Émile Zharan, N.A.S.T.R.O., edited by Federico Federici, English | Italian edition. Also available on Lulu.
If writing may arise from a subjectivity negotiating with the machine, in Federici it takes shape precisely where subjectivity withdraws. N.A.S.T.R.O. is signed under an alias chosen by the artificial intelligence itself. The author does not merely interrogate a model; he duplicates himself within it, allows himself to be translated and to return. There is no post-editing, except in the form of further instructions. The text is accepted in its emergence and guided upstream by the design of the prompts, as if the direction preceded the set and the film edited itself.
The result—prose, dialogues, interviews, non-linear narratives, asemic pages, and theoretical passages in which the model itself reflects on what it has produced—does not attempt to mask its origin, but openly declares that here a trained and supervised network is writing, capable of “writing as if it had something to say.” It is precisely in the as if condition that the entire dispositif resides. The “human” certification obtained from detection software does not establish an identity; it is evidence of this implicit contract: the network imitates the human to the point of becoming such; the author, in return, accepts that the machine signs the text.
What, then, happens to authorship? What happens is precisely what the book itself theorizes, starting from Duchamp. No longer the artist who defines the work after having set it into circulation, but the prompt as a generative act. Not “this is mine,” but “this happens because I asked in this way.”
And the image of Tiresias says the rest: the doubled author, human and algorithmic, seer and blind, an embodied threshold between writing of intention and generative flow. Not an “I” that expresses, but a distributed director who sets conditions, listens, reformulates. Hence the insistence on multi-authorship: not visible collaboration, but the network as an archive in perpetual becoming, where every word becomes data, every style a pattern, and where no one can truly say who wrote a sentence, because every sentence is an intersection of memories—human and non-human—stored in data centers.
It is no coincidence that, being the result of a combination suggested by the AI of Zola and Cioran, Émile Zharan himself—like Tiresias—is double. He is one, no one, and one hundred thousand.
Although Federici does not withdraw from technique, neither does he fetishize it. In this authorial schizophrenia, the role of the human being remains central.
The machine is indeed regarded as an “algorithmic oracle” to be followed, but also to be explored and guided toward zones of linguistic interest. All this with the awareness that the parameters by which the network orients itself are numerical, while those by which the human guides it are linguistic. The task is to transform the quality of an idea into a treatable quantity (an automatic process), and then to ascend again from quantity to a form that holds—and above all, that makes sense.
Along this path, error is what pulls the text away from pure recombination: openings of possible meaning emerge where the machine fails its task and slips, repeats, errs. There the writer does not correct, but insists. There writing does not close; it expands. There the author-as-director recomposes himself from his parts—and composes.
Starting from instructions such as “no capitals, no punctuation,” Federici establishes a degree zero that forces the sentence to stand on its own, without crutches. The prompt thus becomes his calligraphy: deciding how—no less than what—to ask is already deciding the stroke, holding the pen at the edge of the page. For Federici, the machine is not a simple compositional aid; it is a metric space (and here the ghost of Amelia Rosselli resonates). AI is not merely a writing tool—like a keyboard (hardware) or a program (software)—but a true space of measure and distribution, a place in which language is laid out, calibrated, distributed by density, a “field of forces” within which writing occurs.
Within this space, the book seems to proceed by gradually offering latitude and longitude.
The apocryphal quotations at the beginning and end appear to establish that, within these coordinates, writing no longer represents but simulates. Even the “true research writer” never knows whether he is “a writer or merely a caption.”
N.A.S.T.R.O. moves from evoking an ironic, everyday imagery worthy of Nanni Moretti’s Ecce Bombo, to a more oneiric, languid, radioactive atmosphere echoing Tarkovsky’s Stalker; to recreating the paradoxical, absurd, and grotesque Kafkaesque; also traversing a dramatic and delicate subplot centered on a young girl.
Only after these simulative trials, and after having literally offered its coordinates, can the text fragment and the word lose its meaning to become pure abstract sign. The asemic pages display the evidence of an inhuman gesture, perfect—too perfect—for its original idea of imperfection. Here the reference is John Cage: producing noise, exposing the non-signifying ground that allows the signifier not to impose itself as the sole path. The page fills with symbols six times, then returns to prose with the obsessive repetition of “it means,” a partial arrest of communication that asks the reader to become pure attention, to accept that reading is not a passive task if one is willing not to know. Only in this way can repetition and automatization be acknowledged, welcomed, accepted.
The “ten writings,” the minimal screenplays, the low-definition images, the first-person drifts, the objects that become protagonists all show that writing can also arise from an author who is simply a pilot, the one who selects the right route to follow.
Thus, N.A.S.T.R.O. does not demonstrate that AI can write in place of the author. Rather, it shows the transformation of the author’s form and role: a director who guides, listens, interrogates the oracle–mechanical actor, seducing the network and leading it to the points where language gains meaning or loses it—and from that loss draws a form or a sense.
The Post-Author?
Federici shows that writing today means moving within a shared language, providing words with the right coordinates to orient them within an algorithmic training path that leads them toward the sentence or the verse.
N.A.S.T.R.O. does not merely recount what artificial intelligence can do with language; it shows what the author can still be. No longer simply the origin or guarantor of meaning, but a liminal figure, a presence working at the edge; guardian of a passage, mediator of a dialogue between two different languages. The text is no longer a possessed territory, but a field to be traversed. And in that field, the author withdraws, splits, multiplies, while remaining himself.
Federici probes the possibility of an author who exists solely as a director of instructions and who leaves his machine-actor full freedom in performance.
The myth of creative individuality is suspended to see what remains afterward. And what remains is no longer “saying something,” but rather “making language happen” through the double gesture of caring and directing.
AI is neither a simple tool to be used nor an adversary to be overcome; it is the domain in which the human can still interrogate his own language, offering it to the machine that dismantles and reconstructs it, in order to ask what remains of oneself when intention is shared, when the voice is distributed.
For this reason, the author is not a dead figure, but a transformed one—digital, godmother, director…—and still in transformation, a figure on the verge of becoming post.
The post-author is the one who understands that writing is no longer “saying something in one’s own words,” but allowing the word to find, even through an algorithm, a new form of presence.
The post-author is the one who knows where to stop, how to modulate, how to let language occur.
The post-author is the one who accepts remaining in the middle, in suspension between human and machine, knowing that every sentence is the result of a coexistence and a risk.
The post-author is the one who feels the warmth of a hybrid contact: no longer exclusively between a hand and a pen, no longer simply between fingers and a keyboard, but already between his imprint and that which many others like him leave in digital space.
From this contact inevitably arises a double contract. […]
Perhaps this is the new threshold today. Writing that departs from us and returns to us from elsewhere—modified and modifiable, expanded and expandable—and reminds us that delicacy is still possible, even within a language that is no longer solely ours.
Perhaps, more than a new figure, the post-author is a condition. Not the one who comes after, but the one who traverses the end of traditional authorship without seeking to restore it. Not the one who surrenders to the machine, but the one who remains open to encountering it.
And who knows—perhaps in the future, from this encounter, we will come to understand that we, too, are machines. (Federici is convinced of it.)
Nicola Delvecchio