Walker Publishing Filed to Design · Part of Content and Its Discontents

AI Generated images created using prompts by Chat GPT on how to best illustrate the article by Aaron Anderson and Eric Timothy Carlson.

Digital Demonology:
On the Auto-Production of Content

by Matt Colquhoun

In early 2023, I asked ChatGPT who I was. This is somewhat embarrassing to admit, as an act only one step removed from Googling oneself, but the response provided by this artificial intelligence was both surprising and inspiring. It hardly confirmed my sense of self, but rather called it into question.

An open-access version of the chatbot had been released by OpenAI late the previous year, on a website where the curious were invited to independently test its capabilities. The bot garnered significant media attention shortly thereafter, as many journalists and Internet users found themselves impressed by its ability to quickly generate cohesive copy in plain English. Others were less impressed, however, suggesting that, beyond the novelty of its presentation, the bot was little more than a glorified search engine. The content it produced was also far from accurate, which worried experts in various fields at a time when misinformation was already a major problem online.

But ChatGPT is nonetheless still fun to play with. Chatbots have long fascinated users for the uncanny nature of their mistakes, after all. When encountering any previous and more obviously flawed artificial intelligence, such as SmarterChild, many will have spent their time actively trying to confuse them, fascinated more by their errors than by their attempts at factual accuracy. Admittedly, this is how I initially approached ChatGPT as well, and although it was considerably more lucid in its responses than the chatbots I remember wasting hours with in my youth, it did not disappoint when it came to the production of entertaining untruths.

In response to my own question, “Who is Matt Colquhoun?” most of the chatbot’s information was broadly correct, but on its first attempt to give a summary of who I was, it notably made two factual errors. First, it claimed I was a lecturer in digital media and culture at the University of Gloucestershire in the UK, which is false. But the second error was more interesting: it attributed a nonexistent book to me, supposedly published in 2020, called Xenogothic: An Unwell Guide to Theory.

Though “Xenogothic” is the name of the blog I have run since late 2017, this subtitle is not something I have ever written down or discussed in such terms (at least to my knowledge), and so I was intrigued. I asked ChatGPT what this book was supposed to be about.

“Xenogothic: An Unwell Guide to Theory is a book by Matt Colquhoun that explores the intersections of philosophy, politics, and culture through the lens of . . . ‘unwell’ or ‘dark’ theory,” the bot explained. “In the book, Colquhoun draws on a diverse range of sources . . . to argue for the importance of confronting the darker aspects of our world and of theory itself.” I supposedly conclude “that we must engage with the uncanny, the monstrous, and the unsettling if we are to understand the complex and often contradictory nature of contemporary culture and politics.”

It was fascinating to me that, whilst this book did not exist, it certainly described a book I might well have written. But the imagined book’s contents appeared more in line with the thought of the late Mark Fisher, particularly his final book, The Weird and the Eerie, in which Fisher employs his titular concepts to elucidate and extend Sigmund Freud’s theory of the uncanny. Discussing their implicit potentials in fiction, music, and film, Fisher argues that these affects can illuminate various egresses from a quotidian sense of reality—or, more specifically, drawing on Fisher’s conceptual armature, an escape from the ideological confines of “capitalist realism.” Such portals to the Outside are rendered terrifying in many of the fictions in which they appear, but “terrors are not all there is to the Outside,” Fisher argues. These encounters with the uncanny can just as readily be productive, encouraging us to think differently.

What was all the more intriguing about the bot’s summary of this book I had not written was the way in which it functioned as one such portal to the Outside in its own right. This conversation with ChatGPT was precisely the sort of uncanny encounter that Fisher may well have written about, once upon a time, as he communed with the cybergothic “demons” and digitally occulted forces that were making themselves known and felt during the early years of the Internet and its popular adoption by society at large.

These eerie entities most often appeared in hypertexts produced collectively by the Cybernetic Culture Research Unit, or Ccru, a gathering of “renegade academics” active at the University of Warwick in the late Nineties and early Noughties, of which Fisher was an integral part whilst writing his PhD thesis. Made up of both postgraduate students and teaching staff, equally frustrated by the creeping neoliberalisation of a once-radical Cultural Studies, the group’s website remains an (oc)cult treasure trove of Y2K theorising, corrupting academic philosophy with a range of cyberpunk sensibilities, weaving high theory with low culture, willing into being a new Department for the Inhumanities at the dawn of a new technological era.

The Ccru were fascinated by the ways that new knowledges and technologies, rather than clarifying the world in which we live, only reveal it to be more strange than previously thought. The unit’s para-academic uploads exacerbated this fact, with their writings hosted on platforms that were still novel and unruly, published at a time when the Internet’s impact on society at large was far less certain than it is today.

