The Technology of Writing: From the Essay to GPT-3

by Derek Neal

I write this essay as much for myself as for the reader. It is my conviction that one writes to find out what one thinks, not to put down fully formed thoughts that are floating inside one’s head.

Some sort of alchemy occurs when I put pen to paper, or in this case, pen to screen, as I set down the stuff knocking about my brain and give it a more solid, permanent form. But why do I insist that what I write comes from within me? To say that my words flow from my own head, down my arm, and into the writing instrument is simply the representation of a process I don’t fully understand. The bards who sang epic poems in ancient Greece did not view their creations in this way, as coming from within, but as being inspired from without, inspire in this case taking on its original meaning: to breathe into. The poets began their stories by invoking the gods, or muses, in the hope that the spirit might be blown into them, filling them up and allowing them to translate that spirit into words and music for the benefit of an audience. It may be that we could also think of writing in this way. Since I’m writing an essay, I might invoke the spirit of Montaigne, call upon him to breathe life into my pen and help shape my words, but it may also be that literacy itself precludes this, that literacy and the written mode of thought are fundamentally interior activities, a conversation with oneself, and that something about the written word lends itself to being thought of as coming from inside of oneself, whereas the spoken word seems to come from “out there,” with the speaker being a vessel giving form to something of which they are not the origin.

Whatever the case may be, I could use Montaigne’s help in this writing attempt, remembering that “essay,” the genre Montaigne is credited as having invented, means “try” in French. I am trying here to understand why, and to make the claim that, one must still write. Perhaps you did not know this was up for debate, but then you haven’t been paying attention. When I was learning French in college, it was understood by most students that translating essays with websites like Google Translate was futile. The technology at that time would translate word for word from English and so your French would end up following English grammar rules while also featuring French vocabulary that you had no way of knowing. It was clunky, it often made little sense, and the teacher immediately knew it had been translated. This is no longer the case. Translation software now understands how to group words together, and thanks to machine learning and AI, it can create natural sounding language that is more fluent than what most intermediate and advanced language students can produce. Now a teacher knows that a student has translated their writing because it is simply too good.

In addition to translation software, there’s also paraphrasing software, which will take a sentence from an article and spit out a brand-new one, complete with different vocabulary and grammar, all while maintaining the meaning of the original. The new sentence can then be adjusted to meet the user’s taste: more or fewer synonyms, increased creativity, less formality, etc. While someone may use this software to paraphrase an argument from an article they’re using as research in an essay, a more likely use may be to pass off the work of others as one’s own while avoiding plagiarism detection software like Turnitin. This software works by matching strings of text in a student’s submission to other submitted essays or anything on the internet. With paraphrasing software, these programs are evaded. As technology continues to progress, students and teachers enter into an ever-escalating arms race where each new piece of software responds to the advances of previous iterations.

If you think I’m exaggerating, consider the experiment run by Ethan Mollick, wherein he prompted GPT-3, an AI built to model human language, to create an essay question and the rubric for the essay, write the essay, then give itself feedback on the essay. The AI gave itself an 80. One wonders what the AI generated grade complaint would have looked like, but we can be sure that it would have been well written, because the shocking thing about this exercise is just how high quality the work is, not just the essay but the rubric and feedback as well. In the non-academic sphere, there are examples of AI (GPT-2) writing a New Yorker-esque” paragraph, and more recently, a paragraph in the vein of Kafka. The truth is, AI is at the point where it can write better than most of us, and certainly better than the incoming college freshman. So, why write?

In a university writing course, the emphasis is often on improving one’s writing “skill.” These skills are said to be necessary for future jobs in which one must correspond with colleagues, clients, or customers in writing. This framing of the purpose of writing is a microcosm of the university as a whole, as the university has become a place for skill development, which will help one find a job upon graduation and justify the investment of a university degree. However, this justification for writing falls flat when computers can write better than we can. The proponents of writing as an isolated skill are fighting a losing battle in the same way as people who try to defend studying the humanities by saying it will help students be more desirable job candidates. In making this argument, defeat has already been conceded. The marketplace and the study of the humanities are antithetical to one another. In fact, the marketplace and the university are antithetical to one another, if we consider that the original purpose of the university, with humanistic study at its core, was the cultivation of one’s ability to think and reason, not to build marketable skills.

The university should function as a bulwark against capitalism and the idea that the defining value of something is the monetary gain it creates. It still does, to a certain extent. Deep in the recesses of an old hall, a crusty professor may be found, toiling away at something whose utility may not be translatable into dollar signs, but which serves the goals of wisdom and knowledge. These teachers, the ones we remember from when we were students as having impacted our understanding of life in some deep way, as having opened our minds and enlightened us, or shown us the path to enlightenment while encouraging us to walk down it, are still there and will always be there to some extent. The university is a refuge for them from the cruel realities of the outside world. On the other hand, there are people like my neighbor, a post-doctoral researcher in virology who recently attempted to explain why the erosion of tenure was a good thing, as it keeps people “on their toes” and ensures that research is useful and practical, not some nonsense with which an absent-minded professor might busy themselves. My neighbor is a good guy, I assure you, but his views of what matters and what doesn’t is determined by the logic of the market. His argument means that if one’s interests are not en vogue at the present moment, not trendy, not marketable, then they won’t be funded as they won’t fill the coffers. His argument is reminiscent of the gig workers who fight against their classification as employees, helping to destroy that which might sustain them.

