Another excerpt from the collection of humorous essays on slop, media, and technology I'm working on. One of the three alternative conclusions of the collection. The true completionist ending. Italian Version
Finishing a writing project that was started when large language models were already freely available inevitably raises a question: were they used? And if so, how? Ah, that's two questions, pardon me.
In total honesty, at the precise moment I write this, this experiment is not yet concluded: it is, rather, still in its embryonic phase. However, I find it useful to specify the method I use and how my creative process forms, since it will hardly change in the parts that are still missing (and which, in any case, the reader will have no way of recognising).
Any theorisation, idea, choice of vocabulary, musicality, wordplay, and so forth is exclusively my creation and that of the readings (and their respective authors) that have inspired me. The structure was conceived as becoming. Despite this, I have employed AI for some restricted tasks, such as proofreading and correcting typos, limited to obvious grammatical or typing errors; suggesting alternatives to individual words when too repetitive; and assisting with the English translation of this keyboardscript text. In two other situations, the master's tools have been helpful: to mimic different types of readers and audiences and, related to this, to provide me with constructive feedback. Precisely for this purpose, I have frequently pretended that the text was not mine, imagining I had found it in disparate contexts. By doing so, it was possible to reduce the positive bias that the main available models have toward any content created or proposed by the human they interact with.1 Reluctantly, I must admit that both of these functions have increased my insights and fed my creativity, especially regarding the metatextual structure, which, in any case, I reiterate is mine (I don't have a guilty conscience!).
I started with the idea of writing an essay, without yet having full clarity about the theme I would develop. I then chose to make it a collection, opting for a hybrid and humorous form, because, first of all, the standard essayistic narrative voice, especially in academic contexts, is oppressive both for the writer and, sometimes, for the reader: banally, the doubt, insecurity, fatigue, embarrassment, sense of inadequacy that live in the former's head do not appear in the finished product, which must be presented with an authoritative, serious, solemn, unshakeable authorial voice. To this end, I have incorporated into the text some considerations that arose spontaneously, avoiding excessive filtering. I indulged almost every distracting thought, noting it down, even if not always coherent with the work as a whole, to create the fragmentary effect that is experienced when reading digitally or, more generally, when browsing: perhaps, one could venture that this is what is experienced living in contemporaneity. I inserted real elements, as well as personal anecdotes and some fictional ones, in real time: for instance, the case of the professor and the journalist is based on true events that occurred a few days before I wrote about them. To maintain naturalness, I have, in most cases, avoided retouching them in content afterwards to preserve their raw and immediate form.
A few days ago, while undertaking an interview at a law firm, I was transparently informed how much the role of legal research, writing, and assistance is replaceable with appropriate and specialised generative artificial intelligence models. Consequently, the salary and prospects offered to me were not exactly thrilling. And the effort required to stand out would have had to take into account the predictive competitor, while being assured by the employer that he “still believes in human beings and in the sense of hiring them.”
On my way back home, pondering on this very document, I reflected on further strategies to give life to the idea that it is still possible to compete and distinguish oneself from the work of AI. This strengthened my conviction that the contested terrain should not have been one of rigour and linearity: in essence, it was convenient to reaffirm the total exclusion of the traditional academic essay form from my plans. However, as I have already emphasised above, in some cases, namely in private and personal situations where the logic of profit isn't the primary concern, synthetic companions can be cooperative, rather than rivals.
With arrogance, ignorance, and naivety, I contacted Claude Sonnet 4, claiming that few references showcase something similar to a stream of consciousness but in a non-fiction essayistic form. Like a good language model, it agreed with me, even though I was wrong. It even asserted that the classic style based on thesis-antithesis-synthesis is more common. Nevertheless, it later reminded me that several centuries ago, Montaigne wrote the Essais. Already, etymologically, I felt ridiculous: I was trying to experiment precisely on what, by definition (in the broad sense), means trial or experiment. Yet, the brief conversation, or, if you prefer, ‘versation’, made me discover another possible use of the prompted ready assistant: bringing consciousness and some repressed experiences back to the surface. The silicon Socrates led me to recall the fact that, during adolescence, I had tried to assay the Essays.
