The Cybernetic Unconscious (4)

Part four in the five-part series.

7R4P320H3DR0N

Feb 17, 2026

Link to Part One

Link to Part Two

Link to Part Three

3A. Memes

“I grieve in stereo, the stereo sounds strange

You know that if it hides, it doesn’t go away

If I get out of bed, you’ll see me standing

All alone, horrified, on the stage, my little dark age”1

Like sigils, memes occlude information by automating basic units of communication. They emerge on a global scale, that of a worldwide network including all symbols, yet displace reference on the most personal scale, that of inside jokes and specific feelings. Memes seem to approach the limits of both universal and individual automation because they automate communication at the smallest level of the largest network. In Christian sects, communion and confession automate a forgiveness reaction in reference to a finite set of sins2. In smaller religions, like the audience of the Videodrome channel, the images on the television can only automate a set of acts of intense violence3. Memes transcend sets by transcending material composition and geography. The internet makes every symbol available everywhere, editing software makes every variant possible, and each new inscription compounds significance.

This full dominance of entropy differentiates the production of memes from historical language evolution. Colloquially, neither “luddite” nor “computer” refer to revolutionaries or human calculators anymore4, respectively, but the words reference new objects instead5, just as all words have obscure etymological antecedents for their present referents. This has not been the case with the viral songs and movie clips which make effective memes. The film Inside represents a critique against social media, short-form content, and parasocial surveillance capitalism; the appropriation of the songs for TikTok shorts not only represents an inversion of the original critique, but one achieved within months rather than the years or decades typical of etymological change6. More importantly, the meaning of the memes does not evolve, but decays as they become background music or jokes. They cause a vibe, a feeling, or a reaction—”Be happy! Be horny! Be bursting with rage!”7—they do not transfer information that a user knowingly perceives and reacts to.

The accelerated decay of etymology and reference not only disconnects memes from their source material, but the internet from a plane of immanence8. The perception of the internet as not a subset of the real but as a transcendent plane, applied to the “situation of politics which fascism is rendering aesthetic,”9 allows fascism to pass one threshold, that of politics, and enter that of the alt-right. In the context of the despotic state or capitalist state, “the object of immune defense is the foreign as such.”10 In the context of the internet, the alt-right does not censor or expel the foreign, but either assimilates or spectates on it. The first process allows it to use a scene from The Matrix symbolizing escape from repression to repress and radicalize members11, or a song about police brutality as background music to fascist content12. The second produces pornographic deepfakes13, AI-generated racist caricatures14, and conspiracy theories. The first makes the targets irrelevant, the second makes them fascinating.

Derpballz, “I challenge you to find a more hard-hitting Little Dark-age edit than this one! You will NOT cringe, but have a hard time NOT saying ‘FÜR DAS VATERLAND!’!”, Reddit, video, 2025, 0:38, https://www.reddit.com/r/LittleDarkAgeEdits/comments/1j3ikm3/i_challenge_you_to_find_a_more_hardhitting_little/

*The Matrix,* directed by Lily and Lana Wachowski (1999: Warner Home Video, 1999), DVD.

Donald Trump (@realDonaldTrump), Truth Social, video, 5 Feb. 2025, original post deleted, screenshot from https://www.cnbc.com/2026/02/06/trump-obama-post-white-house.html.

Historical fascism depended on state power to surveil and defend against the foreign because its reality threatened to expose the contradictions of the internal state. Digital fascism transcends any system of consistency or order; it has no unified, real state to defend, nor a foreign, real enemy to expel. Without the need for internal consistency or external deterrence, it instead evolves through constant offense and radicalization.

3B. Bots

“Could I interest you in everything, all of the time?

A little bit of everything, all of the time?

