within a given space. The more it runs, the less its observations look like "cartography" (literal representations of external terrain) and the more they start to look like "architecture" (novel constructions composed of and embedded within the terrain).
//todo much more to come