You use AI for code without thinking twice. When the same AI writes text, something feels wrong. The reason is simpler than you'd expect.
A developer commits code written by Claude. The pull request gets approved. The feature ships. Nobody asks who wrote it.
A writer publishes a blog post. Someone flags it as AI-generated text. The comments section becomes a courtroom.
Same technology. Same level of involvement. Completely different reactions.
You've probably used Copilot, Cursor, Codex or Claude to write code this week. You didn't feel guilty about it. You didn't add a disclaimer to your commit message. You didn't wonder if your GitHub profile should carry an asterisk.
But when you read a blog post and suspect it was AI-generated, something shifts. The ideas feel cheaper. The sentences feel hollow. You close the tab with a vague sense of having been tricked.
What changed? Not the technology. Not even the quality.
(If you want to know if and when it's a good idea to disclose you are using AI for your specific work, check the quick, interactive, free app at the end of this article.)
The Double Standard Nobody Mentions
AI-generated code now accounts for more than a quarter of all production code shipped.¹ At some AI companies, the number is closer to 90%. The developer community didn't protest. They celebrated. Faster shipping, fewer boilerplate bugs, more time for the decisions that actually matter.
Meanwhile, an AI-assisted LinkedIn post still gets treated like a confession.
The usual explanations for why people hate AI writing don't survive contact with this comparison. "AI writing lacks soul." So does most code. "AI writing is lazy." But AI coding is efficient? "AI writing deceives the reader." Your users don't know or care whether you wrote their login page by hand.
The gap has a simpler explanation: visibility.
Text Is the Product. Code Is Not.
When you write, the text is the end product. The reader consumes your exact words, your sentences, your rhythm. The writing is the thing.
When you code, the code is invisible. Users see buttons, screens, outcomes. They never see the functions, the variable names, the architectural decisions. Other developers see the code, but they're (mostly) judging function, not authorship. The code hides behind the experience it creates.
The checkout flow runs on code nobody will ever read. The About page gets scrutinized word by word. Same site. Completely different rules.
This is why AI code gets a pass. Nobody consumes the code itself. They consume what the code produces. The medium is invisible, so the method doesn't matter.
Text has nowhere to hide. Every word is on display. And when those words feel generic, the reader feels cheated. The ideas might be fine. The medium failed to carry them.
The Ghostwriter We Already Accepted
Here's what makes the outrage selective: ghostwriting has existed for centuries.
JFK's "Profiles in Courage" won a Pulitzer. Ted Sorensen wrote most of it. Alex Haley wrote "The Autobiography of Malcolm X," one of the most influential books of the 20th century. J.R. Moehringer wrote Prince Harry's "Spare." The entire Nancy Drew series came from a rotating cast of writers behind the fictional "Carolyn Keene."²
The ideas came from a person. The execution got assistance. The reader got value. Most readers didn't care.
AI is the new ghostwriter. Same arrangement, different tool. Yet "I used a ghostwriter" is professional while "I used ChatGPT" is scandalous.
The difference is mostly cultural. We've had centuries to normalize one and months to adjust to the other.
But the parallel has a seam. Ghostwriting was scarce. You hired a specific person with a specific voice. The collaboration was curated, expensive, deliberate.
AI assistance is instant, universal, and essentially free. When ghostwriting was rare, it was a competitive edge. When everyone has a ghostwriter, the edge disappears.
The Intimacy Problem
With AI images and video, the artificial origin is still sometimes visible. Odd fingers. Uncanny motion. With text and code, AI involvement can be invisible. And that invisibility creates a specific discomfort: the feeling of being deceived.
But we only care about that "deception" when the product addresses us personally. Nobody feels fooled by AI-generated code because they never encounter it directly. The anxiety is reserved for text, where the reader feels personally addressed by words that might not have come from a person.
Look underneath and you find something more personal: intimacy. Text creates a relationship between writer and reader. Code creates a relationship between product and user. When you discover the writer might be a machine, the relationship feels violated. When you discover the code was machine-written, there's no relationship to violate.
Reading on Merit
I read everything based on what it gives me. If an article sharpens my thinking, I don't care whether a person wrote it, an AI wrote it, or they wrote it together.
This sounds obvious. Right now, it's actually a radical position.
Most of the stigma around AI writing orbits around authenticity, not quality. "I wrote this" used to describe a process. Now it declares personal worth. When AI threatens that identity, the reaction targets the writing, but the real wound is to the writer.
Here's a thought experiment. Take an article you found valuable. One that sharpened an idea, offered a framework you kept, or reframed a problem you'd been stuck on. Now imagine learning it was AI-assisted.
Does the value disappear? Does the framework stop working?
If the answer is no, then your objection was never about the content.
A ghostwritten memoir that transforms how you think about grief delivers more than a hand-typed blog post that says nothing new. The origin matters less than the outcome.
But "judge on merit" is easier to say than to practice. Nobody evaluates everything from scratch. You use names, track records, reputations as filters. Authorship is a trust signal, not just an ego marker.
The honest version of this position is narrower than the clean one: once you're reading, merit is what matters. But what made you start reading probably involved trusting a name. That trust isn't irrational. It's how attention works when there's more to read than anyone can filter. So trust it, and then concentrate on what you're actually reading.
What Visibility Reveals
We accept AI code because we never see it. We reject AI writing because we always do.
The ghostwriter taboo took centuries to dissolve, and hasn't fully. The AI writing taboo is trying to cover the same ground in months. The double standard will probably narrow as AI writing becomes as routine as AI code already is.
But the question underneath won't dissolve with the taboo. When authorship stops being a reliable signal, what replaces it?
Maybe just specificity. Not a person wrote this, but this particular person noticed this particular thing and couldn't have written it from any other vantage point.
That was always what made writing worth reading. We just didn't have to think about it until the alternative showed up.
Want to find out what your actual specific situation is?
Answer 6 quick questions to get a clear analysis and personalized next steps (with 2254 possible results, including one that matches your case), all here- no gate, no signup, completely free.
Rabbit Hole
If this double standard made you think about your own reaction to AI, AI Is a Self-Esteem Test argues that your response to AI is the most honest measure of where your confidence actually rests.
The tendency to judge new tools by old standards has deep roots. You're Building a Stone Cathedral Out of Concrete explores why every new material first gets used to mimic what came before.
For a different angle on AI and the writing process, AI Writing Companion as Leverage looks at using AI to pressure-test your thinking rather than replace it.
Footnotes:
- As of early 2026, AI-authored code accounts for roughly 27% of production code industry-wide, with the share approaching 50% at fast-adopting companies. At Anthropic, engineers report the figure is between 70% and 90%.
- Ghostwriting in publishing is well-documented. The "Carolyn Keene" pen name has been used by dozens of writers since 1930.