But to return to writing: when writing becomes a skill, we are no longer able to justify its inclusion in the university classroom, as technology can write better and more quickly. Considering this, it is likely that writing and composition will disappear as a subject of study much in the way that other disciplines that have lost their usefulness, such as Latin, have disappeared. If writing can be outsourced to the computer, why write?

In order to understand why we write, why I’m writing this essay, why I implore the students I teach to write as much as possible, we must remember that writing is not simply a tool to express ourselves, but a mode of expression that shapes our consciousness. To write is to think in a certain way that is not possible when communicating orally, and to lose this ability, or to never cultivate it in the first place, is to be a lesser version of the person one might be. This is the claim made by Walter Ong in his seminal 1982 text, Orality and Literacy: The Technologizing of the Word.

Ong reminds us that writing is a technology. Thought of in this way, AI like GPT-3 can be seen not as a replacement for “natural” writing, but as the continuing evolution of literacy, or as Ong says, “the separation of the word from the living present.” AI does not replace the need to learn to write but becomes an activity on its own with its own motivations and goals. Any new technology allows us to reconsider the previous technology it threatens and to view it in a new light; in fact, this is how literacy allowed us a deeper understanding of the ways in which orality shapes the mind and its thought processes. In contemporary media, the revival of vinyl collectors in the face of streaming services shows us that the medium through which music reaches our ears is not neutral; having an algorithm decide the music we listen to has led many people to realize one of the purposes of vinyl, or even of CDs, is to help cultivate a sense of taste and aesthetic judgement of an artistic form. Music on streaming services is often not about the appreciation of the music itself, but music in service of other purposes: music for studying, music for workouts, music for coffee shops. The same case could be made for film.

Let us use the emergence of these new writing technologies to reflect on what literacy is and how it differs from orality. In Orality and Literacy, Ong draws on the work of Milman Parry and Albert Lord, who recorded thousands of hours of epic poetry recited by illiterate poets in the Balkans, to show that the oral mind functions differently from the literate mind. Parry was the key figure in solving the question of the authorship of The Iliad and The Odyssey, revealing that they had no author at all, but would have been recited by countless singers all using techniques of memorization, repetition, and improvisation to tell the story a new way each time they recited it. Ong summarizes oral thought by noting that “knowledge, once acquired, had to be constantly repeated or it would be lost: fixed, formulaic thought patterns were essential for wisdom and effective administration” (24).

In contrast to oral thought, which Ong also characterizes as close to the human life world, literate thought is characterized by its abstractness and its distance from lived experience. In fact, Ong cites Eric Havelock, who makes the case that Plato’s philosophy of ideas and forms was only possible because he had interiorized writing and it had transformed his consciousness. Ong writes that “nothing of Plato’s analytic targeting on an abstract concept for justice is to be found in any known purely oral cultures” (102). Understood in this way, writing comes to be seen as the vehicle through which our contemporary understanding of the world is possible, built as it is upon abstract concepts such as truth, justice, and human rights. We might also note that the Flynn Effect, the phenomenon of increasing IQ scores throughout the 20th century, was noted as societies industrialized and became more fully literate. Since IQ tests measure abstract thinking, it’s no wonder that literate peoples would perform higher on them.

And it’s not just our understanding of the world that is shaped by literacy, but our understanding of ourselves, too. Ong, along with Havelock, notes how writing allows us to regard ourselves from the outside, in effect to make oneself an object of study. In one particularly fascinating passage, Ong notes how the personal diary allows for a sort of introspection impossible to conceive of for the oral mind, but also how the diary writer may become filled with anxieties due to an overabundance of introspection, commonly known as “thinking too much,” hinting at just how powerful writing can be on one’s thoughts.

The most convincing defense of writing is made by Havelock in Preface to Plato, wherein he writes:

there was a state of mind which we shall conveniently label the ‘poetic’ or ‘Homeric’ or ‘oral’ state of mind, which constituted the chief obstacle to scientific rationalism, to the use of analysis, to the classification of experience, to its rearrangement in sequence of cause and effect. That is why the poetic state of mind is for Plato the arch-enemy and it is easy to see why he considered this enemy so formidable. He is entering the lists against centuries of habituation in rhythmic memorised experience. He asks of men that instead they should examine this experience and rearrange it, that they should think about what they say, instead of just saying it. And they should separate themselves from it instead of identifying with it; they themselves should become the ‘subject’ who stands apart from the ‘object’ and reconsiders it and analyses it and evaluates it, instead of just imitating it (47).

If we believe Havelock, most of the advancements of the modern world would not have been possible without writing. Writing allowed us to move beyond unreflective, formulaic thought in favor of abstract, logical reasoning. This is not to say that orality does not have its benefits, or even its advantages over literacy in some aspects, but that to stop cultivating literacy and encouraging writing would be a great loss. In order to keep writing, we must remember why we write—not to communicate, but to think in a way that is not possible otherwise.