1. In the dream, I go to a house and walk into its dimly lighted living room. Two poems are projected on the wall. A stranger tells me, “Your poems are incredible.”

2. How is this possible? I’m not a poet.

3. “We think you should make them longer,” the stranger says. In the dream this annoys me. Doesn’t he know that saying more isn’t what I’m into; saying more with less is.

4. The dream ends. I’m still in the groggy twilight, lying in bed next to Miss Nancy. It’s a quiet early morning in the attic of a garage on an old farm in Rhode Island. We are in town for July 4th to visit family.

5. I consider getting up to write. Because of this damn dream. Because sentences are flooding into my mind. The dream has transmuted into a question: Why write?

6. Don’t do it, I think. You’ll be on a sleep deficit all day. You’ll be short-tempered and rude. You won’t enjoy the day as much with Miss Nancy, combing the beaches of her childhood.

7. Yeah, sure. I quietly leave my bed and step lightly into the next room to tap out notes on my phone about why I write.

8. Then back to bed. Then back out of bed. Good lord, I think. Something is really wrong with me.

9. Why write?

10. Tap-tapping in the living room: this urge, this compulsion to capture fleeting thoughts. I reach for an analogy: butterfly collecting.

11. As a child, seeing a butterfly made me want to catch it. Not to trap it or kill it, but to lure it onto my finger. To pause, appreciate, to hold the moment, because I knew it would not last. Flap flap flap, silently. Then off it goes!

12. As an adult, I became a butterfly collector. Striving to capture fleeting thoughts and words, not just appreciate them and let them go.

13. Collecting is about catching the butterfly in a net, dropping it in a kill jar, and pinning its beautiful corpse on a board for display with a little label. Monarch Butterfly. Danaus plexippus. Family: Nymphalidae. Linnaeus, 1758.

14. Dreams, story ideas, dialog, descriptive phrases, butterflies. Fleeting and precious. Writing kills the moment to preserve it.

15. To write is to make words sit still so other people can appreciate them, the way collectors catch butterflies and pin them on boards, or the way mineral hounds split geodes open with the crack of a rock hammer to reveal beautiful crystals and colors. “Isn’t this beautiful?” you say to your friends. And they say, “That’s so amazing. So beautiful.”

16. If the words, dialog, description were allowed to dissipate without being pinned to a board, would they be any less exciting, less precious, less appreciated?

17. While tapping out my butterflies in the living room, I think of the paintings at Lascaux Cave in France. Hundreds of paintings of animals dabbed in pigment on stone walls 20,000-ish years ago.

18. Did a cave dweller point out the prancing ponies and foraging feral cattle roaming the land to a friend, and say, “Uggg! Come, see! So beautiful. Is perfect moment. Must preserve so future generations of hominids can appreciate.”

19. Why write?

20. This compulsion to record and display fleeting moments—which I now call The Affliction—seized me when I was 10 years old.

21. As a teen, writing became an identity. I didn’t just write stories. I was a Writer. I wrote like crazy. Bad short stories tumbled out of me.

22. Eventually, writing became my living. I sold my first story. Then another. One day wandering around an airport on a reporting trip, I saw a cover story on the magazine stand with my byline for the first time. Wow that’s me there.

23. The thrill wore out. Being a full-time writer and editor became like carrying water back to your village from a distant well in the heat of North Africa. Pay rent… put gas in car… pay for divorce… buy new butterfly net.

24. Why write?

25. I don’t have to write anymore. Now I just write short stories that not many people want or need. I’m back where I started, at 10 years old.

26. I do it without expectation. If someone reads something of mine and likes it, I am grateful. If they don’t, that’s also fine. I’m doing it for the doing of it.

27. It’s also liberating to know it doesn’t matter whether I do it or not. A little nihilism goes a long way in life. Does it really matter if I write?

28. I once heard a Buddhist teacher open a meditation retreat this way: “You are all going to die, probably sooner than you think. So, now what do we do?”

29. I know what I do: I write. But to do it with minimal despair, I resist the delusion that I am a rare specimen that some collector (agent, publisher) will some day come across and put on their shelf for display.

30. I have a cure for the delusion that “my work” will aggregate into a “legacy” that will cast a long shadow on the world after I am gone: a writer’s death meditation. I don’t find it depressing; I find it liberating.

31. Here it is: Everything I’ve ever thought, felt, said, or written will be lost and forgotten. In the decades after I die—despite the thousands of bylines, the cover stories, the embarrassingly well-paid journalism fellowship, the editors who loved me, the colleagues who respected or despised me—virtually no one will even be able to recall my name.