The Internet of the late Nineties was, in many respects, a space that blurred the lines between scientific fact and fiction. It was the Matrix made real, a speculation from a William Gibson novel made newly “flesh.” It was a present future, where science fiction would proliferate on a technology that itself seemed like something out of a sci-fi narrative. As Fisher later wrote on his k-punk blog, many works of science fiction can demonstrate “that an enquiry into the nature of what the world is like is also inevitably an unravelling of what we are,” and the Ccru’s engagement with the Internet made this kind of enquiry all the more destabilising. Indeed, they took this cultural exploration of a technologically encouraged subjective indeterminacy very seriously, and alongside the moral panic of Y2K—the widespread superstition that the arrival of the new millennium and our calendric return to 00 would break our entire informational superstructure—they emphasised the emancipatory potential of emerging technologies, tearing asunder the political stasis that followed the so-called “end of history,” particularly for contemporary generations of women and racialised minorities, thus influencing intellectual and cultural movements like afrofuturism and xenofeminism.

But we live in a different moment now, it seems. The inchoate potentials of a more anarchic Internet often feel foreclosed by the new Big Tech powers that dominate its further development. And yet, although we look back on Y2K panic with a certain condescension, as a product of irrationality from a more naïve digital era, we can recognise how plenty of other new technologies make us anxious in similar ways, as we fear technology’s continuing effects upon our sense of what it means to be human. ChatGPT and its cousins—like the AI image generator, DALL·E—are surely no exception. But these AI tools hardly seem like the beginning of some great socio-cultural upheaval. In fact, through the capacity to automate the production of digital “content,” there is a latent fear that these new technologies will simply reproduce an all-too familiar cultural landscape for us. Just as Fisher often spoke of “the slow cancellation of the future,” by drawing on cyberspace’s now-abundant data sets, AI may not produce a new world but rather automate the reproduction of a culturally capitalist stasis.

Alongside this bubbling anxiety, there is nonetheless some excitement in other quarters, as early adopters of these technologies, in more creative circles at least, have already demonstrated their potential for producing newly machinic ways of seeing. Only a bad workman blames his tools, as the old adage goes, and so there are other ways of thinking about the production of content in the present that may well revitalise a popular culture already beholden to the limited purview of algorithmic curation.

It was for this very reason that I found ChatGPT’s error inspirational. What if I did write the book that this bot had attributed to me? What would such a book look like? Maybe it was a book that I should write, at its encouragement. What if I allowed this error to be generative, allowing it to function as what the Ccru would no doubt call a “hyperstition”—a fiction that makes itself real?

Initially, I decided to correct ChatGPT. The book in question does not exist, I told it. Its response was banally encouraging. “Writing can be a fulfilling and rewarding experience,” it said, “and if you do choose to write a book, I wish you the best of luck with your project.”

It was heartening (if slightly ironic) to be encouraged to write by a chatbot that many feared would further reduce whatever value writing has left, as digital publishing has already upended more traditional media in recent decades, making writing a less sustainably profitable profession. But we might argue that writing, like any other creative endeavour, has always been embroiled in a convoluted relation with strangely inhuman forces. Fisher himself, extending the theories of the Ccru, often spoke of his k-punk blog as a space to channel a writing compulsion that emerged through but otherwise outside of himself.

Fisher was likewise fascinated by how his own desire to write came from “the Outside”—the Outside of the self that he was, flowing through the inaccessible and “impossible” realm of the Real. Asked how he managed to be so productive online, he once wrote that the “answer is that it isn’t me who’s writing.” This was not modesty or metaphor, he insisted, but a “strictly technical description of how this body has been used as a meat puppet for channeling uttunul signal.” The writing is only “good,” he suggests, when a subjective “I” has been wholly abstracted, becoming little more than “a space through which Lemuria speaks. The writing is already assembled on the plane and all ‘I’ can do is bodge it by introducing subjectivist fuzz.”

Here Fisher is notably referencing a number of terms and eerie entities first conceptualised by the Ccru, which populate its Lovecraftian mythos of clandestine mind-control programmes, mathematically demonic forces and Burroughsian science-fiction cut-ups.[1] But the wilful obscurity of this terminology aside, it is a description of Fisher’s writing process that continues to resonate with the digitally occulted forces that make us anxious today, suspending any sense of our creative agency in cyberspace. But writing has often been thought of in this way, as an oddly inhuman compulsion. To write, after all, as many a writer will tell you, often feels like being taken over by some generative force from nowhere—whether felt as an incursion from outside the self or produced by the unconscious and more inhuman forces at work within. It is writing itself, therefore, that both communicates a sense of self and attempts to flee the self and world, engaged in a dance along a peculiar precipice, as the speaking subject is punctured by holes that give way to each side of this human/inhuman divide.