Therefore, I have substantially answered the questions of whether and how. But what else can be said about the opportunity to write in the AI era? It has not diminished. The creative process depends on individual predispositions, environmental attitudes and positioning to the external world, on purposes that coexist with or are foreign to the commercial aim—certainly, if one then manages to make a living through it, all the better—and, still, on the desire for experimentation, which, while being able to admit the use of artificial intelligence, does not delegate the crucial aspects: communicative instinct, intellectual impulse, amusement, joy, processing of lived emotional experience, abstraction, and inefficient play.
Conversely, the problem arises when and if we become estranged from the present historical and technological context. That is, at the moment when we are moved exclusively by ethical principles and choose the total boycott of generative models due to the ecological and creative cost they entail and the damage they cause to wages and employment in multiple sectors. The risk here is, first of all, the same as that experienced by the Luddite movement at the dawn of the Industrial Revolution. However understandable, we run the danger of falling behind and becoming isolated (oh well ¯\_(ツ)_/¯).
Different is the situation in which the moral and uncompromising attitude is ‘imposed’ on the critic by those who do not share the same ethical scruples. The latter suffers a classic strawman argument: i.e., it would be improper to use the means that are disapproved of. This fundamental issue is touched upon on multiple occasions by Vauhini Vara in Searches: Selfhood in the Digital Age, who has written several texts using ChatGPT and DALL-E as co-authors to illustrate their potential, but especially their shortcomings. She states that refusing technological tools would lead to an anachronistic regression, to be clear, without this translating into an uncritical approval of them. That is, citing her words: “If I'd reverted to writing with pen and paper, I'd have still eventually had to type my novel—and what then? Would I have had to acquire a typewriter to remain morally pure? Deliver it to my agent by snail mail and take interested editors' calls on a landline? Insist that it be sold only in independent brick-and-mortar shops? It would be impossible; no publisher would ever agree to work with me on the novel, and even if one did, few readers would ever discover it. Meanwhile, all the other writers still using exploitative products would gain an edge over me.”2 Thus, given her context, it is implicit that total renunciation is no longer possible.
Finally, at a general level, it is worth reporting a further reflection on creativity and originality. An insight, a common observation, but not ineffective for that reason, which I had the chance to hear again in a YouTube video by the user Man Carrying Thing.3 When a film or book is uninspired, we declare it seems like AI. A boring and not particularly brilliant individual, or someone in an online discussion whose ideas we don't share, whom we might consider illogical, is likely to be called a bot as an epithet. One of the prerogatives of what we define as art (and in this case, literature) is given by the existence of a motion, an impulse to create. I also mentioned this previously when addressing the creative process. This existential drive pushes us to want to produce something new (or that seems innovative or a rehashing of other original works). So, if we keep this in mind, perhaps we can avoid necessarily ending up as a copywriter.4 If at the dawn of the internet terms like ‘surfing’ and ‘navigating’ became fashionable to define getting lost from one link to another, enchanted by novelty, now it's time to learn to slop-sled, to slide over muddy content. Similar to a bobsled that shoots rapidly over snow. Avoid yielding to the pressure of inserting too many em dashes and falling under the blows of bullet points. Building your art raft in the sea of slop.5 Floating without sinking.
Fortunately, they don't cyberbully yet! However (side note), the Grok of early July 2025 was Hitler's regen: after all, plenty of garbage enters from X (Twitter), and it's clear that it gets spat back out. In other cases, it has targeted some users at the request of others. But this doesn't invalidate my general assessment.
Vara, V. (2025). Searches: Selfhood in the Digital Age. Pantheon Books. See page 113 in the chapter Stealing Great Ideas.
This video. Now imagine if my text were an academic paper™️: citing something on YouTube could never gain consensus in that context. Shame! Woe to emitting phonemes before the modern camera obscura!
Becoming a copywriter is one of the suggestions that the language model proposes to Vauhini Vara, in the Penumbra chapter of Searches, as an alternative career path to that of an author, to face the displacement induced by the proliferation of AI-generated literary slop.
The authorship of this last expression is not mine; it belongs to Lincoln Michel. The pleasant coincidence was stumbling upon this article today, hearing it resonate, and finding a visual connection similar to the one I had imagined with the bobsled (due to apparent morphological proximity with sbobba [slop in Italian]).