Hello readers!
Back in January I went down the rabbit hole of ChatGPT and Canva MagicWrite, both in the existential sense and implications for my classroom. And it’s not at all that I forgot about it since then, but the firehouse of news coverage and developments in the past few months have been staggering / dizzying. Birgitte Rasine and I connected back in January and when she offered to share her expert insight on the topic, I immediately took her up on it.
I hope you enjoy Birgitte’s thoughts on the debate about generative AI in the fiction landscape!
CEO One: What does [generative AI] mean for the market for fiction writing?
CEO Two: It's going to be crazy. And much harder to break through--and much more important for people who write fiction to be really smart about prompt engineering.
CEO One: I’m reading Babel right now, it could be catagorized [sic] as a type of historical fiction because of its setting, and it also pushes a strong theme of “language as empire”.
To that end—if language is the ultimate tool of empire, and chatgpt [sic] is rising, I infer that tech might now, officially, be making itself emperor. Its [sic] like a coronation of silicon valley.
All the while, SVB…
Man, what a time to be alive !
The above conversation took place on LinkedIn on March 15, 2023 between the head of an environmental testing services firm (CEO One) and the head of a major national publication (CEO Two). It was one of several comment threads (over 100 comments in all) posted in response to CEO Two’s original post: a delighted recounting of how he put together a little book with one of his sons, as a birthday present for the boy’s younger brother—and they used ChatGPT to write a part of the book. It was just one of the many, numerous, multiple posts and articles that have been flying off the keyboards of altogether over-excited humans, about the brave new world of generative AI that we have fallen into.
Except that the CEOs of large national media outlets hold considerably more sway over cultural conversations than, say, an intern at a fish hatchery.
The irony of this LinkedIn post coming from the head of a publication I not only respect but read often, and which has published numerous thoughtful pieces about the impact of generative AI software like ChatGPT, dripped off my screen all over my desk, and onto the carpet.
The comments that followed dripped still more viscous: one person was trying to figure out “the multiplier” for the volume of fiction titles ChatGPT would write, compared to what fiction writers had been able to do up until this point. Another pointed out the glaring reality that no human writer could compete with ChatGPT’s basic fee, which is $0. A healthy number of commenters were deferential to CEO Two, thanking him for keeping them current and inspired, but their between-the-lines reactions were decidedly cautious.
In the original video post, CEO Two concluded that because he had asked ChatGPT to write a portion of the birthday book, there were really three authors: him, his son, and the AI. But, he said, “you can’t really tell the difference.” In the future, he added, “no one’s gonna really know who wrote anything.” [emphasis his]
Think about that for a moment.
“No one’s gonna really know who wrote anything.”
So, Mr. CEO, just like that, you’re willing to flush 5,423 years of our collective human creativity, passion, skill, and effort right down the fiberoptic plumbing system, straight into the gaping maws of a Large Language Model? Perhaps it’s a veiled warning, a public service announcement you’ve quietly wrapped into a selfie video. Or are you just click-baiting us?
Everything Everywhere AI at Once
It wouldn’t be the first time. AI click-bait is everywhere. Everyone seems to be talking about ChatGPT, Google’s Bard—and to a lesser degree the other generative AI tools. This is the cultural ecosystem we in the “modern” world live in now. And dare I say, we’ve been boiled long enough now that hopefully we can finally feel how hot the pot has gotten. Fake news, move over, AI’s the new sheriff in town!
I have spent many hours reading about this new form of AI. Taking it in. Listening. Thinking. I was supposed to be writing this article when I happened across the above exchange between the two CEOs. I spent an hour wading through the seagrass, as it were. Every single comment. Processing the worlds all of these people were coming from, that I had only a sliver of insight into. Then, it exploded.
I scrolled to the very top of the comment pile, and in less than a ChatGPT minute, I let it all pour forth:
To everyone on this thread discussing us fiction authors in terms of dollars and data points: may I remind you we are human. We are not soluble numbers and we are not market share. We are creators, thinkers, travelers through the human spirit and experience. We do not write simply for cash, renown or titles. We write because we feel. We write because stories course through our blood. We write because narratives are intertwined in our DNA.
When you read a story written by a human author, you've been given the blessing (or curse, as the case may be!) of entering their world, their dreams, their journey. If you have lost that connection, I am truly sorry.
If you do not see that this is the most profound form of extractive colonialism humanity has ever seen, you have lost your sight -- and your insight into what being human means.
I know other creators, especially artists who are experiencing the same thing with the generative art AIs, feel the same.
Do not speak for us. Do not gaslight us. The day writers stop telling stories, the artists stop painting, and the musicians stop composing, is the day humanity dies.
No outlines, no edits, just raw emotion.
But let’s back up and be clear-eyed about this. I don’t hate AI. I’m not afraid of AI. Heck, I even worked in AI.1 I don’t blame AI for any of this. It’s all on us humans. Because:
Humans conceived of AI.
Humans wrote the algorithms.