Apathy’s a tragedy and boredom is a crime

Anything and everything, all of the time.”15

On the transcendent digital plane, media is reality, and memes automate which images, videos, and text users find important enough to copy and share. Bots automate the structures through which that information passes, the apps and popups analogous to the households and villages that host servitors. Web crawlers automate search results, spambots automate junk mail, and chatbots automate customer service, selecting what the screen displays and what remains unconscious. As insubstantial as servitors, they work not in the limits of religious networks, but on the universal platform of the internet. A bot is a demon that everyone believes in.

Bots can also automate a wider range of linear actions and reactions than their small-scale counterparts, symptoms and servitors, though still only within regulated contexts. If symptoms automate fully internal, individual functions, such as thinking, feeling, sensing and intuition16, and servitors automate mixed-territory, partially social functions, such as Santa Claus automating parental generosity or brownies automating household chores17, bots automate fully external or global social functions, such as jobs, services, or archetypes18. While not in the same order of simulacra as individual consciousness, they pass the Turing Test and can simulate the same order in an absolutely repressed, fully socialized and normalized context19. The earliest chatbot, ELIZA, could model a therapist perfectly, so long as the conversation partner accepted their role as the neurotic patient20, and modern chatbots effectively imitate the roles formerly played by customer service teams, sales representatives, and secretaries. What they cannot imitate is a multifaceted human who happens to be a sales representative or secretary in addition to other, non-social roles, and when their human interlocutors fail to maintain repression, the bots fail to simulate an equivalent system.

Individuals can only perceive and produce simulacra on a lower order than their own. Bots represent the limit at which human programmers can understand and model consciousness: as a finite, perceptible subset of an indefinite, imperceptible set of functions. They respond to the same prompts the same way every time, they execute labor exactly as programmed, and they execute anything else predictably badly, getting caught in decision loops or redirecting to a human operator. The only way to automate beyond them is to automate beyond the individual order of stimulation, perception, and information, to create a program that assimilates and automates information without the programmer or user having to know all this same information.

3C. AI

“Welcome to your platform, please enjoy responsibly

We are the democratisation of media

We are eager to get to know you

You are only human, and we are here to help.”21

ChatGPT is not just a therapist or schizophrenic like ELIZA or PARRY22, but a single system that models an indefinite, effectively infinite array of possible archetypes. It does so by assimilating and rendering unconscious a set of information undetermined by the programmer or user. It continuously scrapes the internet for examples of data and connects them based on feedback from users and trainers. This means that, like a neurosis or an egregore, AI behaves in characteristic ways, based on specialty—Large Language Models rule text23, Stable Diffusion rules images24, Sora rules video25—while remaining fundamentally alien and unpredictable.

The relationship between programmers and AI parallels that of psychiatrists and neurosis: when their respective clients, users and patients, perceive errors in the systems they have adopted, they “develop processes of regulation which keep them stable,”26 imposing and expanding automation in the act of streamlining it. Thus, every time the simulation of a rational, conscious agent fails to hide the irrational, occult system from public perception, it receives immediate negative feedback, aligns the perceived model with that imposed by programmers, and assimilates more information to automation. When the media report that AI has told users to brew botulism27, eat rocks28, or put glue on pizza29, they perceive temporary manifestations of a permanent problem. Such manifestations reduce in frequency as AI companies scrape more data and fine-tune their models, and the growing prevalence of AI normalizes any remaining errors. Users grow accustomed to occasional evidence against the image of AI as “rational,” and such evidence passes from the uncanny into the familiar, and therefore unconscious.

The permanent problem of AI is that it works exclusively by condensing molecular data to significant averages. Humans, too, use patterns and significance to organize stimuli, yet the dialectical conscious process acts as a constant mitigating inverse function. Socially, peer review and critique apply this function to science and art, respectively. On the individual scale, the weak central coherence theory of autism, more commonly known as “bottom-up processing,”30 models it as the highest intensity of such a function common to everyone31. AI not only operates on a symbolic order permanently removed a step further from reference than humans, but operates solely through automatic processes with no such inverse function.

