The Perfect Prompt Doesn't Exist: Why Your Judgment Matters More Than Your Words
There's no magic formula. There's just you — your taste, your decisions, and your willingness to stay in the loop.
Somewhere out there, right now, an author is Googling “best AI prompts for fiction writing.” They’re scrolling through Reddit threads, buying prompt packs, copying and pasting templates from someone else’s workflow into their own chat window — convinced that the secret to getting great output from AI lives in the exact arrangement of words they type into that little box.
I get it. I really do. When you’re new to this, it feels like there must be a magic formula. Like if you just find the right incantation, the AI will finally give you what you want. But after years of working with these tools — drafting novels, editing chapters, building entire series workflows around AI collaboration — I can tell you with absolute certainty: the perfect prompt doesn’t exist.
What exists is something far more powerful, and it’s something you already have. It’s your judgment.
Jay Rosenkrantz, co-founder of Plot Drive and a recent guest on the Brave New Bookshelf podcast, put it perfectly: the major difference between beginners and pros isn’t better prompts. It’s better judgment about what the story needs. The pros aren’t typing magic words. They’re making better decisions about what to keep, what to cut, and what to push further. And that changes everything about how you think about working with AI.
There Is No Perfect Prompt
There, I said it. There is no “one perfect prompt”, just like there’s no one perfect thriller novel, no one perfect bowl of ramen, no one perfect drink. Everyone has taste, and that taste influences everything — including what you want out of an AI.
Someone else’s prompt might be a useful example or a decent jumping-off point. But what you want from the AI and what someone else wants are two fundamentally different things, because your taste shapes the ask. Your genre expectations, your character sensibilities, the rhythm of your prose, the emotional register you’re reaching for — all of that is baked into what “good output” means to you. And no prompt template can capture that.
This is actually liberating if you let it be. It means you can stop chasing someone else’s formula and start paying attention to what works for your stories and your voice. The prompt is just the starting line. Your judgment is the race.
What Judgment Actually Looks Like
Let me give you a concrete example from my own work right now.
I’m currently writing the penultimate book in one of my series. It’s been a slow slog — hello, Future Fiction Academy deadlines — and so I’m working with Claude to draft chapters, which I then wholecloth edit. Every single one. Because my taste runs different from Claude’s in the way we tell stories.
Claude has a style I like to call “reporting.” He writes like he’s reporting all the events of a scene. And while that’s certainly competent prose, it’s not how I write. I write in a deeply immersive style, almost as if the character is speaking the prose directly to you. Not a narrator telling a story — the character telling you the story. But no matter how many times I reinforce this style difference with Claude, he always falls back on reporting.
So what do I do? I don’t keep re-prompting forever, hoping the AI will magically become me. I step into my role as the author in charge. Claude writes the action of the scene I want — events happening in the order I asked for, character interactions unfolding the way I outlined them — and then I come along and use my judgment on what to keep and what to throw away. I rewrite for voice. I deepen the interiority. I make the character breathe on the page.
The end output is a true mix of us both, but I shine through because the character voice is stronger. After a draft is done, the resulting text is about 40% Claude and the rest me. In earlier versions, that number fluctuated — there were iterations of Claude that could more competently write in the style I wanted. (RIP Claude 3.7. Our time together was far too short.)
That’s what judgment looks like. It’s not about getting the AI to do everything perfectly on the first pass. It’s about knowing what “right” looks like for your story, recognizing when the output isn’t there yet, and having the skill and confidence to close the gap yourself.
Context Is the Real Secret
Now, prompting does matter. Specificity always beats vagueness. You should never prompt something like, “Please edit this chapter to be more concise.” A better choice is, “Please edit this chapter to be more concise while still holding to the same meaning and sequence of events.” The details in your ask shape the quality of what you get back.
But in my experience, the prompt you write in the moment is only a fraction of what determines output quality. The real multiplier is the context you’ve already built around the project. Character sheets. World bibles. Plot structures. Scene-by-scene outlines. Emotional arc notes. All of that rich, detailed documentation that represents thousands of your own creative decisions.
When you give an AI that context, you’re showing it your taste and judgment. You’re letting it see the choices you’ve already made so it can make better choices for you. The AI isn’t guessing in the dark anymore — it’s working within the framework of your creative vision.
This is why two authors can use the exact same model, the exact same prompt template, and get wildly different results. The prompt is the same, but the context — the accumulated weight of their individual creative decisions — is completely different. And that’s where the magic actually lives.
Guide the Water
Jay Rosenkrantz’s advice to authors is to “be like water” — stop fighting the model when it doesn’t give you exactly what you wanted, and instead go with the flow, using your judgment to shape what comes out.