32. Except a niece or nephew who, approaching retirement, may reminisce at Christmas, “Remember Uncle Danny? Wasn’t he weird? Always writing things down. Always chasing butterflies.”

Many thanks to my Substack pal for reading a bloated and unfocused early draft of this essay and insisting that it might actually matter that I write and that my teenage short stories probably didn’t all suck.

Leave a comment

Share

Subscribe now

Prompt: This writer tries to make a case for AI-assisted writing. Other people disagree that AI-assisted writing is writing. They say it is prompting. And they have other arguments. Summarize and evaluate the arguments. Suggest thoughtful questions for readers to consider.

Chad Rye

Damned If You Do: What Happens When You Give the Anti-AI Crowd Exactly What They Asked For

Ya know, there is a particular kind of trap that only reveals itself after you’ve already stepped into it…

Read more

5 months ago · 102 likes · 83 comments · Chad Rye

[Claude’s response, verbatim.]

The Author’s Core Arguments

Rye’s central claim is that he did exactly what AI skeptics demanded — full transparency about his process — and was rejected anyway, proving that the real demand was never transparency but capitulation. He describes a process where he writes extensive handwritten synopses covering character motivation, dialogue, pacing, and theme, then collaborates with Claude to produce prose, which he revises and directs. He estimates he contributes 75–85% of the creative work.

From there, he builds several supporting arguments:

The moving goalposts argument. Each time he answered a criticism, a new one replaced it — from plagiarism, to “handwriting doesn’t count,” to “you’re a technician not a writer,” to environmental destruction. He frames this as evidence that his critics aren’t arguing in good faith but enforcing a predetermined conclusion.

The historical precedent argument. He points to dictation, ghostwriting, heavy-handed editing, and Henry Darger’s collage techniques as evidence that writers and artists have always used tools and collaborators without losing creative legitimacy.

The learning-through-collaboration argument. Rather than stunting his growth, he says working with AI has accelerated it — you can’t effectively direct and revise output you don’t understand.

The class and gatekeeping argument. Drawing on Quinn Que’s commentary, he argues that established figures like Darren Aronofsky can adopt AI with mild scrutiny, while a self-taught writer from rural Florida triggers moral panic. The objection, he suggests, is less about the tool than about who’s using it.

The environmental whataboutism argument. He pushes back on the “Climate Destroyer” label with comparative data — AI’s share of global emissions is a fraction of what coal, fashion, agriculture, and golf courses consume — and calls selective outrage a “cultural argument wearing a green coat.”

The Critics’ Arguments (As Represented and Inferred)

The piece is written entirely from Rye’s perspective, so the opposing positions have to be partly reconstructed. But the core objections seem to be:

The authorship question. If an AI generates the prose — the actual sentences, word choices, rhythms, and images — then the person who wrote the synopsis is doing something closer to directing or producing than writing. A film producer who describes what they want in each scene isn’t the screenwriter, even if their vision drives the project. The question isn’t whether the synopsis involves creative work. It’s whether creative direction is the same thing as writing.

The craft argument. Writing isn’t just knowing what should happen in a scene. It’s the thousands of micro-decisions that happen at the sentence level — which word, which rhythm, which image, what to leave out. If the AI is making most of those decisions, even if Rye is selecting among and revising them, the skill being exercised is editorial judgment, not composition. These are related but distinct abilities.

The displacement concern. If AI-assisted books become normalized and can be produced faster and cheaper, writers who do the sentence-level work themselves face market pressure from people who don’t. This isn’t about one person; it’s about what happens to a craft and a livelihood at scale.

The “slop” concern. Even if Rye is careful and deliberate, normalizing AI-assisted authorship opens the door for the flood of low-effort, AI-generated content that’s already overwhelming platforms like Amazon and Royal Road. One thoughtful practitioner doesn’t address the systemic problem.

Evaluation

Rye makes some points that hold up well. The moving-goalposts observation rings true — online discourse often does shift criteria the moment someone meets the stated standard, revealing that the standard was performative. His transparency is genuine and worth acknowledging regardless of where you land on the larger question. And the class dimension is real; credentialed creators do get more latitude with new tools.

But the piece also has significant weaknesses.

The environmental section is the clearest example. Rye spends roughly a quarter of the essay on comparative emissions data, but the logic is pure whataboutism — the fact that golf courses use more water doesn’t make AI’s water consumption unimportant. He seems to recognize this (”I am not arguing that AI has zero environmental footprint”) but then spends several hundred words doing exactly what he says he isn’t doing: minimizing AI’s footprint by pointing at bigger ones. The section reads like a tangent that dilutes his stronger arguments about authorship and craft.

