You sit down to write, the blank page in front of you simultaneously inspiring and intimidating: The channel is open, the possibilities are limitless. This stage of the process is difficult and anxiety-inducing, but you know it is an unavoidable part of writing.
//Or, maybe, it doesn’t have to be.//
The impact of computer systems acquiring artificial intelligence (AI) is hard to understate. Although conceptual and technological groundwork for AI had been established since the 1950s, progress and usage were largely sequestered in university and government research labs. It wasn’t until the public release of ChatGPT in November 2022 that artificial intelligence began to dominate the popular consciousness. As companies and industries scramble to adapt to a new world, the divide between optimism and pessimism surrounding artificial intelligence deepens. AI forces us to confront questions looming over modern existence: What do we want from machines, and what will we sacrifice for it?
Understanding AI and writing
Although these may seem like unprecedented times, contemporary discourse about AI echoes common concerns from earlier waves of technological development. Historically, new technologies and mediums have stirred fear among artists concerned about their effects on existing forms. 19th-century realist painters cursed the advent of photography, 20th-century filmmakers worried about advancements in computer graphics, and 21st-century writers considered how word processing software changes the way authors write. Through artistic movements and technological revolutions, writing as an art form has adapted and endured.
Recently, the rise of Amazon and social media has tested the literary field, significantly altering the relationship between publishers, authors, and readers, and pushing the novel into the modern economy of attention. Mark McGurl, in his book //Everything and Less//, explains how Amazon and Kindle Direct Publishing (KDP) reorient priority from writing innovation to customer satisfaction, contributing to the commodification of literature. Authors become producers, forced to adapt their image and work to the fast-paced marketplace. Readers turn into superficial consumers, captured by promises of entertainment or enlightenment; thus, the marketplace works hard to retain their attention. This attention-driven model restructures the reading experience, orienting it around convenience and constant engagement. To maintain and expand its customer base, Amazon aims to make reading frictionless, promoting an abundance of genre-fiction that caters to users’ niche interests. To make matters worse, Amazon uses machine learning and data mining—subsets of artificial intelligence—to extract data from users’ reading habits for marketing purposes, perpetuating a never-ending cycle of consumption.
The emergence of fiction-writing AI tools should come as no surprise, considering the commodified logic of modern literary production. One such program, widely regarded as the cream of the AI-writing-tool crop, is called Sudowrite. “Blank page, begone!” touts the website’s homepage, promising it can brainstorm, create story outlines, expand descriptions, generate metaphors, suggest character arcs and plot twists, and edit users’ writing. Since the post-Amazon literary landscape has positioned books and authors more as products and producers, the implementation of tools to maximize efficiency is an expected progression. While KDP opens paths to publication for more authors, the digital marketplace is an unstable one. For Amazon authors like Jennifer Lepp, who makes a living churning out self-published “potato chip books” multiple times a year under the pen name Leanne Leeds, using Sudowrite to accelerate the writing process allows her to keep up with the demands of the marketplace and her audience.
Despite still using the program, in an interview with The Verge, Lepp considered the harmful impact of AI writing tools on her own skills and their broader homogenization of literature.
“I need to pay attention much less closely. I don’t get as deeply into the writing as I did before,” she said. “I think that’s the real danger, that you can do that and then nothing’s original anymore. Everything’s just a copy of something else. The problem is that’s what readers like.”
To make a living off one’s art, artists are tied up in the dynamics of consumer demand. How, then, do we reckon with cultural shifts surrounding literary production, and what responsibilities do readers and writers have to each other?
The promises and the problems
In literature, artificial intelligence makes two enticing offers. First, it claims to eliminate the kinds of reading and writing it casts as mere busywork—proofreading, editing, synthesizing, and summarizing—and free up time and energy for authentic artistic and intellectual exploration. Second, it promises to enhance this exploration, acting as an excitable book club buddy or a thoughtful writing partner.
In an interview with //The Tribune//, Chris Howard, associate professor in McGill’s Department of Philosophy, expressed a cautious optimism about the incorporation of AI tools in creative fields.
“There is still a human element in terms of telling the machine what you want it to produce, and there’s a lot of latitude for creativity in that process [….] You could have an absolute explosion of culture and a diffusion of power from the gatekeepers of culture. Which is really exciting, but I also don’t want to be too rosy [….] I understand the attitudes of artists who feel like there’s some danger to their craft with this,” Howard said.
