The reign of the scriptor

Making the rounds this week was a mundane critique of AI that was notable mainly because it was made by musician Nick Cave, who was unsurprisingly unimpressed by ludicrous ChatGPT efforts to mimic his songwriting. Cave writes:

Songs arise out of suffering, by which I mean they are predicated upon the complex, internal human struggle of creation and, well, as far as I know, algorithms don’t feel. Data doesn’t suffer. ChatGPT has no inner being, it has been nowhere, it has endured nothing, it has not had the audacity to reach beyond its limitations, and hence it doesn’t have the capacity for a shared transcendent experience, as it has no limitations from which to transcend. ChatGPT’s melancholy role is that it is destined to imitate and can never have an authentic human experience, no matter how devalued and inconsequential the human experience may in time become.

The six strings will never draw blood from a machine!

It’s not even that I think this is wrong — of course you can’t ascribe artistic intentionality to predictive text apps — but its pat, reflexive humanism made me want to pit it against Roland Barthes’s 1968 essay “The Death of the Author,” which polemicizes (with an appropriately incalculable degree of irony) against the idea of “authentic human experience” and the lone creative genius as a means of expressing it.

Barthes’s aim in the essay was to shift critical attention away from authors and their intentions so that texts themselves could be interpreted more freely. The cult of personality that grows up around famous writers — the attention Cave’s post received is a good example — obscures the qualities of the writing itself and its generative open-endedness. “To give writing its future, it is necessary to overthrow the myth,” Barthes declares. “The birth of the reader must be at the cost of the death of the Author.”

Writing, he argues, is multivalent and intertextual, “made of multiple writings, drawn from many cultures and entering into mutual relations of dialogue, parody, contestation.” It dissolves the writer into many different speaking subjects all speaking simultaneously, into the incessant interplay of language itself. Readers reintegrate that multiplicity and hold it together in ways that are not constrained by a writer’s intentions; they are actively engaged rather than passive receptacles of the great writer’s aims.

Now we might say “readers” are ones who prompt AI models and make sense of their output. Generative AI doesn’t just produce text; it produces readers ostensibly liberated from the myth of authors. It makes reading itself tantamount to a kind of authorship, free play in a field of signifiers that can never congeal into signifieds.

It’s not hard to see ChatGPT as a literalization of many of Barthes’s provocations about the future of writing: “Writing is the destruction of every voice, of every point of origin.” Yes, ingesting and processing billions of texts into a statistical array will achieve that. “We know now that a text is not a line of words releasing a single 'theological' meaning (the 'message' of the Author God) but a multi-dimensional space in which a variety of writings, none of them original, blend and clash. The text is a tissue of quotations drawn from the innumerable centers of culture.” Yes, we do know that now, and OpenAI has even made a user-friendly interface for it.

The largest language models aspire to re-create the entire galaxy of intertextuality and make it navigable. They will no doubt be able to fill up as many blank pages as you wish with ready-at-hand text, at the click of a button in the “productivity suite” of your choice, all the whimsical play with signifiers you can afford with your stock of OpenAI credits. Just because Author Gods have mistaken the laboriousness of their haphazard excursions across this “multi-dimensional space” for creativity or originality or “authentic human experience,” that doesn’t mean that the rationalized, machine-assisted journeys we can now take are phony. Rather they are a convenient shortcut to experiencing the sovereignty of readership.

To supplant the Author, Barthes proposes something he calls the “scriptor.” “The Author,” he explains, “is thought to nourish the book, which is to say that he exists before it, thinks, suffers, lives for it, is in the same relation of antecedence to his work as a father to his child.” That is basically Cave’s view of writing, though he takes it further, labeling it not an act of gestation but “an act of self-murder that destroys all one has strived to produce in the past” — an endless project of self-overcoming whose meaning and purpose resides in the struggle of the creator to capture their essence, which is everyone’s human essence, really.

