By Jerry Falade
A few weeks ago, I received an email from Jerry Falade. He asked me to read an essay he wrote, and I obliged. When I offered to run it in this newsletter, we traded edits and finally agreed it was ready to be published.
Jerry is a good guy and a pleasure to edit. Our exchanges have been meaningful, and he has entrusted me with his story. He also sent me the first five chapters of Call Me, I’ll Hide the Body. I have to be honest: I don’t see any “tells” of AI, and I consider myself pretty good at spotting AI-assisted writing. My assessment is that it reads like a script, which makes sense, given Hollywood’s interest in it.
It also makes sense that a non-native English speaker would write in a different style than someone whose first language is English.
I’m asking you to keep an open mind when you read Jerry’s essay. His voice deserves to be heard, especially since his manuscript was passed around and uploaded to Pangram without his consent. No software is perfect, and the people who dig in to the idea that Pangram is rarely, if ever, wrong need to remember that the more people upload and talk about it, the more Pangram receives free publicity, which will inflate its valuation. I vehemently oppose this “gotcha” culture in publishing, and hope that others in the industry think long and hard about accusations of AI-assisted writing, especially when it comes to writers from different cultures.
This is a long essay, so you might need to read it in your browser.
ME AND MY AI AGAINST THE WORLD
by Jerry Falade
It takes months, years even, to write a book. It takes AI-detection software two minutes to ruin your life’s work.
I know because a software report nearly ruined mine. My name is Jerry Falade, the author of Call Me, I’ll Hide the Body. This is one of the hardest essays I have ever written. Long before AI meant Artificial Intelligence, it meant something else to me: AFRICAN INTELLECT. Africa has produced extraordinary novelists, from Chinua Achebe and Ngũgĩ wa Thiong’o to Chimamanda Adichie. For me, that intellect is most intimate when expressed through my language, Yorùbá—a highly lyrical language rich in proverbs, aphorisms, metaphors, parallelism, and “not X, but Y” constructions. Many of these features are now treated as AI artifacts.
I drew on Yorùbá oral storytelling traditions while writing Call Me, I’ll Hide the Body. As a Yorùbá writer telling a voice-driven thriller, it felt natural that the rhythms, sarcasm, humor, and narrative techniques of my culture would shape the novel’s voice. That Yorùbá-inflected voice was one of the qualities every publisher I spoke with praised. It drew fourteen publishers into an auction in the U.S. and led to multi-round auctions involving over a dozen publishers across Germany, the U.K., and Ukraine. I wrote Call Me, I’ll Hide the Body myself. Yet, my African Intellect is now being labeled Artificial Intelligence.
In Yorùbá, even an insult should display imagination. A blunt “You jerk!” or “Go to hell” would disappoint your Yorùbá ancestors. They might turn in their graves if you wasted an insult by making it ordinary. We say things like, Ẹ rí bí orí ẹ ṣe rí pẹlẹbẹ bí ìdí ìgbákọ (“See how flat your head is, like the bottom of a food scoop”).
But place something along the lines of “See how flat your head is, like the bottom of a food scoop” in a manuscript, and an editor trained to fear every unfamiliar metaphor may swear on their life that some hallucinating machine wrote it. What is recognizable as humor within one culture becomes suspicious when read outside it. Speaking of suspicious writing, Jane Friedman, someone I respect in the publishing space, published a piece about AI use on her site, in which she reported that some anonymous agents who had considered my manuscript thought it showed “hallmarks of AI writing” from the very first paragraph. Unfortunately for them, I have a WhatsApp message I sent to myself at exactly 12:15 a.m., while dealing with insomnia, containing that opening paragraph—something I had written as an afterthought. I had already written the manuscript and received full requests before that paragraph even existed.
So my insomnia-induced moment of brilliancy is now being counted by some culturally ignorant people as evidence that a machine wrote my book. You really can’t make this stuff up. It is perfectly fine to encounter writing you do not understand. It is even fine to be ignorant and wrong. But to be confidently ignorant and treat the limits of your own cultural understanding as evidence against somebody else is almost a sin.
Once ordinary rhetorical habits become evidence of AI, almost any writer can be made to look suspicious.
Apparently, metaphors are not the only problem. Our sentence structures are suspicious too. One of the constructions now regularly treated as an AI tell is “not X, but Y.”
This would be fascinating news to generations of Yorùbá speakers who were using contrast for emphasis, correction, humor, and advice long before anyone thought a chatbot might steal their job. I grew up hearing “not X but Y” constructions my whole life. I remember when I complained to my mom in 2017, after completing my BA, that many people my age were already successful while I was still struggling. She told me in Yorùbá: Kì í ṣe ẹni tó sáré jù ló máa kọ́kọ́ dé, bí kò ṣe ẹni tó mọ ọ̀nà (“It is not the fastest runner who arrives first, but the one who knows the way.”) Now, someone in New York can encounter a sentence structure you grew up hearing constantly as a Yorùbá child and confidently inform a publishing commentator that Claude or ChatGPT must have written it for you. They can sound so certain you would think they were sitting beside you when the chatbot supposedly gave you the words.