Using two dozen large language models (LLMs), Sudowrite functions almost like an advanced autocomplete, using data analysis and statistical probability to create sentences by guessing, one word at a time, what comes next. Despite being transparent about how the word processing system works on their homepage, in their FAQ, they answer the question “Is this magic?” with “Yes. But so is life, isn’t it?”
To some extent, this is true: AI is still somewhat of a black box. It is hard to dissect exactly how deep learning and neural networks function, contributing to a fervour and mystique surrounding AI that strikingly mirrors religion. Tech companies’ insistence on the inevitable takeover of artificial intelligence bears resemblance to the eschatological story of the Rapture: You must get with the program, or get left behind.
In an interview with //The Tribune//, Alexander Manshel, associate professor in the Department of English at McGill, rejected this narrative and advised against passive acceptance.
“As with any technology, we’re often told that it’s inevitable, so it’s not if it’s going to change everything, but how it’s going to change,” he said. “But, as with any technology, we have a choice as to whether or not we adopt it, and if so, how. I’ve been really heartened by talking to a lot of not just McGill faculty, but McGill students, who see the value in doing their own complex thinking, their own complex writing.”
To truly engage in complex thinking and writing, no part of the process can be dismissed as nonessential. By trying to eliminate or accelerate some parts of the process, AI writing tools contribute to the sacralization and romanticization of writing, detaching the work from the art. The actual process of writing is envisioned as an undesirable, but unfortunately necessary step that artists painstakingly overcome to translate their grand visions and ideas into the world. However, AI cannot replicate the craft of artistic creation or the process of critical exploration, which is where art and ideas get their value.
In a piece for The New Yorker on Sudowrite, Canadian novelist Stephen Marche wrote, “For writers who don’t like writing—which, in my experience, is nearly all of us—Sudowrite may well be a salvation.” Beyond the use of “salvation” revealing a quasi-religious zeal for AI, the sentence also points to a broader tension: The sense of distance from art that can emerge when generative technologies partly or entirely mediate creative work. In the creative process, ideas gain meaning by acquiring shape and structure through negotiating with the strengths and limitations of a certain artistic form. Marche’s framing places primary value on ideas in their abstract, inarticulated state, rather than on the interpretive labour required to realize them. To many, this perspective can appear liberating—writers might not need to be skilled in the craft itself if a system can refine their concepts into readable prose. But what Marche and the broader framework of AI writing tools overlook is that writing is not a means to an end. The process itself is both the labour and the art.
The way forward
Individual writers can often feel insignificant against the larger political, economic, and cultural forces influencing the trajectory of literature. But the issues and questions raised by AI writing tools go beyond the specific concerns of the literary field. All of us who engage in reading, writing, and thinking have a critical decision to make.
In an interview with //The Tribune//, Alex Steele, president of Gotham Writers Workshop, a creative writing school in New York City, discussed what is lost in AI writing.
“AI does not think, it does not feel. It is, quite literally, bloodless,” he said. In his September newsletter, titled “In Praise of People,” he writes, “You, on the other hand, are the proud owner of blood, brain, body, and (most mysteriously) soul. The power of your writing—however imperfect, flawed, messy—will be found there and nowhere else.”
It’s a vital reminder of the purpose of art: To challenge perspectives, deepen understanding, and foster emotional connection through the transmission of human experiences. In writing, we are constantly confronted with the limitations and inadequacies of human language. Amorphous thoughts, feelings, and ideas beam into our brains, and we translate them into the world as we shape them through syntactic structures, diction, form, and style. Through practice and repetition, this process becomes mechanized in our brains. Perhaps this is why we overlook what we sacrifice when we let it become actually automated by machines.
Agnes de Mille recounts a conversation with fellow choreographer Martha Graham in her biography that echoes Steele’s belief in the value of unique human expression.
“There is a vitality, a life force, an energy, a quickening that is translated through you [….] And if you block it, it will never exist through any other medium and it will be lost. The world will not have it. It is not your business to determine how good it is nor how valuable nor how it compares with other expressions. It is your business to keep it yours clearly and directly, to keep the channel open,” Graham said.
AI is, at the end of the day, a technology and a tool. Particularly in the modern labour landscape, its use may soon be unavoidable. But in our personal, intellectual, and artistic lives, we have the power to be intentional about our implementation of AI, choosing how, when, and whether to use it. In our creative communities, we can influence the creation of guidelines, norms, and precedents regarding artificial intelligence. Above all, we must remember what may be at risk, and keep the channel open.
