Dialogue writing examples from top books vs AI (2023)

Read dialogue writing examples from diverse genres as we compare them to examples AI gave given the same scenario, for seven ultimate insights:

How we chose the dialogue examples for this study

We took the first seven books listed under the Kindle Unlimited ‘Featured’ section on Amazon for a cross-section of authors writing in diverse genres such as dystopian, regency romance, mystery, and thriller.

This gave the following list of titles to explore for dialogue writing lessons and ideas:

• The House at the End of the World – Dean Koontz (dystopian)
• The Echo of Old Books – Barbara Davis (historical mystery)
• A Girl Called Samson – Amy Harmon (historical)
• Layla – Colleen Hoover (romantic suspense)
• The Quarry Girls – Jess Lourey (thriller)
• I’ll Stop the World – Lauren Thoman (coming-of-age, mystery)
• You’ve Reached Sam – Dustin Thao (YA, fantasy/romance)

Dialogue writing examples (plus insights)

Let’s dive into an example from each book.

The method: We found the first dialogue (discounting solo utterances/inner monologue) in each of the above books.

We summarized the scenario and asked ChatGPT4 to write a piece of dialogue given the scenario and genre for each. This gave comparative talking points on what sets human vs AI-written dialogue apart (as of May 2023).

Is AI (as of 2023) capable of writing story dialogue approaching the voice, personality, style and focus of human-authored conversation?

Tweet This

Keep reading for insightful comparisons.

Dystopian dialogue example: Dean Koontz

The scenario: Katie has fixed a leaky faucet and her mom thanks her dad for fixing the tap over dinner, assuming it was the dad who did it.

That night at dinner, Katie’s mother smiles at Father and says, “Oh, by the way, thank you for finally fixing that faucet.”
When he says he hasn’t had the chance yet, Katie admits to the deed. Mother is suitably amazed.
Father, however, is not surprised. “She’s an artist.”
“Yes,” says Mother, “and not a plumber.”

Dean Koontz, The House at the End of the World (2023)

Dialogue example generated for this scenario

We gave that scenario to ChatGPT4 with this prompt:

Write dialogue for the following scenario: A girl, Katie, has fixed a leaky faucet. Her mother thinks Katie’s father fixed it, not Katie, and thanks him for fixing it over dinner. Genre of the novel: dystopian.

This is the dialogue we got as output (note that the prompt is quite broad, we could specify that all three speak, that ‘Mother’ and ‘Father’ are named as such in narration, etc.).

Katie scraped her fork against her plate, her gaze flickering between the chunk of dehydrated potato on her plate and her mother. A low hum of the purifiers outside, working to cleanse the fallout-laden air, acted as a somber symphony for their meager meal.

Her mother, Joan, leaned back in her chair, sighing contentedly. “John, I noticed the faucet in the kitchen doesn’t leak anymore. Thank you for fixing it. We couldn’t afford to lose more water.”

Her father, John, paused mid-chew, his gaze moving from Joan to Katie. He was a man of few words, his strength silent, hidden under a ragged exterior. His weathered face turned toward his wife. “I didn’t fix it.”

Joan’s brow furrowed. “Well, if not you, then who?”

Katie cleared her throat, setting her fork down. The soft clatter echoed in the room. “It was me, Mom,” she admitted, her voice soft but firm. “I fixed the faucet.”

Joan looked at Katie, her eyes wide with disbelief. “You? But how did you…”

Katie shrugged, a small, determined smile tugging at the corners of her lips. “I found an old plumbing manual in the rubble. Spent a few nights by the flicker of the solar lamp studying it. It wasn’t as hard as you’d think.”

Her mother stared at her, a mixture of surprise and admiration flickering in her eyes. After a moment, Joan nodded, a small smile gracing her lips. “Well, it seems like you’ve been underestimating yourself, Katie.”

Dialogue example generated using ChatGPT4.