“In complete contrast,” Barthes continues, “the modern scriptor is born simultaneously with the text, is in no way equipped with a being preceding or exceeding the writing, is not the subject with the book as predicate; there is no other time than that of the enunciation and every text is eternally written here and now.” That may as well describe models like ChatGPT, which effectively demonstrate that there is no need for a consciousness or an inner motivation for language to unfold itself; its own structure — analyzed as a closed system of purely probabilistic relations — suffices to make it appear meaningful at any given moment. Everything a model “writes” is provisional and spontaneous, unmoored by any attempt to have a specific “message.”

In another essay, Barthes makes a parallel distinction between “works” and “texts” —or rather Text, with a capital T. “Works” are what authors write, and the reception of them is trapped in the paradigm of trying to decoding the author’s true intentions. The “Text” is more or less what chatbots produce: a play of signifiers that is unrestrained by intention but instead offers an occasion for a reader to perform a synthesis that Barthes likens to the stroll of a “passably empty subject” in an unfamiliar valley. He describes “the Text” as

woven entirely with citations, references, echoes, cultural languages (what language is not?), antecedent or contemporary, which cut across it through and through in a vast stereophony. The intertextual in which every text is held, it itself being the text-between of another text, is not to be confused with some origin of the text: to try to find the “sources”, the “influences” of a work, is to fall in with the myth of filiation; the citations which go to make up a text are anonymous, untraceable, and yet already read: they are quotations without inverted commas.

And what is a large-language model, if not the automated process of that always already accomplished prior reading, which has stripped away the citations and made them anonymous, seeking to save us from the temptation of trying to reverse-engineer “sources” and “influences” and fixed meanings? “Succeeding the Author,” Barthes writes in “Death of the Author,” “the scriptor no longer bears within him passions, humors, feelings, impressions, but rather this immense dictionary from which he draws a writing that can know no halt.” The endless logs of ChatGPT sessions speaks to that. “Life never does more than imitate the book, and the book itself is only a tissue of signs, an imitation that is lost, infinitely deferred.”

I can’t tell if that is supposed to be a good thing — it sounds a lot like the fear that AI will create a recursive postmodern nightmare world of perpetual sameness that people will accept because they no longer remember otherwise or how to create an alternative. But Barthes seems also to be suggesting that this infinite deferral — and not the hallowed sanctum of the solitary Author — is where the genuine space for creative freedom in life lies. To immerse oneself in the Text is to exceed the limits of the individual self and dissolve the distinctions between reader and writer, past and present, creation and absorption, fact and fiction, and whatever other binary you wish to deconstruct. “Writing, by refusing to assign a 'secret,' an ultimate meaning, to the text (and to the world as text), liberates what may be called an anti-theological activity,” Barthes writes, “an activity that is truly revolutionary since to refuse to fix meaning is, in the end, to refuse God and his hypostases — reason, science, law.”

Part of the mission of commercial AI seems to be to betray that revolution. At the time Barthes was writing, the integrity and nobility of the “human person” — potently symbolized in the autonomy and creativity of “the Author” — was viewed as, in Barthes’s words, “the epitome and culmination of capitalist ideology.” It was seen as an alibi for all the exploitation and alienation intrinsic to the administered consumer society and the principle behind market-based ideas of freedom. Thus to create the space for other conceptions of freedom, it was necessary to disenchant the idea of the Author, much as it seems recurrently necessary to debunk the idea of the “artist,” which frequently stands in for a bourgeois fantasy of unalienated work or serves as propaganda for an entrepreneurship of the self.

Foucault, in the 1969 talk “What Is an Author?” made a similar argument, that the author is “the ideological figure by which one marks the manner in which we fear the proliferation of meaning.” He claims that “we can easily imagine a culture where discourse would circulate without any need for an author. Discourses, whatever their status, form, or value, and regardless of our manner of handling them, would unfold in a pervasive anonymity.”

