Creative Writing Techniques: A Novelist's Journey with AI

Creative writing techniques aren't learned from textbooks - they're learned in the field, banging your head against the problems every story throws at you. This article isn't a list. It's the account of someone who discovered those techniques one at a time, the hard way: by failing, starting over, and learning to use artificial intelligence not as a shortcut, but as an ally.

Her name is Elena. She's forty-two, an architect, someone who designs spaces where people live. But for months now, every evening after the last technical drawing has been saved, she sits in front of a blank page that has nothing to do with floor plans and structural loads. She's writing her first novel.

Note: Elena is a fictional character. Her story was created to illustrate how creative writing techniques can be enhanced by artificial intelligence. Every situation, creative block, and solution described in this article is invented to show concretely how these techniques work in practice.

Updated on July 31, 2026: This article tells Elena's narrative journey to concretely illustrate how creative writing techniques can be enhanced by artificial intelligence.

Table of Contents

• The Blurred Idea
• Cardboard Characters
• Wooden Dialogue
• Getting Lost in Bologna
• The World Beyond Sight
• The Prose That Suffocates
• The Doors That Don't Open
• The Metaphors You've Read a Thousand Times
• Cracks in the Wall
• The Story's Breath
• The Journey Continues
• FAQ - Frequently Asked Questions

The Blurred Idea

It was a March evening when the idea hit her - a psychological thriller set in Bologna. A peculiar sound, a deserted portico, the sensation of hearing something you shouldn't be hearing. The emotion was all there, but when she tried to put it on paper she realized she had nothing but fragments: an opening scene, a face, an alley, a sense of danger she couldn't trace to any source or destination.

For two months she filled notebooks and phone notes. Isolated scenes. Rough character sketches. Descriptions of Bolognese locations that obsessed her. She had enough material for thirty chapters, but none of those chapters knew how to connect to the others. The scenes worked on their own, but they didn't create a story.

That was when she opened a narrative AI tool for the first time. Not to ask it to write for her - that would take much longer before she felt ready - but to do something simpler and harder at the same time: see the structure she herself had imagined without being able to recognize it.

She described all the fragments, one by one, and asked the AI: "What thread runs through these scenes?" The answer was a revelation. Among the fragments, a recurring theme emerged: perception as a trap, the thin line between what we feel and what we invent. The AI suggested that the protagonist could be a sound engineer, a professional listener who begins to doubt her own hearing. That single intuition turned chaos into a project. Within a week, Elena had a three-act structure, a narrative arc for the protagonist, and a map of the key scenes.

For a deeper look at how AI can support every phase of writing, from planning to revision, our guide on how to write with artificial intelligence offers templates and practical techniques for every moment of the creative process.

Cardboard Characters

With the structure finally in place, Elena dove into the protagonist: Martina, sound engineer, thirty-eight, recently moved into a historic apartment on Via dell'Indipendenza. She wrote a detailed profile - age, profession, hometown, relationship status. Then she started writing the first chapters and noticed a problem: Martina wasn't a character, she was a mannequin. She moved through the scene, delivered lines, reacted to events, but she had no weight. No gravity.

She reread what she'd written and understood why: she was telling us about Martina's emotions, not showing them.

Before - telling: "Martina was anxious. She couldn't stop thinking about that noise. She felt guilty for calling the police, but at the same time she was convinced she was right. Fear was consuming her."

All true. All accurate. And all invisible, because the reader sees nothing - they're only told what to think.

After - showing with AI: "Martina touched her left earlobe, the gesture she'd had since childhood when she lied. She set down the coffee cup and pulled it back immediately, worried the sound of ceramic on saucer had covered something. The phone was there, on the table. She looked at it the way you look at an animal that might bite. If she called the police again, Marco would give her that look - that half-grimace that meant 'here we go again' - and she wouldn't survive that glance one more time."

The difference isn't just in length. It's that in the second paragraph the reader sees the anxiety through fingers touching the earlobe, the cup set down and withdrawn, the phone eyed with suspicion. They're not told "she was anxious" - they're shown the body of an anxious person, her involuntary gestures, her emotional geography. The AI had suggested Elena translate every emotion into a concrete physical gesture, a sensory detail, a micro-action that revealed the inner state without declaring it.