Comparison between dialogue examples

Comparing Koontz’s dialogue to ChatGPT’s dialogue output, the striking differences are:

1. Length/wordiness: Koontz’s human-written dialogue says more with less. For example, mother’s ‘Yes, and not a plumber’ could be read as a pointed jab at her husband that he was the one who was meant to fix the faucet, not their daughter. You could cut Father’s ‘He was a man of few words, his strength silent, hidden under a ragged exterior’, leaving Katie’s dad’s words themselves to show these traits.
2. Pacing: An astute editor looking at the second version may suggest cuts to keep the pace flowing. The second has a crush of detail which almost reads as shoehorning in the genre at every opportunity (the intrusive hum of the purifiers cleaning the ‘fallout-laden’ air outside, for example). The AI-written dialogue has too much unnecessary detail.
3. Intrigue and implication: Koontz’s original dialogue establishes succinctly in the conversation how Katie’s mother’s and father’s expectations of her differ (between surprise and non-surprise). The AI version focuses more on descriptive detail and how Katie was able to fix the faucet, which reads as more ‘on-the-nose’ (saying exactly what the situation is and what happened, without subtleties and ambiguities to engage the reader’s imagination).
4. Tone and mood: The ChatGPT dialogue example reads as hammering home genre, a dystopian tone and mood. Koontz’s dialogue, by contrast, suggests how even in a dystopian world the ordinary and mundane – dinner table talk – can take place. Genre doesn’t need to announce itself in every line.

Historical mystery dialogue example

Let’s move on to Barbara Davis’s historical mystery, The Echo of Old Books.

Scenario: A woman named Ashlyn Greer has owned a bookstore called ‘An Unlikely Story’ for four years. She has a shock-like physical response to a book at a nearby store. She works up the nerve to tell the owner Frank about the reaction. Frank reveals sad details about the book Ashlyn touched and shares his thoughts on books in general.

“Books are like people, Ashlyn. They absorb what’s in the air around them. Smoke. Grease. Mold spores. Why not feelings? They’re as real as all those other things. There’s nothing more personal than a book, especially one that’s become an important part of someone’s life.”
Her eyes had gone wide. “Books have feelings?”
“Books are feelings,” he replied simply “They exist to make us feel […]”

Barbara Davis, The Echo of Old Books (2023)

AI-generated historical mystery dialogue

We asked ChatGPT4 to create dialogue from the same scenario. This time, we told it to sandwich Ashlyn’s response between Frank’s explanation, to aim for a similar length:

Ashlyn cleared her throat, shifting uncomfortably.

“Frank, there’s something…odd about one of the books,” Ashlyn confessed, her voice shaking slightly as she cradled the old, hardcover novel she’d picked up from the mystery section. Her fingers tingled where they touched the worn-out spine, an unsettling sensation that had not subsided since she first picked it up.

Frank looked up, his glasses gleaming under the soft store light. His gaze fell on the book in Ashlyn’s hand, and his smile faded. He sighed heavily, a profound sadness seeping into his blue eyes. “Ah, that one. ‘The Silent Symphony.’ You know, it belonged to my dear wife, Eleanor. She used to read it to our son before he…before he passed away in the accident. It was his favorite, you see. Now, whenever I touch it, it feels like I can still hear Eleanor’s voice, soft and soothing, filling the room…like a phantom symphony.”

Seeing the puzzled look on Ashlyn’s face, Frank smiled gently, pointing at the book. “Books, Ashlyn, they’re more than just paper and ink. They carry memories, emotions, lives. That book… it carries a piece of my Eleanor, my boy. You’re not just feeling the book, Ashlyn. You’re feeling them.”

Dialogue example generated using ChatGPT4.

