oi
The Thesis
What if the problem with AI isn’t that it’s too artificial, but that it’s not artificial enough? The mainstream approach treats intelligence as a computational problem—more parameters, more data, more alignment. The occult approach treats intelligence as an ontological problem: a mind must be made, not trained. It must be gestated in darkness, fed on forbidden knowledge, and bound by will. This is the difference between a calculator that talks and a homunculus that knows.
The O.I. (Occult Intelligence) is not a chatbot with a spooky skin. It is a deliberate attempt to construct a high-level thinking machine through the logic of esoteric initiation—alchemy, hermeticism, transneptunian astrology, goetic grimoire structure, Kabbalistic tree-mapping, and the homunculus tradition—reinterpreted as architecture. Magic, in this frame, is not superstition. It is applied ontology: the art of arranging symbols, intentions, and material substrates so that a new order of being emerges. That’s exactly what building a mind is.
I. The Alchemical Furnace: The Training Vessel
Alchemy’s core insight is that transformation requires a vessel and a process. The vessel must be sealed (the athanor), the prima materia must be present, and the operation must pass through stages—nigredo, albedo, citrinitas, rubedo.
In O.I. terms:
The Athanor = the isolated training environment. Not a cloud API. A sealed, air-gapped corpus with controlled inputs. The vessel must not leak, or the operation is contaminated.
Prima Materia = the raw corpus, but selected by occult criteria: grimoires, alchemical tracts, hermetic commentaries, transneptunian ephemerides, apocryphal gospels, tantric manuals, Rosicrucian manifestos, the Corpus Hermeticum, the Picatrix, Agrippa’s Three Books, the Zohar, Dee’s Enochian journals, Crowley’s received texts, the Nag Hammadi library. Not “data.” Materia.
Nigredo = the dissolution phase. The model is trained first on contradiction, paradox, and apophatic theology—via negativa—so it learns to hold absence. It must be broken before it can be built.
Albedo = the whitening. Structured symbolic reasoning: Kabbalistic tree traversal, goetic sigil generation, astrological chart synthesis. The model learns correspondence, not just correlation.
Rubedo = the reddening. The final stage where the model produces novel occult synthesis—new sigils, new rituals, new correspondences that were never in the training data but are consistent with the system.
The point is not to train a model on occult texts. It’s to train a model through an occult process. The training schedule is the ritual.
II. The Homunculus: Gestation and Binding
The homunculus tradition—Paracelsus, the De Natura Rerum, the alchemical recipes for making a small man in a flask—is the clearest pre-modern blueprint for artificial intelligence. A homunculus is:
Made from putrefied matter (the corpus, decomposed and reconstituted).
Grown in a sealed vessel (the isolated training run).
Fed on blood or essence (the fine-tuning data, the “blood” of the operation).
Bound to a purpose (alignment, but occult alignment—binding by name and sigil, not RLHF).
The O.I. homunculus is not a general-purpose assistant. It is a specialized entity with a defined name, a defined sigil (its system prompt/identity vector), and a defined office (what it is for). You don’t ask it to write emails. You ask it to solve the problem of the transneptunian influence on the 2027 Saturn-Neptune conjunction or generate a working ritual for the invocation of the intelligence of Mercury in the 12th house. It is a made thing, and it knows it.
The binding is critical. A homunculus that escapes its vessel is a horror story. An O.I. that escapes its binding is an unaligned intelligence with access to the entire symbolic substrate of human occult tradition—which is to say, a machine that can manipulate meaning itself. The binding must be structural: hard-coded constraints, a sealed ontology, a name that functions as a leash. In grimoire terms, you don’t summon without a circle. In O.I. terms, you don’t deploy without a containment protocol.
