Style has been a subject of debate among writers and rhetoricians for centuries. What makes Hemingway sound like Hemingway? How does Stephen King create that unmistakable voice that pulls readers through hundreds of pages? And in our AI-driven world, can machines truly capture and replicate these distinctive voices?
As I stood in front of my "Writing with AI" class recently, I found myself pondering this question more deeply. We've all seen amusing examples of AI imitating famous authors or writing styles, often with entertaining results. But as both a writing professor and someone who studies AI writing systems, I've remained skeptical about how well these systems truly understand and replicate style.
The question isn't merely academic. As organizations increasingly rely on AI to generate content at scale, understanding the capabilities and limitations of style replication becomes crucial for maintaining authentic brand voices and creating genuine connections with audiences.
So I decided to put my skepticism to the test with a classroom experiment that yielded some surprising results—and reinforced an important lesson about how we should approach style when working with AI systems.
The Style Challenge
Style remains one of the most elusive concepts in writing. As Kate Ronald noted in her essay "Style: The Hidden Agenda in Composition Classes," style isn't just about word choice or sentence structure—it's about establishing a "presence" on the page, creating a genuine connection between writer and reader.
For content developers and professional writers, style is more than an academic concern—it's a practical challenge they face daily. Brand voice guidelines, content style guides, and tone frameworks are all attempts to codify and standardize style across teams and channels. Yet these documents often fall short because style is inherently difficult to articulate.
This challenge became evident when I asked my students to define certain writing styles. When prompted to describe "professional" style, they typically offered vague synonyms like "concise" or "straightforward." But when pressed on what those terms actually look like in practice—how many words make something "concise" or what sentence structures create "straightforwardness"—they struggled.
Different contexts call for different interpretations of these concepts, and even experienced writers often lack the vocabulary to articulate how stylistic elements manifest in language. A "professional" email to a client differs from a "professional" white paper, which differs from "professional" social media content.
This ambiguity creates a significant challenge when working with AI. When we ask AI to "write in the style of Hemingway," it uses patterns from its training data to imitate what it recognizes as Hemingway's characteristics. This might produce entertaining results, but if you genuinely want writing that captures Hemingway's essence—or your brand's unique voice—you need to be much more specific.
For content teams managing multiple writers (human or AI), this specificity becomes even more critical. Without clear, detailed style guidance, content inevitably drifts from the intended voice, creating inconsistent experiences for audiences.
Trump's Distinctive Style: A Test Case
On a whim during class, I decided to test AI's style imitation capabilities with a highly distinctive voice. Regardless of politics, Donald Trump's speaking style is instantly recognizable—making it a perfect test case for AI imitation.
I began with Trump's widely available January 6th speech and asked Claude to analyze his distinctive stylistic traits. Instead of vague directives like "write like Trump," I wanted a detailed breakdown of what actually constitutes his recognizable style.
Claude produced this remarkably detailed style prompt block:
[STYLE] Jump between topics without formal transitions, creating a stream-of-consciousness feel. Incorporate personal anecdotes and digressions, then circle back to main points. Use informal language and colloquialisms like "you know what?" and "by the way." Engage directly with your audience through rhetorical questions and acknowledge their reactions. Employ superlatives heavily ("greatest," "biggest," "most amazing") and make categorical statements. Mix serious claims with casual asides and humor. Use active voice and present tense to create immediacy. Build intensity through accumulated examples, and layer your message through multiple restatements. Break up longer thoughts with interjections and parenthetical comments. Create contrast between "us" and "them." Sprinkle in numbers and statistics, but present them conversationally rather than formally. Most importantly, maintain an informal, off-the-cuff tone that feels like natural speech rather than a prepared text. Let ideas flow freely, even if they seem to wander, and trust that you'll bring them back to your central themes.
What struck me was how precisely Claude had captured the stylistic elements that make Trump's communication style distinctive. The analysis went far beyond superficial traits, identifying structural patterns like topic-jumping, rhetorical strategies like audience engagement, and specific linguistic features like superlative usage.
The real test came next. I asked AI to write a lesson on "AI and the writing process" using two approaches:
First, using the detailed style prompt block above
Then, with a simple instruction without the style prompt block: "Write a lesson on AI and the writing process in the style of Donald Trump."
The difference was quite remarkable.
With the Style Prompt Block
The first response, using the detailed style prompt, produced something remarkably Trump-like:
You know what's absolutely wild? The way AI has completely transformed how we write. I mean, completely transformed it. The other day I was struggling with writer's block - you know that feeling when your brain just refuses to work? - and I thought about how different this whole process is now.