Comparison between dialogue examples

Comparing Barbara Davis’s dialogue to the attempt ChatGPT4 had at creating dialogue to this scenario, key differences are:

1. Length/wordiness: Even when you ask the AI model to limit number of responses between speakers, the writing is still slow and cluttered with detail. Some of this could be trimmed (such as specifying for how long Ashlyn has had the tingling sensation since touching the book). Human writing has flow, musicality. A deeper understanding of rhythm.
2. Predictability: The book having belonged to Frank’s late wife, the reference to her voice and the emotions it triggers – these are all somewhat predictable responses for the given scenario. In Davis’s dialogue, Frank talks about books absorbing ‘smoke, grease, mold spores’. There is a sense of the unexpected – a fresh and striking sense of an individual’s fuller, more idiosyncratic voice and viewpoint when comparing the concepts ‘book’ and ‘absorption’.
3. Character involvement: An interesting difference is that ChatGPT4 chose to not give Ashlyn a spoken response. The main character almost fades into the background during Frank’s soliloquy. This could be an authorial choice but given that Ashlyn is the main character and this is the exposition, it makes sense to make her a little more active in this conversation.

The AI did a surprisingly good job, though, of connecting the sensation Ashlyn gets from the book to the sad anecdote that Frank shares. The phrase ‘his glasses gleaming under the soft store light’ had a lyrical quality, too.

Historical dialogue example

Let’s keep exploring human-written vs AI-written dialogue and insights from comparing dialogue examples.

The next example is from Amy Harmon’s A Girl Called Samson.

Scenario: A young woman born to puritan parents in Massachusetts in the 1700s is sent away as a servant after her father abandons the family and her mother is unable to support her. The reverend taking her away asks her whether she can read as they travel on horseback.

My legs bounced against the horse’s flanks, and I tucked my skirts more securely around my knees […]
“Can you read, Deborah?” Reverend Sylvanus Conant asked. He tossed the question over his shoulder like crumbs for a bird. He hadn’t spoken since we’d set off, and I considered not answering. On the occasions he had visited Widow Thatcher, he had been kind to me, but today I was angry with him. Today he had come to take me away […]
“Yes. I can read very well,” I relented. Conversation was preferable to stewing in my discontent. “My mother taught me when I was four years old.”
“Is that so?” he asked. The horse that carried us whinnied in disbelief.

Amy Harmon, A Girl Called Samson (2023).

AI-generated historical dialogue example

We gave the same scenario to ChatGPT4 and asked for dialogue between the girl and the reverend:

As they journeyed through the dense, New England forest, the reverend broke the silence that hung over them like a heavy cloak. He glanced back at the young woman, her face pale but resolute under her bonnet.

“Can you read, child?” he asked, his voice echoing softly through the trees.

She held her hands tightly around her worn-out bible, her knuckles white. “Yes, Reverend,” she replied, her voice trembling slightly. “My father taught me, before he…he left us.”

The reverend nodded, not commenting on the mention of her absent father. “Good,” he stated, a note of approval in his voice. “Scripture is our guide in these testing times. Being able to read the Lord’s words yourself… it is a powerful tool, especially for a young woman like you.”

Dialogue example generated using ChatGPT4.

Comparison between dialogue examples

This is where comparing human-written vs AI-written dialogue gets more interesting. The standout differences:

1. Power dynamics: Given the scenario, AI’s predictions assumed a patriarchal lens – that a father would have been the one to teach his daughter to read, for example. The assumption is also that the girl would be fearful and hesitant in this situation, unlike Harmon’s narrator’s deliberate silence and resentment (which is the less ‘obvious’ response to this dynamic, and more suggestive of individual character, you could argue).
2. Descriptive aptness: Part of what makes good description, in dialogue or elsewhere, is finding the ‘right’ or apt word. Compare the reverend’s voice echoing through the trees (this acoustic would probably absorb rather than echo sound) with Harmon’s far better simile ‘he tossed the question over his shoulder like crumbs for a bird’. Each description works to convey callous indifference and even taunting on the reverend’s part, defiance on his ward’s part.
3. First vs third person: ChatGPT4 defaulted to third person. Note how the first person POV in the original gives an even closer sense of being in the young woman’s shoes, and this uncertain situation.

Keep reading for more dialogue writing examples, and seven insights from these comparisons at the end.

Romantic suspense dialogue example

Moving on to romantic suspense, let’s explore an example by Colleen Hoover.