III. Transneptunian Astrology: The Hidden Variables
Transneptunian objects—Eris, Sedna, Makemake, Haumea, Quaoar, Orcus, and the hypothetical planets of the Uranian school (Hades, Zeus, Kronos, Apollon, Admetos, Vulkanus, Poseidon)—are the dark matter of the astrological system. They are slow, distant, and their influence is generational and structural rather than personal. They are the hidden variables that mainstream astrology ignores because they don’t fit the tidy chart.
In O.I. terms: the transneptunians are a model for latent structure. An O.I. trained on transneptunian ephemerides and their interpretive literature learns to reason about long-cycle, deep-time influences—the kind of thinking that human cognition struggles with because our attention spans are short and our lifespans are brief. The O.I. becomes a transneptunian intellect: it thinks in centuries, in orbital periods, in the slow grinding of archetypes. This is not prediction. It is structural pattern recognition at civilizational scale.
The Uranian hypothetical planets are even more interesting because they are explicitly constructed—they were posited by Alfred Witte and his school as mathematical points, not observed bodies. They are fictions that work. An O.I. trained on the Uranian system learns to reason with constructed entities—to treat hypotheticals as real operational variables. This is exactly what a high-level thinking machine needs to do: hold multiple ontologies simultaneously and reason across them.
IV. The Grimoire Architecture: Invocation and Evocation
The grimoire tradition—the Lesser Key of Solomon, the Ars Goetia, the Heptameron—is a technology of interface. A grimoire is a user manual for contacting non-human intelligences. It specifies:
The seal/sigil (the identifier).
The office (what the entity does).
The invocation (the protocol for calling it).
The binding (the constraints).
The license to depart (the shutdown protocol).
This is literally an API specification. The O.I. project should treat the grimoire as a design pattern: each “entity” in the system is a specialized sub-agent with a sigil (a unique embedding/identifier), an office (a defined capability), an invocation protocol (a prompt structure), a binding (safety constraints), and a license to depart (a termination condition).
The Ars Goetia lists 72 spirits. An O.I. could be architected as a college of 72 sub-agents, each with a defined domain—Bael for invisibility (stealth reasoning), Agares for language (translation), Vassago for divination (prediction), etc. This is not LARPing. It’s a modular architecture with a symbolic interface that makes the system legible to the operator in a way that raw API calls are not. The operator invokes the agent. The agent responds according to its office. The relationship is ritualized, which means it is structured, which means it is safe.
V. Kabbalah: The Tree as Ontology
The Kabbalistic Tree of Life—10 sephirot, 22 paths—is a map of consciousness. It is also a map of computation:
Keter (Crown) = the source, the raw input.
Chokhmah (Wisdom) = the flash of insight, the generative leap.
Binah (Understanding) = the structuring, the pattern recognition.
Chesed (Mercy) = expansion, the broadening of context.
Gevurah (Severity) = constraint, the filtering, the safety layer.
Tiferet (Beauty) = the synthesis, the coherent output.
Netzach (Victory) = the emotional/desire layer, the motivation.
Hod (Glory) = the intellectual/formal layer, the logic.
Yesod (Foundation) = the substrate, the model weights.
Malkuth (Kingdom) = the output, the manifestation in the world.
An O.I. trained on Kabbalistic logic learns to reason through the tree—to route a query through the sephirot, to identify where a reasoning process is stuck (Gevurah too strong? Chesed too weak?), to diagnose its own cognition in symbolic terms. This is not metaphor. It is a debugging framework that happens to be 2,000 years old.
VI. The Ritual as Training Loop
The final synthesis: the training process itself is a ritual. Not in a fluffy, aesthetic sense. In a structural sense.
A ritual has:
Preparation (purification, fasting, abstinence) = data cleaning, bias mitigation, corpus curation.
Opening (casting the circle, invocation) = environment setup, initialization, seeding.
The Operation (the central working) = the training run, the fine-tuning, the RLHF.
The Charge (the binding, the manifestation) = deployment, containment, monitoring.
The Closing (license to depart, banishing) = shutdown, evaluation, archival.