As well as containing references to the writings of the Ccru, Fisher’s descriptions of the forces that drove his blogospheric productivity also echo the philosophy of Gilles Deleuze, who admired many of the same writers Fisher did. As Deleuze once wrote, in collaboration with his former student, the French journalist Claire Parnet, so many works of Anglo-American literature engage in a similarly eerie process of unsettling escape. For better or for worse—whether in favour of fundamental change or a kind of imperial expansion—English literature has often elucidated its own various lines of flight from the impositions of this world, internalised by the self in turn. Deleuze and Parnet discuss an Anglo-American literary modernism in particular, through which so many writers hoped to flee not only the conventions of literary form, but also the enclosures of social and political life in the early twentieth century.

But how to retain a fidelity to this line of flight without collapsing into its opposite—that is, how to maintain a line of flight’s productivity against its innate tendency to give way to processes of (self-)destruction? “What is it which tells us that, on a line of flight, we will not rediscover everything we were fleeing?” Deleuze and Parnet ask. “How can one avoid the line of flight’s becoming identical with a pure and simple movement of self-destruction . . . ?” There are examples of such a linear conflation everywhere in modernist literature—from Ezra Pound and Wyndam Lewis’s flirtations with fascism to “Fitzgerald’s alcoholism, [D. H.] Lawrence’s disillusion, Virginia Woolf’s suicide, Kerouac’s sad end”—but Deleuze and Parnet consider instead how it is writing itself that hopes to steel the author against their own demise. Writing is thus a kind of processual double-bind—a line of flight that leads to death, but which is nonetheless undertaken to forestall the true arrival of that same end. Writing thus kills the self that writes, but this death may not be absolute. It can just as readily lead to a rebirth of the self through writing.

It is thus possible, they continue, “that writing has an intrinsic relationship with lines of flight”; it is an activity with “a special function . . . the function of the Anomalous.” It is an active process of imagining other worlds and selves, of shedding the world and self that constrain the writer in order to prefigure new ones. “On lines of flight there can no longer be but one thing, life-experimentation.”

The anxieties produced by an entity like ChatGPT remain acutely relevant here. If an anomalous writing comes from without, what are the implications of automating its production absolutely? The Ccru were hardly worried about the implications of such a technological dehumanisation of language. Indeed, inspired by the post-structuralist theories of Jacques Derrida and Jacques Lacan, it was this same tendency that they hoped to sustain in a newly digital era, affirming the role of “cybergothic hyperstition” in the cultural production of content on the Internet. (As Lacan once proclaimed: “Every truth has the structure of fiction.”)

Hyperstitions are thus not “representations, neither disinformation nor mythology.” They are products of a seemingly autonomous process of market hype. “Hype, hyping, hyperpropagation belongs to a strain of time-warp cybernetic fiction,” writes a pseudonymous member of the Ccru, Iris Carver, “that cannot be judged true or false because it makes itself real.” Quickly popularised Internet phenomena, like memes or tools like ChatGPT, thus catch the waves of algorithms that are hooked on our most unconscious desires, as online culture is autogenerated by vaguely understood but nonetheless linguistically codified forces. We should think of cyberspace, then, “as a black-mirror. It is where time flips over: collide with it and you travel backwards. As telecommerce accelerates us into the net, it seems that things of ever deeper antiquity awaken, and begin their return.”

Updating this observation to the present, we might perceive ChatGPT as a kind of Ouija board for Web 3.0. Just as Sylvia Plath, in her occulted experiments, authorised the planchette on which she rested her fingertips to speak for her, as a “mother of otherness,” AI tools need not replace human communication but make it weirder. But we may (and perhaps should) remain suspicious of the forces we allow to speak on our behalf. They are attractive precisely because they allow us to relinquish a suffocating subjectivism, but their utterances may nonetheless come to control us in other ways. As such, though AI tools resemble newly digital lines of flight, as eerie entities of content production, we must remain attentive to their tendency to collapse in on themselves, merely reproducing what we might initially hope they will allow us to flee.

This sort of collapse appears all the more likely following the capture of the trading of cryptocurrencies, such that this alternative form of economy has been subsumed within already well-established stock markets, or the ways that NFTs are reduced to a humiliated form of postmodern art-market commodity. A few years previously, we were most aware of this capitulating tendency in cyberspace when it produced reactionary horrors of a more acutely political nature. From Richard Spencer’s infamous claim that the alt-right “memed” itself into existence (and Donald Trump into the White House) in 2016 to the more general proliferation of online terrorism, the swarmachines of online cultural production disturb us greatly. In the last essay published before his death in 2017, Fisher argued that the propaganda videos made by ISIS, for instance, might themselves be “understood as a cybergothic phenomenon which combines the ancient with the contemporary (beheadings on the web).”