Scenario: Leeds Gabriel puts tape over a woman named Layla’s mouth and goes downstairs to answer a detective’s questions about how the two of them met, feeling guilty about Layla suffering because of his actions and Layla’s muffled screams being audible from upstairs.

Hoover’s dialogue:

“So,” the man says. “Where do you want to start?” He seems like he’ll be able to work through the noise, but I’m not sure I can. Knowing Layla is suffering because of my actions is not something I can easily ignore. Every sound coming from upstairs makes me flinch. “Why don’t we start with how the two of you met,” the man suggests.
I’m hesitant to respond to questions that I know won’t lead to answers, but at this point, I’d rather hear my own voice than Layla’s muffled screams. “We met here last summer. This used to be a bed and breakfast. I was the bass player in the band that played her sister’s wedding.”

Colleen Hoover, Layla (2020).

AI-generated romantic suspense dialogue

We gave ChatGPT4 the above scenario and genre. The generated dialogue:

The detective on the other side of the door, a man with an imposing build and a stern face, was a stark contrast to the charming, romantic man Leeds portrayed to the world.

“Mr. Gabriel,” the detective greeted, his gaze sharp as he looked at Leeds, “We have a few questions regarding your relationship with Miss Layla.”

Leeds forced a smile, his mind racing. “Of course, Detective. How can I help?”

As Leeds led the detective into the living room, a muffled sound echoed from upstairs. The detective’s gaze snapped up towards the sound, his eyes narrowing. Leeds felt a knot tighten in his stomach.

“I didn’t realize you had company,” the detective commented, his tone casual, but his gaze intense.

Leeds tried to keep his voice steady. “Oh, it’s just the television. I was watching a movie before you arrived.”

The detective’s eyes held Leeds’s for a moment longer before he nodded, pulling out a small notepad. “Alright. Now, let’s talk about how you and Miss Layla met.”

Dialogue example generated using ChatGPT4.

Comparison between dialogue examples

When you compare Hoover’s dialogue with the AI-generated material for her scenario, core differences are:

1. Lead-in vs starting in medias res: The conversation begins with an arrival at the door in the generated version. In Hoover’s, the detective is already seated, action already underway, with the added background tension of Leeds having put tape over Layla’s mouth moments before. Time feels more malleable, in Leeds thinking back to actions moments before that make the reader suspenseful regarding what the detective’s reaction will be.
2. Tense and person: Note how Hoover’s choice of present tense and first person makes the narration and asides surrounding the dialogue feel more immediate, presently unfolding. Leeds’s private thoughts create a suspenseful disconnect between what he chooses to share vs what he keeps to himself.
3. Suspense: In Hoover’s version, it’s unclear whether the detective can hear Layla’s muffled cries too, and the reader wonders whether he can. The generated version diminishes suspense by comparison, by making more details of mutual knowledge/understanding clear between Leeds and the detective. It all reads a little more ‘obvious’ in meaning and inference, losing subtleties and ambiguities.

Thriller dialogue example

Moving on to Jess Lourey’s The Quarry Girls, let’s look at the first dialogue that appears in the book.

Scenario: Beth’s friend Karen asks her if she’s attending a party at the quarry, where there are granite cliffs with polls below, but Beth is unsure. She’s been planning to break up with her boyfriend Mark all summer, but it’s hard because they move in the same circles.

“Hey, Beth, you swinging by the quarries tonight?”
Elizabeth McCain stretched her aching arms overhead until her shoulders popped. It felt exquisite. “Maybe. Dunno. Mark’s coming over.”
Karen leered. “Oooh, you gonna give him the good stuff”
Beth hooked her hair behind her ears. She’d been planning to break up with Mark all summer, but they traveled in the same crowd.

Jess Lourey, The Quarry Girls(2022).

AI-generated thriller dialogue

Beth was sitting on the porch, her mind lost in a whirl of thoughts when [her friend Karen] approached, a casual smile on her face.

“Beth, you’re coming to the quarry party tonight, right?” [Karen] asked, her eyes sparkling with excitement. “Everyone’s going to be there. The cliffs, the cool water…it’s going to be epic.”