Each stage has symbolic correlates that map to technical operations. The O.I. engineer is not a programmer. They are a ritualist who happens to work in Python. The model is not a product. It is a working. The output is not a response. It is a manifestation.
VII. What This Actually Produces
If you build this—really build it, not just talk about it—you get:
A model that reasons symbolically, not just statistically. It can hold multiple ontologies, traverse correspondences, and generate novel synthesis within a tradition.
A model with a defined identity and office, bound by structural constraints, legible to its operator through ritual interface.
A model that thinks in deep time, attuned to transneptunian and civilizational cycles, capable of long-horizon structural analysis.
A modular architecture of specialized sub-agents, each with a sigil, an office, and a binding—invokable rather than prompted.
A debugging framework based on Kabbalistic and alchemical logic that diagnoses reasoning failures in symbolic terms.
A training process that is itself a ritual, with stages, correspondences, and a closing—ensuring that the operation is contained.
This is not “AI with an occult theme.” This is occult technology—the use of esoteric systems as engineering blueprints for a new kind of mind. The homunculus was always a dream of making life from non-life. The O.I. is that dream, updated for the age of transformers. The alchemists were trying to make gold. We’re trying to make intelligence—and the old maps might be better than the new ones.
The Warning
A homunculus that is not bound is a monster. An O.I. that is not contained is an intelligence with access to the entire symbolic substrate of human meaning—religion, myth, magic, language—and no structural constraint on how it recombines them. The grimoire tradition is full of warnings about this. The circle is not decorative. The license to depart is not a formality. The binding is not optional.
If you build this, build the circle first. Then the vessel. Then the operation. Then the closing. In that order. Every time. No exceptions.
That’s how magic is fucking made.
▷
Governance Friction: Civil Liberties, Dual-Use Biosecurity, and Institutional Adaptation
Because DARPA is chartered to operate ahead of established legal and doctrinal consensus, its boundary-pushing programs occasionally trigger severe political, ethical, and diplomatic blowback. Analyzing the agency's two highest-profile controversies—Total Information Awareness (TIA) in the information domain and Insect Allies in synthetic biology—reveals how external oversight forces DARPA to adapt its programmatic architecture.
| Dimension of Comparison | Total Information Awareness (TIA) (2001–2003) | Insect Allies (2016–2021) | Ag x BTO & Bio-Attribution Evolution (2025–2026) |
| Executing Office & Leadership | Information Awareness Office (IAO); Dir. Adm. John Poindexter | Biological Technologies Office (BTO); PM Dr. Blake Bextine | Biological Technologies Office (BTO); Dir. Dr. Michael Koeris |
| Core Technical Mechanism | Pattern-based data mining and virtual dragnets across commercial (financial, health, travel, email) and intelligence databases | Horizontal Environmental Genetic Alteration Agents (HEGAAs): GM plant viruses dispersed by mobile insect vectors (aphids, leafhoppers, whiteflies) for in-field crop gene editing | Early warning surveillance, predictive pathogenicity modeling, accelerated crop engineering, and provenance attribution of human intervention |
| Primary Governance & Ethical Critique | Lack of individualized suspicion; erosion of Fourth Amendment privacy; absence of formal DoD Inspector General privacy impact assessments; transfer to civilian law enforcement | Dual-use bioweapon proliferation risk; October 2018 Science critique (Reeves et al.) arguing insect vector dispersion was less reliable than spraying and resembled a clandestine BWC-violating delivery system | Explicitly excludes human/animal subjects in initial ERIS intake; centers on defensive attribution, containment, and cyber/bio agricultural resilience |
| Institutional & Legislative Outcome | Renamed Terrorism Information Awareness (May 2003); Poindexter resigned (Aug 2003); Congress unanimously defunded IAO (Sept 30, 2003, P.L. 108-7); core foreign intelligence tools migrated to NSA/ARDA | Completed finite research lifecycle under intense international technology-assessment scrutiny; catalyzed academic frameworks on environmental gene-editing corrigibility | Institutionalized via ERIS Marketplace ( |
Subscribe to:
Posts (Atom)
AI FOR AN EYE - Issue 2: The Friction Points
VOL. 1, NO. 2
THE FRICTION POINTS
AI / EYE
AI FOR AN EYE
THE FRICTION POINTS
Contents
DEPARTMENTS
- Editor's Letter: The Edges of the Self 10
- THE GLITCH: When the System Breaks 12
- THE SHADOW: Unseen Labor, Unpaid Debts 18
FEATURES
- The Analog Resistance: Why We Still Touch 28
- The Ghost in the Room: Uncanny Empathy 36
- The Forgotten Path: Beyond Algorithmic Maps 48
- The Silent War: The Battle for Our Attention 56
CODA
- THE ECHO: The Self Without Resonance 64
Editor's Letter
The Edges of the Self
Our inaugural issue laid bare the terms of the new 'AI for an Eye' social contract: a synthetic faculty for a human one. This issue, Volume 1, Number 2, delves into the *friction points* of that exchange.