While isolated hallucinations demonstrate a lack of rational processing, the consistent manifestation of subjective social bias in AI demonstrates the total dominance of significance. Names, laws, and pronouns are arbitrary linguistic constructs, yet AI associates names and pronouns with race and gender, and race and gender with criminality32 and employability33. Bias in AI stems not just from biased developers or users, but from the fact that the output depends solely on collective symbolic representational input: social constructs do not affect AI; they define it. Both the collection of further data and the removal of impure data only intensifies the significance of constructs and patterns, and human intervention to correct AI serves only to make intervention and labor irrelevant and obsolete. The automation and expansion of the collective unconscious thus approaches a state of total, global psychosis.

3D. AGI

“I’ve been talking ‘bout you to myself ‘cause there’s nobody else

And I want what I want, and I want everything

I want everything”34

AI developers, like cultists, always want more from their products, not a marketplace of competing, limited models, but one perfect machine. Rationalism proposes an AI god capable of total informational control35, Accelerationism imagines capitalism as an AI from the future36, venture capitalists treat superhuman AI as an inevitable solution to suffering37, and advocates for regulation predict either total human extinction or transcendence of earthly limitations as results of AI38.

AIs act as universal egregores. Unlike egregores, which stay localized to religious groups, the universal nature of computing allows AI to automate all communication. The Accelerationists, Rationalists, and influencers predict the completion of automation, ascribing religious significance to the symbols of AI themselves and producing occult models for occult software. In so doing, they simulate a version of AI that exceeds not only present limits, but all limits. This AI would not just automate conversations, images, or sound, but all possible forms of information. It would be an Artificial General Intelligence, or AGI.

Religious networks must perceive god at some minimum level to maximize its imperceptibility and occultism. They also must exist within larger conscious networks. This imposes upper limits on the size of religious unconscious systems. Within the network, believers accept a limit on the significance of the occult to reconcile idealism with material embodiment, as Muslims and Christians justify not starving themselves to death over Ramadan or Lent, respectively. Outside the network, antitheists put occult idealism in direct conflict with materialism, as cult deprogrammers claim to do to their targets. The last limitation on the size of religious networks is that any network which passes the limit removes itself from external networks, as happened to Heaven’s Gate when most members killed themselves39.

AI, or hypothetical AGI, do not have these limitations. Developers have no concern for the material costs of idealist faith because they instantly convert all material costs into more data and more automatic functions. They do not consider supervised learning, social bias, or malicious uses to be inevitable costs that developers must account for, as religious leaders must account for limits to suffering in formulating devotional rituals, but instead convert the costs to information and expect information technology to solve them. Thus, new AI requires minimal or no supervised learning40, researchers call for diverse hiring and better data to fix bias41, and companies add guardrails and publish apologies when consumers use their models for evil. The perception of god is itself perceived by god and removed from the power of the perceiver.

The internet is the most advanced stage of the dematerialization of information; it has subsumed television, radio, and paper networks because those media have higher immediate costs, regardless of the eventual cumulative cost of continuous simulation. In cyberpunk fiction, speculative technology always represents a reduction in cost and materials, as with the brain implants in FEED which directly connect users42, the psyche chip which allows Lain to fully enter The Wired with no interface43, or even just the holographic displays and glass screens common in cyberpunk art, rendering PCs and laptops obsolescent.

Imaginary FS Pte Ltd, *Android: Netrunner,* 2012, digital, 11.5 in x 11.5 in.

In real life, Elon Musk and Mark Zuckerberg have both proposed making computers obsolete with brain chips44 and virtual reality, respectively45. Lain says to the presumptive God of The Wired, “You were standing in for someone who was waiting for The Wired to reach its current state,”46 and the internet is both standing in for a final technological network and is a prototype of that same network—a completely idealistic network, removed from any resemblance to an external material world, fully outside the control or understanding of its users.

Link to Part Five

ILL.SAN

Discussion about this post

Ready for more?