I love that metaphor, but I’d reframe it slightly. You’re not the water in this scenario. You’re the one guiding the water. You can contain it. You can coax it into places it doesn’t usually go. You can redirect it entirely when the terrain calls for it. But guiding water is a lot of energy and work, so you have to be strategic about where you spend that effort.
Pick the paths of least resistance for the parts of your process where you don’t need ultimate control. Let the AI find its own way through those sections — it might surprise you with something better than what you had in mind. Then save your energy, your focus, and your editorial insistence for the moments that truly define your story. The character voice. The emotional turning points. The line that has to land exactly right.
You’re still in control — you’re just being smart about where you apply it.
Every Word Is Your Responsibility
I’m not here to harp on the anti-AI crowd. That’s boring and not why any of us are here. But I do want to say this clearly, because it matters: never get lazy with the AI.
Check its output. Every single time. Be the human with taste, with judgment. Never take yourself out of the process. Yes, some parts of your workflow can be automated based on your detailed instructions — and that’s fine, that’s efficient. That’s smart. But in the end, every word on that page is your responsibility. Your name is on the cover. Your readers are trusting you to deliver a story worth their time.
This is the part that gets lost in the discourse sometimes. The “human in the loop” isn’t a buzzword or a disclaimer you slap on your process to make critics happy. It’s the whole point. You are the reason the output is good. Your taste, your editorial eye, your understanding of what your readers want and what your story needs — that’s what transforms competent AI prose into a book that people love.
The AI is a collaborator. A powerful one. But it’s not the author. You are.
Building Your Judgment
If you’re newer to writing and you’re thinking, “But I don’t have strong editorial judgment yet” — that’s okay. Taste isn’t something you either have or you don’t. It’s something you build over time.
I’ve been publishing fiction for over a decade, and there are things I wrote early in my career that I absolutely wouldn’t write that way now. My taste has evolved. My craft has deepened. My understanding of what makes a story resonate has grown through years of writing, reading, learning, and — yes — making mistakes. That’s how it works for everyone.
My advice for building that judgment: learn the craft independently of the AI. Pick up books by authors and academics you trust and familiarize yourself with good storytelling. Study structure. Read widely in your genre. Pay attention to what pulls you in as a reader and what pushes you away. That foundation of craft knowledge will make every interaction with AI more productive, because you’ll know what “good” looks like before the AI ever generates a word.
And trust your taste. Even if it’s still developing, it’s yours. Lean into it.
Your Individuality Is the Point
At the end of the day, we are all unique human beings, and the way we talk to AI is going to be different. There’s no one right way to write a book, and there’s no one correct way to prompt the AI. The authors who thrive with these tools are the ones who stop looking for someone else’s formula and start paying attention to their own instincts, their own creative vision, their own sense of what a story should feel like on the page.
Learning to lean into your individuality is tough. It takes courage to trust your own taste when everyone around you is waving prompt templates and insisting there’s a “best” way. But it’s so rewarding when you see a final output that you love and you’re proud of — knowing that the AI helped you get there, but you made it sing.
Stop chasing the perfect prompt. Start trusting the person typing it.
What’s one moment where your judgment completely changed the direction of an AI output? I’d love to hear your stories — drop them in the comments!



The water/guiding water reframe is the sharpest move in here, and the 40/60 breakdown earns the argument in a way that abstract claims about judgment never quite do. You're writing from inside the practice and it shows.
The place I'd push: you get close to something harder in the 'building your judgment' section and then pull back into reassurance. The implication you don't quite follow is that the foundation has to exist before the tool, not alongside it. Judgment isn't just something you apply to AI output — it's something you develop through the friction of not having the shortcut available. The resistance of the blank page, the dead end that won't resolve, the draft that's wrong in ways you can't yet articulate but have to sit with until you can — that's the training ground. Speed removes exactly the conditions under which that capacity grows.
Which means the writers who start with these tools before they've built that foundation aren't just getting mediocre output. They're potentially training themselves out of the thing that would let them recognize it as mediocre. Your advice — learn the craft independently of the AI — is right. It's also probably the hardest sell in this entire conversation, because the tools are right there and the craft takes years.
None of that undermines what you're saying. It just suggests the stakes are higher than 'trust your taste.' The taste has to be built somewhere the AI can't reach.
As for a specific moment in working with AI, you already pointed at it, it is often the other way round. When the AI, in my case Claude as well, comes up with an idea, a direction I would have never thought of.
What also works nicely, for me, I build extensive character sheets, between 2k to 5k of words. I feed that into a GPT and have discussions with my characters, feed them with situations. I take what I like and forget the rest. But it is often quite eye-opening.