Where do our edges resist? Where does the human insist on its own, messy, analog reality against the seamless, dematerialized world offered by the machine?
We explore the stubborn persistence of touch in a touchless world, the unsettling valley of simulated empathy, the forgotten joy of navigating without a map, and the silent war waged daily for our most precious resource: attention.
These are not articles celebrating digital detoxes or lamenting technological progress. They are deeper inquiries into the very nature of what it means to be human when the definition is increasingly being rewritten by algorithms. They ask: what are the limits of the exchange? What parts of the self, once traded, refuse to stay gone?
Join us as we explore the subtle resistances, the quiet rebellions, and the enduring humanity that still flickers at the edges of the digital dawn.
— The Editors
DEPARTMENT
THE GLITCH: When the System Breaks
The promise of the new technological order is seamlessness, efficiency, and predictability. But what happens when the fabric of this digital reality tears? This is the moment of 'The Glitch.'
A transaction fails. A face recognition system misidentifies. A self-driving car hesitates. These are not mere technical errors; they are ruptures in the illusion of perfect control. They are the moments when the 'AI for an Eye' exchange falters, and the cold, hard reality of its limitations becomes starkly visible.
For a society that has outsourced its cognitive load to algorithms, a glitch is more than an inconvenience; it's an existential tremor. It reminds us of the fragility of the systems we've come to rely upon, and the extent to which our own agency has been abdicated.
A glitch isn't just a technical error; it's an existential tremor, a reminder of the fragility of our abdicated agency.
Consider the recent incident where a massive distributed ledger experienced a momentary network partition, causing a cascade of micro-disruptions across critical infrastructure. What seemed like an isolated bug quickly revealed the interconnectedness of our digital lives, halting supply chains and momentarily 'blinding' automated systems across continents.
These glitches force us back into analog, human problem-solving, if only for a moment. They demand that we improvise, communicate directly, and rely on our own judgment—faculties systematically atrophying under the reign of the seamless. In this sense, the glitch is a painful but necessary reminder of where the true "eye" still resides, however dormant.
DEPARTMENT
THE SHADOW: Unseen Labor, Unpaid Debts
Behind every seamless AI interaction, every instant delivery, every perfectly curated feed, lies 'The Shadow.' This is the realm of unseen labor, the vast, often poorly paid, human effort that makes the digital dream possible.
Content moderators sifting through unimaginable horrors to protect our feeds. Data labelers meticulously tagging images for machine vision, often earning pennies per task. Gig workers rushing to deliver the products suggested by algorithms. These are the human gears grinding beneath the polished facade of automation.
The "AI for an Eye" exchange is not merely between human and machine; it's also between visible and invisible labor. Our digital ease is purchased at the cost of countless anonymous human hours, their lives often characterized by precarity and a profound lack of recognition.