Wooden Dialogue

If the characters were still weak, the dialogue was a disaster. Marco, Martina's husband, spoke like a psychology textbook. Every sentence was clear, polite, perfectly articulated. The problem is that couples married for fifteen years don't talk like that - they speak in allusions, unfinished sentences, things they say and things they don't, and the latter carry twice the weight.

Before - dialogue without subtext: "Martina: 'I heard that noise again.' Marco: 'What noise?' Martina: 'The one from downstairs.' Marco: 'Maybe you should talk to someone about it.' Martina: 'You don't believe me.' Marco: 'I didn't say that.'"

It works. You understand what's happening. But there's no tension. It's dialogue that could belong to anyone.

After - with subtext: "Martina didn't touch her fork. 'You didn't sleep.' It wasn't a question. Marco looked up from his plate. 'Neither did you.' 'The noise from downstairs...' 'I heard it once too. It's probably the pipes.' He poured himself some water, with a calm that didn't belong to him. 'Are you sure you want to stay here?' Martina looked at him. It wasn't worry in his voice. It was something else. It was prayer."

Now there's something underneath. Marco doesn't say "I'm worried about you" - he pours water with a calm that doesn't belong to him. Martina doesn't say "you're as scared as I am" - she reads a prayer in her husband's voice. The AI had helped Elena build dialogue by thinking about what characters don't say: the subtext, the swallowed words, the silences that mean more than sentences. And it had suggested giving each character a distinct voice - Marco is terse, avoids direct questions; Martina is observant, reads between the lines - so the reader can recognize them even without dialogue tags.

Getting Lost in Bologna

Bologna isn't just a setting for Elena's novel. It's a character in its own right - the porticos, the towers, the particular light that filters through the vaults. But the more she wrote, the more she got lost in the details. In one chapter, Martina crosses Piazza Maggiore in five minutes. In another, she stops at a kiosk on Via Rizzoli that Elena had placed there without verification. And when a beta reader pointed out that bus 36 hadn't run down that street since 2019, Elena realized she had a bigger problem: her narrative world wasn't consistent.

The AI became her archivist. Elena created what she came to call the "Bologna bible": a document with the map of named streets, transit routes during the story's period, active shops and bars, market hours, and architectural features of the buildings. Every time a new detail entered the story, she added it to the document and asked the AI to check it against previous chapters.

It worked. Within days, the novel's world was as solid as the foundations of a building Elena herself would have designed. And she gained something extra: detail by detail, Bologna became a place the reader could inhabit, not just visit. If you want to understand how to build believable settings within a compelling plot, our guide on how to write a story addresses exactly this - the coherence between world, characters, and narrative.

The World Beyond Sight

Elena had the opposite problem from what she expected. Her descriptions of Bologna were beautiful - visually. The porticos, the squares, the sunset light on the towers: everything rendered in crisp images. But something was missing. Rereading the chapters, she realized her novel's world was a painting: you looked at it, but you couldn't step inside.

She asked the AI to reread a scene - Martina walking under the portico at evening - and identify which senses were being ignored. The verdict was blunt: sight present, hearing hinted at, smell absent, touch absent, taste absent. Four out of five senses forgotten.

Before - visual only: "The portico stretched ahead of her, a gallery of shadow and light. The brick columns followed each other in rhythm, and at the far end the street disappeared into darkness."

After - multisensory: "The portico stretched ahead of her, a gallery of shadow and light. Her heels changed sound under the vaults - from sharp to reverberating - and the rhythm of her steps became part of the walk. Something was dripping from a gutter, somewhere behind her, with the regularity of a clock. The smell of wet asphalt mixed with that of coffee from the bars that had closed not long ago, a warm scent that dissolved in the cold air. She brushed a column with her palm: the brick was rough, damp, and it left a dark mark on her skin."

Now the reader no longer looks at the portico - they walk inside it. They hear the change in the sound of footsteps, the dripping, smell the coffee and asphalt, feel the roughness of wet brick on their hand. The creative writing techniques that engage all five senses transform a description into an experience, and the AI can function as a sensory reviewer that flags which senses you've forgotten.