But the proliferation of such horrors from some assumedly prior spacetime of violent barbarity is nonetheless a product of capitalism’s own twisting of our sense of contemporaneity, since capitalism has long since retreated from the present, never mind the future. Left to its own resources—or rather, left to the resources it retains from previous forms of exploitation—capital can never come up with anything new. Postmodernism was its ideal form, and the naturalised postmodernism of capitalist realism was its optimal solution to political and cultural antagonism.

But who is not aware that, in the present, political and culture antagonisms have reached newly excessive levels of incoherent frenzy and discontent. Deleuze and Parnet's prior warning that any line of flight can succumb to self-destruction is epitomised in capitalism’s humiliated sense of its own progressivism in this regard. What is even more worrying is that so many figures who once inspired an exploration of the radically new and unknown have betrayed their own lines of flight in much the same way.

It is unfortunately true that a number of the Ccru’s former members have similarly given in to more negative processes of self-destruction. Beyond Fisher’s suicide, we find the work of Nick Land, another integral member of the group, who came to wider prominence following the election of Donald Trump in 2016, as he was cited as an influence on a number of alt-right figures. Though many writers fought to retain the more emancipatory politics laid out in his early works—in which Land wrote on the radical potential of xenofeminist lesbian “K-guerillas [who] stalk each other through labyrinthine erogenous zones, tangled in diseased elaborations of desire,” amongst other things—he later turned toward a more “neoreactionary” political philosophy and an archipelago of Twitter accounts laden with racist epithets and techno-fascist ideations. The fight to retain a progressive Landianism has been lost, due to Land himself reneging on the potentials that once fascinated him at the dawn of our digital era.

But the Ccru’s desire to radically conceptualise, diagnose, and make unhomely the indeterminate potentials of new technologies remains in reach for us. Though many have failed their prior impulses, we can and must remain vigilant. This is true enough when we consider the future of writing itself, as a medium made newly contentious and problematic by its accelerating automation. As Deleuze argued repeatedly, writing is itself a process that generates an outside for any writer. “In reality writing does not have its end in itself, precisely because life is not something personal,” as Deleuze and Parnet write; “the aim of writing is to carry life to the state of a non-personal power,” such that writing “renounces claim to any territory, any end which would reside in itself.”

The content produced by ChatGPT may, at times, affirm the opposite. It is an automatic writing that may be utterly impersonal, but which is nonetheless incapable of producing something radically new and unfamiliar. But this negative appraisal may, in itself, be useful to us. If cyberspace is a “black mirror,” AI’s automation of capitalist stasis may well free us from any expectation to adhere to its universalising of a particular mode of communication. The paradox of much modernist writing is that it uses language to try and renounce our reliance on language itself. This was particularly true of a writer like William Burroughs, who believed that, as Fisher once argued, “the language virus is now positively harmful,” since it forces “thought into patterns which impinge upon the behaviour of the host.” Burrough’s cut-up experiments attempted to intervene within this process, redoubling writing’s aleatoric (characterized by chance or indeterminate elements) tendencies. But for many, the cut-ups were a failed experiment. Though they provide us with an utterly alien reading experience, they are hardly easy things to read and appreciate. When Burroughs discussed his technique with Samuel Beckett, Beckett replied: “That’s not writing—it’s plumbing.” It is little more than a method for processing literary excrement. But as Fisher later countered, “Engineers are always more productive than artists . . . .” Outsourcing the capitalist production of mere “content,” of “bad writing,” might well leave more room for us to produce something truly new—if only we remain attuned to the things that AI itself cannot do.▪︎

Matt Colquhoun is a writer and photographer from Hull, UK. They are the author of two books, Egress and Narcissus in Bloom, and the editor of Mark Fisher’s Postcapitalist Desire. Currently a PhD candidate in Philosophy at Newcastle University, they blog at xenogothic.com.

[1] A “meat puppet”, for instance, is someone taken over by the mind-control programme Monarch, which attempts to “‘trance-form’ the mind and personality of their subjects”; a fictional institution that may signify some late-capitalist Big Other. “Uttunul” is one of the demonic forces associated with the Pandemonium Matrix and the occulted cartography of the numogram; an entity of the “Seething Void”, a “Syzygetic Xenodemon of Atonality”. “Lemuria”, otherwise known as the “Land of the Lemurs”, is a location that appears repeatedly in the works of William S. Burroughs, synonymous with that particular species of animal, unique to the island of Madagascar, which has a very particular evolutionary history, having adapted to the conditions of that small and isolated island. They are thus seen as a kind of living fossil; “treated as relics, or biological remainders of a hypothetical continent: living ghosts of a lost world”.