Who Wrote This?

Yes, I wrote this. But the fact that I now feel compelled to say so is a problem.

This is an essay about provenance, and about the most recent dilemmas regarding human authorship. The condition circles around the medium itself. The very presence of words on paper or screen sits at the heart of its own dysfunction, at least in the sense of understanding communicative authenticity and artistic credibility.

As Marshall McLuhan observed decades ago, the medium is indeed the message. This lands with particular resonance in the conversation about AI, human creativity, technological advancement, and, more specifically, generative AI’s disruption of creative industries. Some things to flesh out:

The barrier of the medium is most evident within its own parameters. Let’s look at this from the inside out to gain perspective. When I hear a musician performing with sincere intention, and this is most vivid when it is live, in person, with sound emanating from the acoustic body of the instrument itself, and with high levels of expertise, it is resoundingly a human creation. No confusion.

Likewise, consider a live theatrical performance: actors breathing human nuance and characterful interpretation from a script. Definitely no AI there. Or a painter working directly onto a canvas. Or a photographer working in a darkroom, developing and printing under the glow of a red lamp and the chemical smell of photo-developing solutions. These art forms are all components of a human condition that runs centuries old. Just on that alone, how could they be anything other than human in their condition?

What’s the thread line through this? Human embodiment. Kinaesthetic experience. Real-time witnessing. Not only do these qualities validate expression and communicate trustworthiness, but they also provide a strong causal link between ourselves and our lived experiences. We are watching something, someone, passing information across time in front of our very eyes and ears, and eliciting a response to what might closely resemble an intimate communal relationship. We are being touched, and if we so choose, we might reach out and literally embrace the source of the art with our own hands. This is the fascination that feels visceral when we approach a work of art. Indeed, this is the mechanism of art galleries and museums and concert halls: the opportunity to witness emotions and experiences from others across time, in the immediately physical present.

So powerful are these functions that the demand to technologically expand the ability to share artful and emotional information beyond its physical timeliness recurs through imaginative innovations and technical developments.

Printing press technology released literature from monastic scribes to broader geographic and demographic audiences across future generations, with profound effect: the Bible, the Qur’an, Don Quixote. A bit later came the ability to record audio. Music that otherwise could only be heard in person was now in people’s homes, and is now ubiquitous to the point of super-saturated availability in every corner of the world.

The same happened with quill and ink, fountain pen to ballpoint pen, typewriter to word processing software. And now, Grammarly to AI chatbots and a whole host of assistive technologies for writers, editors, and the like, from enthusiast to professional. Yet right here, this, this is what I want to explore, disassemble, and analyse with both precision and humanity. Because, apparently, the world of literature and book publishing at large is currently undergoing its own moment of disruption at a fundamental level.

As a writer preparing to publish my first novel, after a lifelong career of working with writers, technical, copy, and creative, and with accomplished editors with agents and book deals and fine arts degrees from leading universities, I am feeling a new kind of pressure and anxiety. That my work will be perceived at least as AI-assisted, and worse, as AI slop. A term being thrown around liberally, and with clear intention and deliberate consequence, to condemn AI technologies to the point of anathema. But who is this serving, and why is it so fervent?

I’ve talked in other essays about the gatekeepers: literary agents, editors, and publishing houses. There’s certainly enough disruptive alarm, with the industry inundated by an overwhelming number of low-quality manuscripts that are bottlenecking resources and the vetting process. The most recent case of this has to do with the modestly popular novel Shy Girl, which moved from self-publishing toward a deal with one of the major book publishers before collapsing under allegations of generative AI involvement. After internal review, the publisher withdrew the book, while the author disputed personal responsibility and described a more complicated production chain.

Whatever the final truth of that case, its cultural function was clear enough: it became a cautionary spectacle for burgeoning writers carving their own paths, and a signal from industry professionals about who controls the chokepoints of literary legitimacy. Which presents an interesting insight, both deeply conservative and also contradictory.