The Prose That Suffocates

After weeks spent making everything vivid, Elena fell into the opposite trap. The prose had become too literary, too laden, too full of itself. Every paragraph was a monument to adjectives, every sentence a stylistic exercise. She read a passage aloud and found herself running out of breath - not because the prose was gripping, but because it was suffocating.

Before - baroque prose: "Martina's pupils, dilated by the astonishment and anguish that gripped her chest like a vise of black velvet, settled on the crimson horizon of the Bolognese sunset, where the centuries-old towers stretched their Gothic shadows like the fingers of forgotten giants, while the surrounding portico - that secular cathedral of brick and patience - seemed to hold its breath along with her, in a silence that was both prayer and condemnation."

All that weight just to say one simple thing: Martina watches the sunset and she's troubled. The AI helped Elena cut:

After - lean prose: "Martina watched the sunset beyond the towers. The shadows of the porticos stretched across Via Indipendenza like ink stains. The noise of the city had disappeared, and for a moment she had the sensation that Bologna itself was holding its breath. Then the phone buzzed in her pocket. The spell broke."

Same scene. Same emotion. But now it breathes. The words don't pile on top of each other - they arrange themselves, and leave room for the reader to enter.

The Doors That Don't Open

The first chapters were there, but Elena wasn't satisfied with the openings. Every chapter began with a description, or with Martina waking up, or thinking about something. It worked, but it didn't hook. A distracted reader might have closed the book.

She asked the AI to generate five alternative openings for chapter 4 - the one where Martina hears the noise for the first time at night - each with a different approach: one starting from a reversal, one from silence, one from dialogue, one from action, one from a physical detail. Then she compared them.

Before - weak opening: "Martina was in bed but couldn't sleep. The noise from downstairs had started again, that kind of rhythmic scratching she couldn't identify. She turned the other way, pulled the blanket over her head, but the sound came through everything."

After - opening with a hook: "The silence came before the noise. And that was the real problem. Because the noise, Martina had learned to manage. The silence she couldn't. The silence meant something had stopped - and things that stop, in the dark, usually don't stop for anything good. Then the scratching started again, and for the first time Martina thought that maybe she should have been afraid of the silence more than the sound."

The difference isn't just in the words. It's that the second opening creates an expectation - the silence is the real problem, not the noise - and then flips it again. The reader has to keep going to understand why. Effective openings don't tell you what happens: they promise that it's worth finding out.

The Metaphors You've Read a Thousand Times

Halfway through the novel, Elena noticed an insidious problem: the metaphors. "Her heart raced wildly." "You could cut the silence with a knife." "His eyes were like pins." They weren't wrong, but they were invisible. The reader would glide past them without registering anything, because they'd read them in a hundred other books.

She asked the AI to suggest comparisons tied to Martina's world - a sound engineer - and the results changed everything.

Before - cliché metaphor: "Fear cut through her like a knife."

After - original metaphor, tied to the character: "Fear had entered her like an infra frequency - below the threshold of hearing, but powerful enough to make every surface vibrate."

Or:

Before - generic metaphor: "The silence was total."

After - specific metaphor: "The silence had cracked like a soundproof wall pushed past its maximum load."

Some worked. Others were too technical for the average reader, and Elena set them aside. But the ones that worked - the ones that belonged to Martina's world - gave the novel a new stylistic coherence. The metaphors were no longer decoration: they were characterization. They told you something about the person who thought them.

Cracks in the Wall

When she finished the first draft - seventy thousand words that still felt miraculous - Elena read the whole thing in one go. And she found the cracks. In one chapter, Martina lives on the fourth floor. In another, the third. In one chapter she says she's an only child. In another, she mentions a brother named Luca. In one chapter it's March 15 and there's a thunderstorm. In another, the same day, the sky is clear.

These were errors that an attentive reader would notice immediately, and that would destroy the credibility of the entire story. She copied the whole manuscript into an AI tool and asked for a consistency report: characters, timeline, setting, factual details. In twenty minutes, she had a document listing every inconsistency, chapter by chapter.