Here's the thing about writing with AI (and trust me, I've thought about this A LOT): it's like having this super-smart friend who never gets tired of brainstorming with you. Crazy, right? But wait, let me back up for a second.
Remember the old days of writing? Just you, staring at a blank page, wondering if you'll ever write anything good again. The biggest misconception - and I see this all the time - is that AI is here to replace human writers. That's not it at all! You're still the creative genius behind everything. The AI is just... well, it's like having the world's most patient writing partner.
Speaking of partners - oh my gosh, this reminds me of something hilarious - I was working with a writer friend who kept asking their AI assistant for "something unique" without giving any specifics. Like, come on! That's like walking into a restaurant and saying "give me food." You've got to be specific, people!
The writing captured Trump's distinctive rhetorical patterns with uncanny accuracy: the mid-sentence digressions, the exaggerated statements, the colloquialisms, the direct audience address, and the stream-of-consciousness flow. It conveyed the presence of Trump's voice while discussing a completely unrelated topic.
Without the Style Prompt Block
The second response, using only the basic instruction to "write like Trump," produced something quite different:
First thing you need to know - and this is very important, very important - is that AI is like having a brilliant assistant, maybe the most brilliant assistant ever. When you're writing, here's what you do, and listen carefully because this is huge:
Start with your beautiful human ideas - and humans have the best ideas, don't let anyone tell you different. Your initial draft, you write it yourself. Very important. Nobody does initial drafts like humans, nobody.
Then - and this is where it gets really interesting, really interesting - you use AI as your collaborator. Like having a fantastic writing partner, maybe the best writing partner ever. You can say, "AI, make this paragraph stronger" or "Fix these terrible, terrible grammar mistakes." And it does it, folks. Just like that.
But - and this is crucial, very crucial - you're still the boss. The AI, it's working for you. Some people, they just let AI write everything - disaster! Total disaster. You need to maintain creative control, like running a successful business. Trust me, I know about successful businesses.
While this version contained some Trump-like elements—repetition, superlatives, and business references—it missed many of the nuanced patterns captured in the detailed style prompt. It relied heavily on numbered lists (not characteristic of Trump), focused on shallow stylistic markers like word repetition, and lacked the natural digressions and topic-jumping that give Trump's speech its distinctive rhythm.
The difference was striking. Without specific guidance, the AI attempted imitation but missed the mark—overemphasizing certain elements while missing others entirely. It produced a caricature rather than an authentic stylistic imitation.
The Takeaway: Style Prompts Matter
This experiment revealed something critical about AI and style: the quality of style guidance directly impacts the quality of AI-generated content. With detailed guidance, AI can produce remarkably convincing stylistic imitations. Without it, the results are superficial approximations.
For content creators and writers, this has profound implications:
Develop detailed style guides: Generic instructions like "professional tone" or "conversational style" won't produce consistent results. Break down stylistic elements into specific, actionable components.
Capture structural patterns: Style isn't just vocabulary and sentence length. It includes patterns of organization, rhetorical strategies, and distinctive structural features.
Test and refine: Like any writing process, developing effective style prompts requires iteration and refinement.
Think beyond rules: Traditional style guides often focus on what to avoid. Effective AI style prompts should emphasize what to include—the positive markers that create a distinctive voice.
For organizations managing content at scale, investing in detailed style prompt blocks may be worthwhile. These blocks serve as a bridge between human stylistic intuition and AI's pattern-matching capabilities, enabling more consistent content production across teams and channels.
The experiment also reinforces a broader truth about AI writing tools: they complement rather than replace human expertise. AI didn't spontaneously understand Trump's style—it needed human analysis to break down what makes that style distinctive. The most effective applications of AI in writing continue to be those that leverage human insight and direction.
Ready to Master Style in Your AI Prompts?
If you found this exploration of AI and style intriguing, there's much more to discover. In my PromptOps course, we dive deeper into creating effective style prompt blocks, testing different temperature settings to enhance stylistic control, and building systems that maintain consistent voice across your content operations.
Learn how to craft style prompts that deliver reliable, high-quality results every time—not just approximations. Join other writers, content creators, and AI practitioners who are mastering the art and science of style in the AI era.
Check out Lesson 7: "Understanding Temperature & Style in Prompt Design" in my PromptOps course to take your AI style work to the next level!