More fundamentally, the piece never seriously engages with the strongest version of the opposing case. The real question isn’t whether Rye works hard or cares about his stories — he clearly does both. The question is whether the thing he’s doing is the same thing as writing prose, or whether it’s a genuinely different activity that deserves its own name and its own category. He compares himself to authors who use dictation or heavy editing, but those analogies have limits: a person dictating a novel is still generating every sentence themselves, and an editor reshapes existing prose rather than generating it from a synopsis. The AI collaboration he describes is doing something qualitatively different from either of those precedents, and the essay would be stronger if it wrestled with that difference instead of treating all tool-use as equivalent.

The Pharisee metaphor is vivid but also does some heavy rhetorical lifting — it positions his critics as hypocrites motivated by self-righteousness rather than people with a legitimate philosophical disagreement about what authorship means. It’s effective as persuasion but less useful as argument.

The outline-writer anecdote is actually the most interesting moment in the piece, and it’s underexplored. That person’s question — “did AI write that book or did I?” — gets at the real heart of the debate. Rye answers quickly (”I think he wrote it”) and moves on, but there’s a rich and genuinely unresolved question there about where the line sits between scaffolding and structure, between directing and creating.

Questions Worth Sitting With

On authorship: If you write a detailed synopsis and an AI writes the prose, and you then revise the prose — at what point does the balance tip from “I wrote this with AI assistance” to “AI wrote this with my direction”? Is there a meaningful line, or is it a spectrum?

On craft: Is learning to direct and revise AI output a form of learning to write, or is it learning a different skill? Can both be valuable without being the same thing?

On precedent: Rye compares AI collaboration to dictation and editing. But when you dictate, every word originates in your mind. When you edit, you reshape existing human prose. AI-generated prose originates from neither of these processes. Does that distinction matter, or is the creative vision upstream of the prose what ultimately counts?

On transparency: If Rye’s transparency was met with the same hostility as silence would have been, does that prove the demand for transparency was dishonest — or does it just mean transparency doesn’t resolve the underlying disagreement?

On scale: Even if one careful, deliberate AI-assisted author is doing something defensible, does normalizing the practice create conditions that harm writers broadly? Can you evaluate an individual case without considering its systemic implications?

On gatekeeping: Is the insistence that “writing means producing the prose yourself” a defensible standard, or is it an arbitrary boundary drawn to protect a particular mode of production? How do we distinguish legitimate craft standards from credentialist gatekeeping?

On the outline writer’s question: If someone uses AI to generate a draft and then rewrites every word, most people would say they wrote the book. If someone uses AI to generate a draft and changes nothing, most people would say they didn’t. Somewhere between those poles is a line. Where is it, and who gets to draw it?

Subscribe now

Share

I am 23, sitting at the dining room table, typing the final draft of an article about marine life on coastal jetties on the north shore of Long Island.

My older brother—one of the cool kids, one of the “heads” at our high school, one of the teens who went to Led Zeppelin concerts and smoked weed and got laid, and was just beginning his drug career—walks up behind me, looks over my shoulder, sneers, “Feathered dive bombers! Faggot!” and smacks me in the back of the head.

Don’t ask me about the logical connection between homosexuality and using figurative language to describe seabirds. I have no idea.

I hadn’t learned yet that “dive bomber” is not the best way to evoke the image of a seagull, but the article was accepted and published—my first-ever professional magazine feature sale ($75). No college degree, no training, no advice from my parents other than, “Just make sure you learn a trade so you can always pay your bills.”

Writing wasn’t what they had in mind. More like carpentry. I ended up as a carpenter of words instead.

As I stumbled into a very unlikely and largely unplanned career as a writer and editor, my brother descended further into addiction, crime, and violence, and spent much of his adult life behind bars. He died alone in a Brooklyn hospital from a head injury.

That’ll teach him to make fun of my similes.

Subscribe now

Share

When Trevor Cohen kindly invited me to join his collaboration and write about my daily life as a writer, I thought, “Whoa, that would be very boring. I sit motionless in front of a computer and write, occasionally pausing to pick my nose or check to see if any of you are paying attention to me on Substack.” Oh, you too?

But I did find myself inspired by the challenge of choosing an adjective to describe myself, “Day of the ____ Writer.” I chose disciplined.

Disciplined? Whaddya sayin’ Junk Man? That the rest of us aren’t? Nah. It is just the best word I could think of. I know it sounds a little self-congratulatory.