She corrected the floor errors. She decided that brother Luca really did exist in the protagonist's backstory - and mentioned him back in chapter 2, where he hadn't been before. She standardized the weather on March 15. The AI hadn't written anything - it had read, with a patience no human would have had for seventy thousand words, and it had found the points where the wall had cracked. This is one of the creative writing techniques many authors underestimate: AI as first reader, the one who sees the cracks you can no longer see because you've been staring at them for months.

The Story's Breath

The last problem was the most subtle. The novel was long, it was consistent, the characters worked, the dialogue had tension, the descriptions were alive. But something was off. Some sections were too slow - three chapters in a row of reflection without action. Others were too fast - two chapters of action without a moment to breathe.

She asked the AI to analyze the pacing, chapter by chapter, classifying each section as "action" or "reflection" and flagging where the balance was off. The result was a visual map of the entire novel's rhythm - and for the first time Elena saw what she'd been feeling as she read: chapters 7, 8, and 9 were a stretch of reflection without a single wave. Chapter 12 was a wall of action without a pause.

She inserted a short action scene breaking up chapter 8. She added a reflection scene in chapter 12, where Martina finally stops and thinks about what she's losing. When she reread it, the pacing was different. The reader could breathe, could tense up, could ride the wave. It was no longer a plain or a cliff face - it was a story that moved the way working stories move.

The Journey Continues

What happened to Elena isn't different from what happens to anyone writing a story for the first time. Creative writing techniques aren't formulas to memorize - they're problems to solve, and you solve them by writing, failing, rereading, and rewriting. Artificial intelligence didn't write Elena's novel. But it helped her see what she couldn't see alone: the structure hidden in fragments, the emotion hidden in gestures, the subtext hidden in silences, the cracks hidden in the words.

If you're writing your first story, don't try to apply all these techniques at once. Choose the one that addresses the problem you have today - flat characters, wooden dialogue, descriptions that are visual but mute - and use it. Then move to the next. Like Elena, you'll discover that creative writing techniques enhanced by AI aren't a trick for writing better: they're a way of seeing more clearly what you're already writing. If writing a thriller is your goal, our complete thriller writing guide covers genre-specific techniques from plotting to pacing.

FAQ - Frequently Asked Questions

How can AI help develop realistic characters?

AI analyzes the psychological traits you've defined and helps you translate them into concrete actions, gestures, and reactions. Its main role is turning "telling" into "showing": it suggests physical details, bodily reactions, and specific behaviors that make a character three-dimensional. The final choice - which details to keep and which to discard - remains yours.

How do you maintain setting consistency in a long novel?

By creating a reference document - a "world bible" - that collects every detail of your setting: streets, transit, businesses, rules. AI consults this document to verify coherence between chapters and flag inconsistencies. It's an essential principle, especially in genres with heavy world-building demands like fantasy or science fiction, but it applies to any story - because even a present-day Bologna requires the same care as an invented world.

Can AI suggest original metaphors and avoid clichés?

Yes, as long as you ask for comparisons tied to the specific world of your character. The most effective metaphors aren't the strangest - they're the most faithful: if your protagonist is a sound engineer, a metaphor drawn from her professional field will be original and coherent at the same time.

Can AI replace the work of a professional editor?

No. AI identifies inconsistencies, coherence errors, and repetitive patterns across large volumes of text at a speed impossible for a human reader. But aesthetic judgment, narrative sensitivity, and stylistic choices remain human skills. AI prepares the ground for the editor - it flags the problems, it doesn't solve them.

How do you balance AI use without losing your authorial voice?

Think of AI as a mirror, not a pen. It shows you your work from different angles, but the choices remain yours. Accept suggestions when they resonate, reject them when they betray your vision. To go deeper on finding and maintaining your authentic voice, read our guide on how to find your own writing style.

Related Articles

• How to Write with Artificial Intelligence
• How to Write a Story
• How to Find Your Own Writing Style

About Books Maker: Our team is made up of AI professionals. Together with expert writers and authors, we created booksmaker.ai to help our users achieve their publishing dreams by leveraging the power of Artificial Intelligence tools for every stage of the process, from idea to book creation.