Humans achieved phenomenal things with some of these algorithms (protein-folding party, anyone?)
Humans “trained” generative AI models (e.g., large language models or LLMs) on the collective body of work of other humans (writers artists designers and other creative types). They didn’t tell us what they were doing, and they certainly didn’t ask our permission.
Humans released ChatGPT (and other gen AI) into the public sphere aware of their potential misuse and without the proper guardrails.
The other humans took to it like moths to a flame.
One more thing ~ let’s zoom in a bit on CEO Two’s comment that fiction writers will need to be “really smart about prompt engineering.” I do want to be clear I’m referring specifically to literary fiction authors here… not marketing content creators, not copywriters, or any other type of business or marketing writers (that’s another post for another time, and, spoiler alert, generative AI can be a good thing).
An engineer is not a writer, and a writer is not an engineer. That is not to say an engineer couldn’t also write books, or that a writer couldn’t also do engineering stuff. What I’m saying is that writing is not engineering. Words are not data points and sentences are not predictive models. Storytelling isn’t an algorithm.
What the CEO is saying, however, is that technological progress has come for us literary types. Convert or die!
To prompt or not to prompt? That is not the question.
The idea that I should give up my skill, my passion, and my agency as a writer, as an author, to a piece of software that probably got fed on some of my work without my permission, and now wants to regurgitate it in my voice, tone, structure, and style to anyone who plugs in my name, is appalling. This is a profoundly dehumanizing and disempowering form of intellectual colonialism. The speed with which so many people have jumped aboard this runaway train without checking the final destination printed on their tickets, is shocking. Understandable from the perspective of human behavior, but still wild to witness.
Storytelling is being human. It is not about data. It is not about money. The fact that many of us make our living from it (money) and track its success or readership (data), doesn’t make it innately about those things. The innate meaning and value of stories is human connection. This is how our culture evolved. This is how we have shared what it means to be human, from one generation to the next, from one century to the next, from one culture and country to the other. This is what storytelling is about.
And that is what we are giving up when we let algorithms write for us.
You might say, oh don’t be such a Luddite. Technology is evolving, it’s useless to be afraid of progress! But this has nothing to do with being a Luddite or resisting “progress.” I’m not advocating for going back to the typewriter and making books by hand. There is zero question that being able to write this essay on a computer instead of having to write it out by hand and then mail it to a publisher who would manually lay it all out, is exponentially faster and easier, and that that technology, that progress, is the source of this explosion of expressed thought and all this content we have seen in the last few decades. For better, or for worse :)
But you cannot compare a word processing program to generative text algorithms. It’s a completely different level of progress. MS Word, Google docs or any of the other word processing programs do not write for you. They do not take your agency as a writer away. They do not appropriate the work of countless known and unknown writers and authors, decompose it into words and patterns, and disgorge it back up on your screen in the form of newly formed phrases and sentences that are based on their predictive algorithms. Yes, predictive. Not original, not creative, just predictive.
Personally I don’t mind—at all—if a gen AI writes some of the more administrative or iterative content for me. I would much rather be working on my novel (which I’m not doing right now because I’m writing about an AI that I don’t want to write my novels for me).
But we do need to be very nuanced and cautious about how we use these generative text algorithms. Some may counter that writing prompts does take effort and insight, and that there is still the work of editing the final product. Yes, but it is still not original writing. In fact, you are no longer the author. You are a fact checker, a researcher, an editor, a bot babysitter—and likely the credited “author” if you put your name on it. But to be truly a writer, you need to do the work. A singer only deserves that title if she sings—not if she creates a synthetic voice which she modulates. An athlete only deserves that title if he puts in the hours—and hours and hours—it takes to train his body and mind.
Remember, too, that when you use the bots to write, they learn from every prompt and every piece of text you run through it. You are effectively empowering the algorithm to take agency away not just from you, but from other writers.
Well, how about just using ChatGPT to generate ideas? To flesh out my narrative? To just, you know, get me going when I can’t think of a way to open a chapter?
I’m not your editorial mama and I can’t control what you do. What I can say is, to all of you writers and poets out there, don’t lose that powerful creative spark. Sure, see what fresh new ideas the AI can spin up for you, but don’t give up your agency. Guard your unique voice, your originality, your creative expression and style. Above all, don’t let go of your ability to imagine, to wander, and to wonder. Because that—that is what makes us human, and that is what makes us storytellers.
Birgitte Rasine is an award-winning literary author (The Jaguar and the Cacao Tree, Verse in Arabic, and other works), former journalist for the Hollywood trades, and the author of The Muse here on Substack.
For full transparency, I (Birgitte) worked in AI, for Google. I was creative writer #4 hired on the early Google Assistant team, and helped launch the Assistant in Latin America, France, and French Canada.
Thank you so much for sharing such a considered and thoughtful discussion about this, I feel much the same as you and it makes me sad (but doesn’t surprise me) to hear that people high up see things that way.
Oh the CEOs....sigh.