One aspect of my writerly discipline—ahem, “an organized practice with its own methods and standards”—is that I read great writing continually, every day, no matter what, like a starving man. I read magazines, newspapers, novels, and short stories. In print and on Substack. Thanks to a digital Libby account with the DC Public Library, I can check out loads of things I can’t afford to buy.

But I read the way a writer reads: to notice and understand what the author is doing, so I can learn from it and apply it to my own work. I steal the occasional adjective.

Reading like a writer is part of my “creative habit,” which Twyla Tharp wrote about in her insightful little book of the same name. She said that creativity isn’t a mystical and ineffable inborn talent but a practical, daily habit you can cultivate. I agree!

I am certain that much of my creative habit comes from my former life as a freelancer. I wrote for science magazines full time. And more! So much more than full time! I needed to publish 60,000 to 80,000 words a year to make a living. This went on unceasingly for decades.

Subscribe now

What does that require?

You write when you want to and when you don’t. When you are tired and when you are rested. When you can’t think of anything interesting to say. When you love the story or it bores you to death.

Parent dies—shut up and write.

Marriage detonates like a Hiroshima bomb—shut up and write.

Drink yourself into rehab—get sober, go to AA meetings and shut up, shut up and write.

Then rewrite every first-draft catastrophe until it looks publishable. Get edited. Rewrite more and keep answering editorial queries until they run out of them.

Sound like fun? It requires a certain madness, and also fear. Lots and lots of fear.

When I walked away in 2024, I was done. No joy left in it. I didn’t have a single em-dash clause left to give.

Now that I’m back where I started as a teenager—writing short fiction—I can rely on a well-trained, biological Large Language Model in my head.

After the messy first draft, usually purged in one or two vomitous bursts, I keep drafting until I run out of ways to write it. A draft a day until I am done. It’s not unusual for me to write fifteen drafts.

That’s what it took last month to get this story to the point where I thought it worked, to where I couldn’t think of any other way to write it.

Then I publish it on Substack. For me, hitting the Enter key is like typing the last period at the end of the last sentence: fini. Most of the pleasure is in “having written,” as Dorothy Parker famously said.

Just a final word about gratitude. I’m grateful simply to be free to scratch out my half-assed short stories and publish them here, whether anyone reads the stories or not. Mostly they don’t, and I’m okay with that.

My job every day is to enjoy this. To connect with other writers I can commiserate and bullshit with. To feel the thrill of small breakthroughs. To experience the satisfaction and pride of becoming incrementally more competent at a very difficult thing. And to be intoxicated again by being a writer, the way I felt in the early years when I made my first sales in national publications and saw my byline in the magazine rack at the airport.

Thanks for listening. I hope to have my first (and last) novel done in time to file for Social Security.

Share

Leave a comment

I knew I had to tear it down and rebuild it. After revising, it somehow ballooned to over 4,000 words. The story had to be 2,500 words for submission.

I cut it to 2,400. I thought, Wow, this is so much better. But despite the better angel on my shoulder begging me to stop, I added back 145 in editing. I turned it in.

With a paradoxical postpartum enthusiasm, I felt like writing more about one of my characters. I fished out a 10-year-old attempt at flash fiction. It came in at 525 words. It was tighter, more focused, and a better read.

I posted it. But it kept calling to me from across the room. It said: You have two characters vying for the reader’s attention. Focus on one. And make it shorter.

It was getting late. So I opened it on the platform and edited live. FUCK ME, I think. What am I doing? I am such a nut.

I cut it to 366, and again, it seemed better.

But then I got this idea for a short Note about the madness of writing and then unwriting. I dictated notes to my recorder app. Four or five rambling sentences. I went to bed.

This morning I wrote the Note. Every time I touched it, it swelled like a mosquito bite you keep picking at. It came in at 255 words.

I start thinking about how long a Substack Note should be. Ideally, 50 words? 100 at most? So I cut it to this:

I wrote a 2,500-word story drowning in exposition and lacking tension to draw the reader through. After revision, it was 4,000. I carved it down to 2,400. Tinkering raised it to 2,545. Then I turned an old sketch about one of the characters into a 525-word short, published it, but then cut it to 366. I wrote a Note about writing and unwriting, which came in at 255. Now I’m trimming this Note too, chasing the perfect length. It is 86 words now, including THIS sentence.

50? Would it be better at 50?

Tell me about your madness of writing and unwriting. Keep it to 75 words. Or less.

Subscribe now

Share

.