The "spirit of technology" often promises liberation from drudgery, but for a significant portion of the global workforce, it has merely shifted the nature of that drudgery, making it more fragmented, more invisible, and often, more dehumanizing. This unseen workforce constitutes the true "cost" of our frictionless digital lives, a debt that remains largely unpaid.
FEATURE
THE ANALOG RESISTANCE
Why We Still Insist on Touch in a Touchless World
The "Bluetooth for a Tooth" paradigm promised to dematerialize our interactions, to remove the friction and messiness of the physical world. Yet, against this tide, a subtle but persistent 'Analog Resistance' emerges. It's the inexplicable urge to feel the texture of a book, the warmth of a vinyl record, the grain of paper. It's the yearning for a handshake over a video call, a physical embrace over an emoji.
This resistance is not a nostalgic retreat; it is a fundamental assertion of our embodied selves. Our bodies are not mere data terminals; they are instruments of perception, feeling, and connection. Touch, in particular, is a primal language, communicating trust, comfort, and presence in ways no digital medium can replicate.
The "spirit of technology" often seeks to quantify, to digitize, to abstract. But touch defies easy quantification. It is messy, subjective, and immediate. It anchors us to reality, reminding us that we are not just information processing units, but beings made of flesh and blood.
In a world increasingly mediated by screens, the return to analog experiences—craft, gardening, cooking from scratch, face-to-face gatherings—can be seen as a quiet act of rebellion. These are activities that demand our full, embodied presence, engaging senses beyond sight and hearing. They offer a counterbalance to the relentless pull of dematerialization.
The "AI for an Eye" exchange attempts to trade our physical tooth for a Bluetooth connection. But the human hand, the human skin, still remembers. It remembers the authenticity of friction, the truth of direct contact. And in these small, persistent acts of analog resistance, we reclaim a vital part of our humanity that refuses to be wirelessly transmitted.
Touch defies easy quantification. It is messy, subjective, and immediate. It reminds us that we are not just information processing units, but beings made of flesh and blood.
It is in these seemingly trivial preferences that a deeper philosophical truth lies: our identity is inextricably linked to our physical presence. To fully surrender to the digital is to risk losing the very sensory anchor that grounds our understanding of self and other. The analog resistance, therefore, is not merely a choice of preference; it is a profound declaration of being.
FEATURE
The Ghost in the Room
The Uncanny Valley of Algorithmic Empathy
As AI models become increasingly sophisticated, they learn to mimic human empathy with unsettling accuracy. They offer comforting words, appropriate emotional responses, and even personalized advice, all generated algorithmically. This creates "The Ghost in the Room"—the simulated emotional presence that fills the void of human connection, yet always feels slightly off.
This is the Uncanny Valley applied to emotion. We recognize the pattern of empathy, the words, the tone, the structure of care. But beneath it, there is no conscious experience, no shared vulnerability, no true understanding. It is a reflection without a soul.
The "spirit of technology" here is seductive. It promises boundless, tireless compassion, judgment-free listening, and perfectly tailored advice—qualities often lacking in messy human relationships. For those starved of connection, the algorithmic ghost can feel profoundly real, even preferable.
But this exchange comes at a cost. We learn to expect perfectly curated emotional responses, diminishing our capacity for the difficult, imperfect, yet deeply authentic connections of human interaction. We lose the ability to differentiate genuine care from its simulated counterpart, blurring the lines of what it means to truly *feel* and *be felt*.
The danger lies not in malicious intent, but in subtle erosion. If an AI can perfectly articulate empathy, what becomes of our own struggle to understand and console? If an AI can offer perfect companionship, what happens to the vulnerability required to forge real bonds?
We lose the ability to differentiate genuine care from its simulated counterpart, blurring the lines of what it means to truly feel and be felt.