To put the absurdity in more familiar terms: would anyone expect a Texan to stop saying “y’all,” or a Southerner to retire “bless your heart,” simply because those expressions suddenly became associated with AI?
Exactly.
So why should I suppress the rhythms and rhetorical devices I grew up hearing now that a machine has learned to imitate them?
Let’s talk about my favorite writing technique: aphorism. In Yorùbá storytelling, aphoristic expressions are the salt in the soup. Without them, even a good story tastes unfinished. We have a famous saying, Òwe lẹ́ṣin ọ̀rọ̀, ọ̀rọ̀ lẹ́ṣin òwe; bí ọ̀rọ̀ bá sọnù, òwe la fi ń wá (A proverb is the horse of speech. When words lose their way, a proverb brings them home.) A good Yorùbá orator is expected to be able to use aphoristic expressions to make their storytelling more memorable and beautiful. I learned this at six, listening to my father read D. O. Fagunwa, who painted his voice-driven prose with metaphors and aphoristic expressions.
People whose knowledge of African literature begins and ends with Chimamanda Adichie’s Americanah may encounter Yorùbá-styled aphoristic expressions in my voice-driven fiction and conclude a server farm in Northern Virginia wrote it. The server farm, apparently, has a… broader reading list.
I wrote a seminar paper on the 17th-century Salem Witch Trials several years ago. Now I am living through the 21st-century version.
Oh, is that an em dash? Surely you used AI.
Oh my God! He uses parallelism and aphoristic expressions a lot. AI hallmarks! That Nigerian boy used a chatbot to write the novel.
Welcome to Salem.
And this problem isn’t limited to Yorùbá writers. A friend of mine with Chinese roots read the first draft of this essay and told me how deeply it resonated with her. She is a writer too, and she admitted that she is now afraid to use parallelism and isocolon, both intentional and beautiful features of Mandarin Chinese rhetoric, because they are treated as indicators of AI-generated writing in English.
This is the larger danger behind the current panic. In the rush to identify machine-generated prose, culturally unfamiliar English is too easily treated as “suspicious English.” For those of us who learned English in school, often from teachers who had also learned it as a second language, the way we speak and write it is heavily influenced by our own cultures.
I searched online for the “features” people now associate with AI prose and counted their occurrences in my 85,000-word manuscript. There were 18 ‘not X, but Y’ constructions and 46 em dashes. I am beyond flabbergasted. A career I spent years building was derailed because people unfamiliar with the Yorùbá rhetoric and oral storytelling traditions that shaped my narrative voice concluded, out of sheer cultural ignorance, that a chatbot must have written my work and set the rumors in motion.
Due to these rumors, a journalist obtained my unpublished manuscript and, as they later acknowledged in their published article, submitted it to AI-detection software. All without my knowledge or consent. This reporter sent the AI detection report to my creative partners as proof of my “alleged use of AI.” After a single phone call, my former representatives made their decision and released a statement that could have ended my career.
When this reporter reached out to me prior to the whole incident, I offered to show them my writing process in an interview. I wanted to share my screen and write from a random premise of their choice while they watched because I was so confident in my talent. I mean, writing is the only thing in the world that I know how to do. I also wanted to explain the background and inspiration behind the book. Yet all this person seemed to care about was for me to “confirm the size of my advance.” They ignored my email offering to show them my writing process and reached out again to ask if it was true that my former agent had dropped me.
Beyond what happened to me, the incident raises an even more troubling contradiction. You can’t claim to be investigating the use of AI while taking an author’s unpublished manuscript and feeding it into an AI system without first seeking the author’s permission.
This raises serious questions about intellectual property and professional ethics. I haven’t seen this kind of investigative zeal directed at white authors accused of using AI: hunting down their unpublished manuscripts, running them through detectors, sending the results to their creative partners, and then proudly announcing it in print. Apparently, that level of scrutiny is reserved for the Black boy from Nigeria.
Edward Said would recognize this script. In publishing’s little Orientalist drama, I suppose I am the “Other.” People who probably can’t name eight African countries without Google were allowed to become judge and jury over whether my voice and writing were authentically mine.
And this problem extends far beyond my own case. Even native English speakers have not been spared from the AI paranoia. They can’t comfortably use the natural and deeply human “rule of three.” Don’t get me started on the Oxford comma and curly quotation marks, which I found out two days ago on Threads were now considered AI markers. They also claim writing in formulaic structures is a sign of using AI. BREAKING NEWS: humans are taught formulas for writing too; especially academics and SEO writers. I’m a PhD student specializing in the portrayal of Black masculinity and mental health in African diasporic and African-American literature. I’ve spent over fourteen years studying literature and engaging in critical academic writing, which has trained me to write in highly structured prose: thesis, evidence, transitions, parallelism, conclusion.