Yet one could imagine Microsoft incorporating a version of this idea into their marketing materials for the coming ChatGPT-enhanced version of Office. If, as Barthes claims, “it is language which speaks, not the author” and that to write is “to reach that point where only language acts, ‘performs,’ and not ‘me,’” then why not put “language” to work instead of human employees? Most writing tasks are purely functional and formulaic, and the inadvertent injection of human personality into them may actually impede their effectiveness. You don’t need the “audacity to reach beyond your limitations,” as Cave puts it, to conduct most business correspondence. Rather than having to mystically draw from the ambient field of platitudes when we have some mundane writing task to perform, AI models can just disgorge them as required. The commercialized version of the death of the Author would thereby help further the instrumentalization of language, which is not the “birth of the reader” so much as the mutual obsolescence of reading and writing in favor of operational code that parses itself.

Commercial AI seems to appropriate and recuperate Barthes’s kind of critique — it turns poststructuralist ideas about the decentered subject and intertextuality into potential business models. It seems like an extension of Luc Boltanski and Eve Chiapello’s argument in The New Spirit of Capitalism that neoliberalism gained traction by seeming to address the “artistic critique” of capitalism (the idea that the problem with capitalism is that it makes life “inauthentic”) while marginalizing the “social critique” (the fact that capitalism exploits and perpetuates social divisions and inequality). That played out in social media (something I wrote about a lot a decade ago).

Commercial AI addresses (and tacitly appropriates) the critique of the appropriated artistic critique, that“artists” are just ideological show horses for neoliberalism and what is necessary is an attack of aspirational individualism and the valorization of the “creator class.” Rather than prompting entrepreneurial selves to monetize their subjectivity, as with the business models of social media, generative AI wants to harness the value of negating subjectivity altogether, part of the behaviorist turn marked also by the saturation of society with other kinds of algorithmic decision-making. This means that there is no escape from the productivity demands of having a personal brand or being online through embracing “vibes” or “corecore” or some other “death of the author”-like expression of the surrender to sorting algorithms. That just feeds a different tech profit center.

But at the same time, AI hype builds itself in part on fear mongering over its supposedly irresistible capacity to overwrite and abolish human creativity, an effort to foreclose on the possibilities of viewing generative models as liberating us from the onus of subjectivity and the exclusions of the category of the “human” and bourgeois possessive individualism and so on. The hype feeds on, depends on, reactions like Cave’s, which reinstate the Author but only as the remote, rarefied figure that is out of reach for most people, or serves as a tyrannical ideal driving them to “do what they love” until they are bled dry. A subterranean appeal of generative AI is the promise that it will level down the “elites” without requiring anything from “non-elites” — a pseudo-revolution from above to dissipate revolutionary ferment.

The wonder that “AI” evokes comes preconditioned and constrained to fundamentally boring feats of content production, showing us what it looks like to create without imagination so we convince ourselves that it would be more expedient to give up on that possibility altogether. As John Herrman pointed out in this New York magazine piece, the new wave of AI presents itself as a kind of “magic” that quickly reveals itself as a management tool: “The funniest chatbot in the world will lose its mystique pretty quickly when your employer decides it should make you 45 percent more productive or when it shows up in Microsoft Word dressed as a paper clip.” But that could also be felt as a relief. No one is consigned to jouissance by the new technologies, just an intensification of ordinary anxieties. Our subjectivity isn’t dispersed by using generative AI but expedited. In streamlining and functionalizing our relationship to language, it makes the whole process of being a subject constituted by language that much more painless and rote. That wasn’t the ideal that “The Death of the Author” was aiming for, but I do wonder if generative models extinguish the dream that essay articulates by fulfilling it.

It seems more likely, however, that generative models (and the hype surrounding them) will not lead to “the death of the Author” in the sense Barthes means, but will re-enchant the various illusions of “real presence” and human agency. One could foresee the proliferation of automated text prompting a kind of human “pivot to video” and visual evidence of human performance will become a requisite component of exchanges commercial and personal that are supposed to have value or be meaningful or resonate with “authenticity.” Text would cease to be a viable mode of “human” interaction; there will be “nothing outside the text” because there will be nothing left inside it.