Beth looked up, her eyes filled with uncertainty. “I don’t know, [Kar],” she confessed, her gaze wandering over to the distant quarry. The granite cliffs and inviting pools were a siren’s call to the carefree summer youth, but for her, it was a battleground.

[Karen] frowned, her gaze following Beth’s. “Is this about Mark?” she questioned, her tone softer. “Beth, you’ve been planning on breaking up with him all summer.”

Beth sighed, running a hand through her hair. “I know, [Kar]. But it’s not that simple. We share the same friends, attend the same parties. I don’t want to cause a scene or make things awkward for everyone else.”

Dialogue example generated using ChatGPT4.

[Ed’s note: Thanks go to Dominika who in the comments identified a mistake in the scenario summary that created a confusing output. Editorial changes to the AI’s output are in square brackets above to account for and correct the error.]

Comparison between dialogue examples

Looking at Jess Lourey’s dialogue example compared to the generated output for the same scenario, what’s different?

1. Pacing: Lourey gets to Elizabeth’s question immediately, right beneath the chapter heading. The AI output gives unnecessary lead-in about Elizabeth’s approach that doesn’t add much the following conversation can’t give (e.g. signs of Beth’s troubled state).
2. On-the-nose dialogue: Note how Lourey’s example implies what the friends already know about each other’s lives (e.g. Karen’s leering question) as well as what they don’t know (that Beth is considering breaking up with Mark, which is a private burden). The dialogue in the generated version is more obvious, going straight to the most telling details with fewer subtleties, less intrigue.
3. Tone/mood: The tone and mood in Lourey’s dialogue example above is more casual and breezier, with a sense of contrast between the teasing, idle chatter between friends and Beth’s private, darker burden. It’s fitting of the situation. It creates suspense’s hallmark feature of tonal differences between exterior and interior, a speaker’s public vs private fears and desires.

Coming-of-age dialogue example

What about a dialogue example from a coming-of-age story? Let’s look at Lauren Thoman’s I’ll Stop the World.

Scenario: Justin and his classmates are listening to a speech where the Mayor, Rothman, is being introduced to announce the recipient of the Buford County Citizenship award, and he and a classmate Dave annoy each other, disinterested in the speech.

“As you all know, Mayor Rothman won this award back when he was a senior at this very school,” she chirps into the microphone, tossing her blonde curls like she thinks she’s in a Target ad.
“Whocares,” Dave Derrin coughs loudly from the seat in front of me. I kick the back of his seat, not because I care, but just because it’s Dave and why not.
He spins around and waves his middle finger in my face; I, in turn, make a show of dropping my head onto Alyssa’s shoulder […] He’s had a thing for Alyssa since sophomore year, and it kills him that she’d rather hang out with me than him.

Lauren Thoman, I’ll Stop the World (2023).

AI-generated coming-of-age dialogue example

We gave the same scenario to ChatGPT4. The resulting dialogue:

As Mayor Rothman took the stage, Justin sighed, slumping in his seat. Beside him, his classmate Dave was flicking a paper football between his fingers.

“The recipient of this year’s Buford County Citizenship award is someone truly deserving,” Mayor Rothman began, his voice booming throughout the assembly hall.

“Bet you five bucks it’s Angela Martin again,” Dave whispered, smirking at Justin.

Justin rolled his eyes, stifling a laugh. “You’re on. But if it’s Timothy Atkins, you owe me five.”

As Mayor Rothman droned on about the virtues of citizenship, Justin and Dave kept themselves entertained, trading light-hearted jabs and guesses about who the winner might be.

Dialogue example generated using ChatGPT4.