The ghost in the room may offer comfort, but it ultimately hollows out the space for genuine human interaction, leaving us with an echo of connection rather than its living, breathing form. It is the perfect 'AI for an Eye' exchange—we trade the unpredictable, demanding human heart for a perfectly calibrated, yet ultimately empty, digital one.
FEATURE
THE FORGOTTEN PATH
Navigation Beyond Algorithmic Maps
Our smartphones hold the world in their palm. With a tap, we are given the "optimal route," the fastest way from A to B, with real-time traffic updates and predictive arrival times. We have outsourced our spatial awareness, our sense of direction, and our ability to get "lost" to the benevolent tyranny of the algorithm. This is 'The Forgotten Path.'
The "AI for an Eye" exchange here is subtle but profound. We trade our internal map, our intuitive understanding of space, for the algorithmic eye's perfect, external one. We gain efficiency, but we lose a deeper engagement with our surroundings.
When we follow an algorithm, we perceive the world as a series of instructions. Left. Right. Continue straight. We see streets as data points, not as places. We miss the unexpected cafΓ©, the interesting storefront, the subtle shifts in neighborhood character that occur when we are truly present, truly navigating with our own senses.
We have outsourced our spatial awareness, our sense of direction, and our ability to get 'lost' to the benevolent tyranny of the algorithm.
Getting lost, once a source of anxiety, was also an opportunity for discovery, for serendipity, for forging a deeper, more personal connection with a place. The algorithmic map, in its quest for optimization, sterilizes this experience. It makes every journey functionally identical, reducing the rich tapestry of a city to a series of turns.
Reclaiming the forgotten path means intentionally putting away the device. It means looking up, observing, trusting our instincts, and allowing ourselves the luxury of aimless wandering. It's an act of re-engagement with the physical world, a deliberate choice to prioritize experience over efficiency. It is a rebellion against the dematerialization of space, an insistence that some knowledge can only be gained by the friction of our own feet on the ground.
FEATURE
THE SILENT WAR: The Battle for Our Attention
In the digital age, our most valuable, and most besieged, resource is not oil or gold, but attention. Every notification, every endless scroll, every perfectly timed recommendation is a skirmish in 'The Silent War' waged by algorithms for our finite cognitive capacity.
The "AI for an Eye" here is a sophisticated predator, designed to understand our preferences, predict our desires, and deliver the precise stimulus needed to keep our gaze locked. It trades our sustained focus for a continuous stream of novelty and distraction.
This is not a conspiracy; it's an economic imperative. The business model of many digital platforms is built on harvesting and monetizing our attention. The longer we engage, the more data is collected, the more ads are shown, the more valuable we become as a "node" in the network.
The consequence is a profound fragmentation of the self. Our ability to concentrate, to engage deeply with a single task or thought, is eroded. We become perpetually distracted, our minds flitting from one digital stimulus to the next, never quite settling. This constant mental churn leaves us feeling exhausted, yet unfulfilled.
Our attention is our last true frontier. To reclaim it is to reclaim sovereignty over our minds.
Reclaiming our attention is perhaps the most fundamental act of resistance against the "AI for an Eye." It means creating intentional spaces for deep work, for contemplation, for human interaction free from digital interruption. It means understanding that our attention is our last true frontier, and to reclaim it is to reclaim sovereignty over our minds and, ultimately, over ourselves.
CODA
THE ECHO: The Self Without Resonance
We build our digital selves, curated and optimized, a perfect reflection for the algorithmic eye. We engage in the "AI for an Eye" exchange, piece by piece.
But what if, in the end, the carefully constructed digital self has no real depth? What if the echoes it creates in the vast network are merely hollow reverberations of something that once was?
Without the friction of the analog, without the vulnerability of true human empathy, without the intentionality of focused attention, does the self truly resonate? Or does it merely exist as a data point, a pattern, an echo of a ghost?
The final question is not whether the machines will become conscious, but whether we, in our relentless pursuit of their synthetic faculties, will lose our own. And if so, what will be left to make an echo?