Not to be misunderstood, I am entirely on the side of purely human-authored creativity, be it in literature, music, photography, or visual design at large. I have executed many creative projects across these mediums throughout my career, recognised and ostracised across decades. All the while using tools and technologies to assist in production and refinement. But always, always with authorial control. Anything less than a fundamentally human creative process simply doesn’t make sense to what I consider sincere and transparently vulnerable art.

But I am reminded of a recent talk I attended with the renowned New York Gen X musician and writer Richard Hell. He’s long been revered for his irreverent punk rock sensibilities and erudite poetic prose. Something curious and contradictory revealed itself early in the interview, though.

When asked about his musical career in the ’70s, his highly charismatic creation story, he quickly downplayed his musicianship, saying something to the effect of, “I was always a terrible musician.” This holds true to the punk rock ethic. And the middle-aged artful audience responded with reasonable recognition of this ethos. The interview then went on to dig into his prose: projected clips of passages from his books, followed by conversation and admiration of the nuance and idiosyncrasy of his technique, and his refinement with word structure.

This was my favourite bit, certainly impressive, and what I was there to hear. But I couldn’t help but consider that, if this exercise had been done with any of his, or anyone else’s, punk rock music, it wouldn’t make sense. It would be incongruous, egregious, and directly in opposition to the very nature of what the ethos stands for. Yet not in this case. So, why?

Growing up as a classically trained musician from an early age, I felt this, having suffered the post-punk Toronto Queen West music scene in the ’90s as a working Chapman Stickist. Not “cool,” and arguably discreditable from that perspective. Notions that still linger locally, decades later.

Punk music is allowed to be technically crude when its credibility comes from attitude, embodiment, and risk. These are qualities of innovation being accessed with immediacy and technology: electric guitar barre chords, distorted amplifiers, and visible exertion with raw emotional intention.

But literary prose is judged through an older machine: technique, sentence control, textual intelligence. So Richard Hell can say, “I’m a terrible musician,” and that becomes part of his punk credibility. But if he said, “I’m a terrible writer,” I suspect the literary audience would not treat that as charming in the same way. Written words do not carry the same embodied alibi.

Punk can be deliciously crude because its credibility is visceral and immediate. But it’s borrowed. It’s just rock’n’roll. Although that’s also reductive. I once had the opportunity to participate in a Chippewa powwow, grounded in cultural millennia, and have never felt anything more emotionally raw and “punk” since.

Literary prose cannot make the same claim, because words and literature have long been a mediated premise. I’d argue that book publishing doesn’t democratise literature. In fact, it reveals that authority is a required affirmation. AI is not breaking literary provenance so much as revealing that the evidentiary chain was always fragile. The industry’s mechanisms did not solve that fragility; they merely stood in for it, substituting institutional affirmation for the human verification written text could never fully provide. Not very punk, at all.

While still maintaining my preference, and personal preference will always decide in the end, generative AI, like typewriters in the ’30s, drum machines in the ’80s, digital cameras and Photoshop in the 2000s, will not break human creativity, or literature. It will cause the loss of professional vocations associated with, and dependent on, the raw human channel of creative expression, outside of and before curation, barbarous as it might still seem.

It’s another limited argument, really. It reveals that literary authority has always rested on a fragile substitution: institutions standing in for the absent body of the author. In a culturally relevant and impacting way, this mirrors other tensions between composers and musicians, actors and producers, programmers and end-users.

One might argue: AI tech is all big business, the big man. Again, that is limited. Open source tech, and specifically LLMs and generative AI source code, is also ingeniously driven by grass-roots communities. Look at Hugging Face, or almost anywhere on GitHub, and you’ll see a technology emerging out of the masses of cultural innovation.

AI is giving us the chance to ask: What if literary authority can only survive by mistaking curation for truth? What if the disruption is not coming from outside the gates, but from a fragility that has been inside them all along?

The crisis is not simply that machines can produce language. It’s more that literature has always asked us to trust an absent body. AI has not invented that absence. It has merely made it impossible to ignore.