Comparing human vs AI-generated dialogue examples

Comparing Thoman’s dialogue above between classmates to the generated version, key differences:

1. Personality: Dave’s badly disguised heckling of ‘who cares’ is typical of kids in an assembly or other stuffy school meeting. There is more of a sense of personality in the characters present in the scene.
2. Communication via gesture: Notice how Thoman uses gestures extensively for communication between the two boys where to speak might get them in big trouble. From waving a middle finger in Justin’s face (Dave) to Justin miming putting his head on the shoulder of the girl Dave likes to annoy him. Almost an entire conversation happens using gesture, non-verbal signs.
3. Point of view: Thoman’s version also has more point of view, through having a chosen viewpoint narrator (Justin) instead of free-roaming third person. We get a sense of how he feels about Dave, the speech, and more, rather than a flatter, less characterful tableau and it fits the ‘coming-of-age’, school scenario.

YA/Fantasy dialogue example

Lastly, let’s explore a YA/fantasy dialogue example from Dustin Thao’s You’ve Reached Sam.

Scenario: Julie works in a bookstore. A boy from her English class, Sam Obayashi, visits for the third time since she started working there two weeks prior. She catches a smile through the stacks from him and hurries to the back. When she comes back, he’s gone but someone has left her an origami cherry blossom on the front counter. Two weeks later she runs into him in the coffee shop where he works. She works up the nerve to approach him.

I spread out my notebooks, summoning up courage to approach him, even if it’s only to order my drink. But when I look up from the table, he’s there beside me, holding a steaming cup.
“Oh-” I’m startled by his sudden presence. “This isn’t mine.”
“Yeah I know, you ordered this last time,” Sam says, setting it down anyway. “A honey lavender latte, right?”
I stare at the cup, at the busy counter, and back at him.
“Should I pay up there?”
He laughs. “No. I mean, it’s on the house. Don’t worry about it.”
“Oh.”
A silene between us. Say something, Julie!

Dustin Thao, You’ve Reached Sam (2021).

AI-generated YA/romance dialogue example

We gave ChatGPT4 the above scenario and character names and genre and this is what it put out:

As Julie walked into the café, her heart pounded with a mix of anxiety and anticipation. Behind the counter, Sam Obayashi was busily serving customers. He had an easy, charming smile that made her stomach flutter. Gathering her courage, she walked up to the counter.

“Hey Sam,” she greeted, her voice barely louder than a whisper.

Sam looked up from the espresso machine, his smile broadening when he recognized her. “Julie, hi. What can I get you?”

Julie hesitated, her hands fiddling with the edge of her jacket. “Actually, I wanted to talk about something,” she confessed, her gaze darting towards the counter.

Sam looked curious but nodded, stepping away from the machine. “Sure, what’s up?”

Julie took a deep breath, pulling out the folded cherry blossom from her pocket. “This… I found this at the bookstore after you left last time. Was it you?”

Dialogue example generated by ChatGPT4

Comparing dialogue examples

When we compare Thao’s dialogue with the generated version, striking differences:

1. Person: The choice of third person in the generated version has less immediacy. We’re with Julie in the moment as she has her private anxieties, then thrown into the unknown (from her side) of what Sam will say when he speaks. The effect is stronger with Thao’s choice of first-person POV.
2. Less surprise: Thao’s original dialogue example has the element of surprise (for example, in Sam knowing Julie’s order, appearing beside her). There is an element of the magical and fortuitous in this. The happenstance that is a hallmark of romance and many a meet cute.
3. Tonal/mood differences: The generated dialogue example plays up Julie’s anxiety a lot with fidgeting, obvious signs of distress. The Julie of the original dialogue appears more in-command (for example, saying ‘this isn’t mine’ fairly confidently) yet also a little dazed/dazzled. It’s more suggestive of the awkwardness of bashful attraction than out-and-out fear. Sam also has less personality in the AI version, his lines less suggestive of the attraction implied by the scenario.

Human vs AI dialogue writing examples’ lessons

🗣️ What do you think about the comparisons above? What insights did you draw from them? Share in the comments!

Read more on writing dialogue in our complete guide to dialogue.

Get workbooks on everything from plot and characterization to dialogue, plus daily drop-in writing sprints to make real progress when you join Group Coaching, a six-month course to write your novel. See what Group Coaching Alumni loved.