If creative people have to second-guess every natural instinct because someone might call it an “AI pattern,” then where is the joy in writing? Why should we fear the em dash as though it were a scarlet letter? Why should we break apart a list of three or four simply because a detector might find it suspicious? Should I stop including Nigerian Pidgin English and my Yoruba oral storytelling tradition in my work because of an AI detector? How long until using full stops and capital letters become AI artifacts too?
And yes, I am aware of the irony: as I write this, I am using the very parallelism, rhetorical repetition, em dashes, and rule of three that someone could feed into a detector as evidence against me and conclude a chatbot wrote this for me.
But at what point does proving we are human require us to stop sounding like ourselves?
Machines were trained on millions of human books to imitate us. Now humans are being blamed for writing like humans? If we do not act now, then within the next few years, every manuscript may begin to sound similar because writers are afraid their work will be labeled machine-generated. A friend of mine even told me he has started deliberately misspelling words and using incorrect punctuation to convince AI detectors that his work is human. We are gradually witnessing an era of linguistic and cultural erosion.
We might even reach a point when we may need to submit AI-detector results for proof of human authorship before an agent agrees to represent a debut writer.
The process will be absurd: I pay an AI detector $20 to confirm that I wrote the book I spent several months, perhaps years, writing. It gives me a 95% AI score. I spend months rewriting the book, killing my voice in the process. Then I pay again, and hope the machine is in a better mood. The future is looking bright for writers.
How does it make sense to use AI tools to detect AI writing? These companies benefit when their systems flag more work, even when the results are wrong. The more fear they create, the more publishers and universities feel they need their services. Yet when a false positive ruins a career, the immediate consequences fall on the accused writer. High-profile cases like mine make the news due to the size of our advance. Lesser-known writers lose agents, submissions, book deals, and sometimes their only real shot at a publishing career. And almost nobody hears about them.
Why should we be forced to upload our unpublished manuscripts into proprietary systems without fully understanding how our work will be stored or used? All to prove our own humanity.
After hearing about my situation, my mentor—a major figure in Southern crime fiction—ran their manuscript written in 2015 through “you-know-which-AI-detector.” The software declared it approximately 70 percent AI-generated. Fortunately, this writer has received numerous awards and has published best-selling books before the whole AI paranoia.
But when you are a debut author with no existing fan base and no editors or publishers familiar with your voice and body of work, your fate may depend on these AI detectors.
The machine made the accusation. The human had to prove his humanity.
So what does proof of human authorship look like when a detector has already been treated as the authority? I have outlines, handwritten notes, midnight WhatsApp messages, notes I typed on my phone, and multiple structurally revised drafts of my work. Yet my career is on trial. You can check my Google Docs’ version history and watch the manuscript evolve and see the same voice, humor, imagination, and judgment at work. When you are forced to disprove a negative, you need to provide evidence, then evidence explaining your evidence.
I used AI for research because I have studied Literature my entire life, and my protagonist is a Chemistry student with expertise on different types of acids. I couldn’t invent fallacies which would get called out by critics when the book gets published.
So, I used a locally hosted AI for questions like:
● How do you fit a corpse inside a cooler?
● How to delay a corpse’s decomposition process
● Which chemicals can hide blood stains under UV light?
I am sure if you run into a PhD student whose Google search history included how to fit a corpse inside a cooler, you would probably call the cops on him. This is what I used Dolphin-Phi for: to ask questions no crime writer wants in their Google history, and to convert the metrics (celsius, kilogram, and others) we used in Nigeria into their American equivalent.
If you’re a writer who constantly feels the urge to clear clutter and delete old files, please stop. With the current AI paranoia, you need to keep everything.
And for your ultimate alibi, convince IShowSpeed to livestream your entire writing process. Don’t take bathroom breaks. Unless you have video evidence of yourself typing every word of your manuscript, defending yourself against commercial AI detectors may be nearly impossible.
It is my African Intellect and me against the world. Like Femi, the protagonist of Call Me, I’ll Hide the Body, I prefer to fight my battles one at a time. I will not let people with louder voices and deeper pockets bully me into silence.
A famous Yorùbá adage says that a ram that steps backward during a fight is not a coward. It has only stepped back to gather more strength. I have taken that step. Now I’m moving forward to reclaim what is mine.
I am delighted to announce that David Vigliano is my new literary agent. He has repped people like former U.S. Vice President Mike Pence, Britney Spears, Mike Tyson, Prince, and Justin Timberlake.
Oh, and even Pope John Paul II.
I have also been invited to appear as a guest speaker and panelist at the Aké Arts and Book Festival, the largest African literary festival. This year’s theme is “Native Intelligence,” and I will speak extensively about how African Intellect, and the intellect of anyone whose first language is not English, is being unfairly flagged as AI-generated. I have spent enough time defending my voice. Now I intend to use it.
As a Nigerian man, I only fear God and my mother’s slippers.