Table of Contents
TL;DR: AI writes like a brochure because it’s trained on averages. It hits the common arguments, lands the common conclusions, and never says the one specific risky thing only a person who lived the work would know to say. The shallowness is structural, not editable. You can’t fix it by rewriting a few sentences. The fix is starting from your real material, in your real voice, and writing it yourself. Here’s the mechanism in detail, why no amount of prompting solves it, and the cost a reader pays when they pick up the result.
Here’s an AI paragraph on running a customer service operation:
Effective customer service operations require a strategic balance of technology, well-trained staff, and clear processes. Leading organizations leverage data-driven insights to optimize the customer journey, ensuring consistent quality across all touchpoints. By empowering frontline agents with the right tools and authority, companies can drive both customer satisfaction and operational efficiency, creating sustainable competitive advantages in today’s dynamic marketplace.
Here’s a paragraph by a human who has run a customer service operation:
The first thing you learn running a customer service team is that the worst hour of the week is Monday morning at 9, because that’s when every weekend’s worth of broken purchases lands at once, and your best agents are still on coffee. The second thing you learn is that the agents who solve the hardest problems are the ones who don’t read the script, and your job is to figure out who they are before HR fires them for not reading the script.
Both paragraphs are technically correct. One is dead.
The deadness has a name. The shallowness of AI prose is the failure mode that breaks more attempted books, articles, and reports than any other, and it cannot be edited out. Here’s why.
What shallowness actually is
Shallowness in AI prose isn’t a vocabulary problem. It’s not a grammar problem. It’s not a tone problem. You can read AI output and find no specific sentence you’d cut. Every sentence is technically fine. Every paragraph holds together. The piece reads like a piece somebody would write.
The problem is what isn’t there.
The AI paragraph on customer service above tells you exactly nothing you didn’t already know. The strategic balance of technology and people. The data-driven insights. The customer journey. The frontline agents with the right tools. None of it is wrong. None of it is specific. None of it is the thing only somebody who actually ran the operation would know to say.
Shallowness is the absence of the specific, risky, slightly-controversial detail that a working human writer puts in because they earned it. Monday morning at 9. The agents who don’t read the script. The HR problem of keeping the good ones. None of those details show up in an AI paragraph because none of them are average. They’re the things you only know if you sat in the chair.
Why this happens
The mechanism is structural to how large language models work, not a flaw that can be fixed by better prompting or more training.
A language model produces text by predicting the most likely next word given everything that came before. Across millions of training examples on a given subject, the “most likely” word is, by definition, the average. The word most writers would use in that position. The phrase most writing on the topic includes. The conclusion most pieces in the genre reach.
The system isn’t choosing to be average. The system has no concept of choosing. It’s a probability machine producing the highest-probability output. The highest-probability output, across millions of training examples, is the consensus. The consensus is the average. The average is shallow.
This is why prompting doesn’t fix it. You can prompt the AI to “write in a punchy specific voice with concrete details,” and the AI will produce text that looks like text with concrete details. The details will be the average concrete details people write when they’re trying to look specific. They won’t be the ones only somebody who lived the work would know.
The system can fake the texture of specificity. It cannot produce specificity, because specificity comes from lived experience, and the system hasn’t lived anything.
Why no amount of editing fixes it
People try. I see it constantly with clients who’ve drafted something with AI and want to “warm it up” before publication. The instinct is to read the draft and replace the most obviously average sentences with sentences that have more voice.
It doesn’t work. The reason it doesn’t work is that the shallowness isn’t in the sentences. It’s in the structure underneath the sentences. The AI organized the content the way AI organizes content. It hit the common points in the common order. It made the common arguments. It included the standard examples. It reached the standard conclusion. Rewriting a few sentences leaves the structural shallowness in place.
I covered this in a piece on AI-written books that get sent to me for review. The author tries to save the chapters by adding a story here, a joke there, a personal anecdote in the middle. The result is better than raw machine output and still no good. It reads, in the words I used in that piece, like a hollow thing wearing a couple of human accessories, because that’s exactly what it is. The full version is in You Used AI and It Shows.
The only thing that works is rebuilding from real material in your real voice. That isn’t editing. That’s rewriting from scratch with the AI draft as a discarded scaffold.
What readers actually notice
The reader can’t always name what they’re noticing. They can feel it. The piece reads competent and they put it down without finishing. The book is technically about a subject they care about and they don’t recommend it. The article ends up in their tab pile and gets closed without action.
If you ask them why, they shrug. They might say “it didn’t grab me” or “I wasn’t really into it.” What they sensed was the absence of the specific. They expected, on some pre-articulate level, that a person who knew the subject would say something that surprised them, that committed to a position, that risked being wrong, that included a detail only that person would have included. The absence of those things produces the disengagement. The reader didn’t notice the AI. They noticed the work was hollow and stopped reading.
This is the part most people writing about AI prose get backward. The argument is usually “readers can tell when AI wrote something.” That overstates it. Readers can tell when something is hollow. They go looking for why, and they find the AI. The hollowness is the signal. The AI is the explanation they land on second.
The implication is dangerous, because it lets writers convince themselves that if they just hide the AI better, they’ll be fine. They won’t. The reader’s response to hollowness is the same whether the hollowness came from AI, from rushed human writing, from a writer who didn’t know the subject, or from any other source of vacancy on the page. The fix isn’t to hide where the hollowness came from. The fix is to write something that isn’t hollow.
What working writers actually do
Working writers, the ones whose books and essays land, are doing something specific that AI cannot do.
They are including the one detail only they would have included. The 71-year-old client I work with on a memoir is doing this on every page. He includes the specific street his uncle worked on in 1972. The fact that his mother kept a glass jar full of buttons and no two were ever the same color. The argument with his father that lasted three years and ended over a single sentence neither of them remembered later but he can still quote. Every page has at least one detail like that, and the details are what make the page worth reading.
The AI would have written the same chapter as “a complicated relationship with my father that shaped my early adulthood, marked by both warmth and disagreement.” Technically true, completely useless. The human writes “my father said one thing in 1976 that I haven’t been able to get past and I’m seventy-one and probably never will.” Specific, risky, says something only this writer could have said. The full profile of how he does this is in The 71-Year-Old Memoirist Who Uses AI Better Than You Do.
That’s the skill. The skill is including the things only you would have included. The skill is not faked by the machine, ever, because the machine has nothing only it would have included. It has averages.
The mechanism, named
Shallowness is the structural inability of a language model to produce specificity, because specificity comes from lived experience, and the language model has lived nothing.
You can’t prompt around it. You can’t edit around it. You can’t sprinkle around it. You can’t disclose your way around it. The only fix is starting from your real material in your real voice, which is the work AI was supposed to save you from doing, and which is the work that produces something worth reading.
This is the failure mode The Death of Thinking is about at scale. What happens to a profession, an industry, a culture, when the loudest output in the room is the structural average of everything written before it, dressed up in clean grammar, with nothing of any specific person in it. The answer is not encouraging, and it’s the answer the next decade will produce unless individual writers, individually, keep doing the work the machine cannot do.
That work is yours. It starts with the detail only you would have included.
Frequently Asked Questions