Transcription
Look, AI is not going to magically turn you into a great writer. But what it will do is make great writers absolutely terrifyingly awesome. The people who already think deeply, who already care about the craft, those are the people who are going to get to strap on an exoskeleton into their brain, essentially using AI.
My name is Jason. I wrote 14 books the traditional way and then I spent 3 years over at Kindlepreneur teaching people how to write and publish books before going full-time on this channel. So, I've lived on both sides. There's the grind-it-out, no AI, stare at a blank page world and then this new world where AI can type almost as fast as you can think if you use it correctly.
In this video, I'm going to give you 11 specific ways to use AI to become a better writer. Five of these are going to be on how you can get AI to write better and six are going to be ways to use AI to make your writing better. The goal is simple. We want the AI to do the heavy lifting, but the human stays in charge of the thinking and the art.
Let's start with making AI less dumb. Number one, give AI one job at a time. Most people talk to AI like it's some magical intern who can do everything in one shot. Like come up with an idea, structure it, write it, and polish it. That's how you get generic, mushy, and middle-of-the-road slop. Instead, we want to treat AI like a specialist. In one phase, it's only helping you generate ideas. In another, it's only organizing what you've already said into a structure. And in another, it's only tightening sentences and cutting fluff. When the model has one clear job, it stops trying to be clever and starts being precise. This is why I use automation so I can have one step for everything. The strategy is to have one brain for one task. Break your process into stages and assign AI exactly one stage at a time.
Number two, stop dumping your whole manuscript into the AI and instead give it structured context. So, what I mean by this is that yes, technically some tools will let you paste in huge amounts of text, even your whole book if you want, even a whole series. And yes, that's usually a bad idea for two reasons. When you overload it with text, it has to average across everything, and so you dilute the prompt overall, you dilute what matters, and you burn a ton of tokens and money for very fuzzy understanding on the AI's part. Instead, you want to summarize and structure. For a book, write short summaries of each chapter. And yes, you can use AI to help you make these summaries. Summarize what happens, who changes, what promises you made to the reader, etc. And when you need AI's help with a new chapter, you give it those summaries instead of the full text of everything. Think of it like bringing a friend up to speed. You don't read them your whole manuscript. You say, "Okay, in chapter one, X happens, and in chapter two, Y happens." You want AI operating off of that same kind of concise overview, so it can actually reason about the story instead of drowning in raw text.
Number three is to test different models instead of being married to one brand. Most writers pick one tool and assume it's the best because it's either popular like ChatGPT, maybe they think because it's expensive it'll be the best one like Claude, or perhaps someone on YouTube said it's the best, but different models have different personalities and they're better for different things. Some are more chatty and creative, some are more logical and structured, some are way cheaper and 90% as good for a specific task like writing prose for instance. In fact, I've found that a smaller, cheaper model like what AI has put out can sometimes give you prose that feels cleaner and less AI-ish than the big-name one like Claude. The strategic move is to run small experiments, take the same small task and compare outputs across models. Ask which one feels closest to how I would have done it. Then match the tool to the task. And this is what I do literally with every automation I build. I take every single step and I test it across about a dozen different models to see which output for that specific step I like the best. And I'm constantly going between different models and different model providers. So, you might have one that is your go-to for structure, another for your prose, another for a more like technical cleanup. You don't need to be loyal to any one particular brand. You just need the right tool for the right job.
Number four, refuse to let AI be better than you are. If you never develop your own taste and your own understanding of structure, your own storytelling instincts essentially, AI will outpace you. If you do develop these things and you add AI on top, you will outpace everyone else. AI magnifies what's already there. And if your skill is low, AI just lets you create more low-quality content faster. If your skill is high, AI lets you express that skill at a much higher volume. So, the rule is to use AI to accelerate your learning and output, never to avoid learning and output. Every time you reach for AI, ask, "Am I using this to skip thinking or to get more at bats with my thinking?" If you hold that line, AI becomes a force multiplier. If you cross it, AI quietly replaces you.
Number five, build a living style guide so AI can actually sound like you. When you tell the AI to just write in my style, it has no idea what that means. You have to teach it. And so, instead you want to treat your voice like something that you can document. You do that by picking a handful of your best pieces. So, this could be chapters you've written, emails you've written, even blog posts. Study them for patterns, and you can use AI to also do this analysis for you, but make sure you are heavily involved in this process. You want to know like how long your sentences usually are, how often you crack jokes, how formal or casual you are, whether you lean on stories or pure explanation, uh whether there are specific phrases you love uh or phrases you don't love. And you turn all of those observations into a list of rules and a style guide. Then, whenever you use AI, you anchor it in that guide as well as samples of your writing to sort of demonstrate what your what that guide is meant to to show. You're basically saying, "Before you start typing, this is the persona that you're adopting." So, over time, as you evolve, you update the guide and it grows alongside you. As a result, AI stops sounding like generic internet voice and starts sounding a lot more like you on a good day.
By the way, if you made it this far into the video, you might be a good fit for the Story Hacker membership. Link down below to get started.
So, let's go into part two of this video where we're going to talk about how to use the AI to improve your writing instead of improving the AI's writing. Cuz now we we need to flip it. So far, we've been making AI better. Now, we're going to make you better using AI as a coach, mirror, or training partner.
So, tip number six is to use AI as your interviewer to kill writer's block. Most people think that they have an idea problem. And they actually don't. They have more of an extraction problem. Your ideas are there, but they might not just be top of mind for you. So, when you're just staring at a blank page, you're asking your brain to decide what matters, structure it, phrase it beautifully, all at once. And of course, that's going to cause you to freeze. This is where AI becomes your interviewer. You tell it what you're trying to write about. You let it ask you a series of smart and targeted questions, and you answer them one by one the way you'd answer a curious friend. You stop worrying about perfect phrasing, and you just talk. So, this might be stories that you'd forgotten that come up or opinions you didn't know you had on the surface, or connections between ideas that appear because you're in conversation and not isolation. And when you're stumped, you can look at the examples of how other people might answer and react to them. Or you can ask the AI to give you examples, and usually when it gives you options, you can say, "No, that's not it. Here's what I actually think." That "No, here's what I really mean" moment is often where your real voice is showing up. Cuz remember, you're not asking AI to write for you, you're using it to pull the book that's already in your head into the open.
So, number seven is to let AI critique you, but of course never abdicate judgment. AI is great at spotting certain problems. For instance, repetition or overly long tangled sentences or confusing transitions. It can also do a decent job at spotting obvious clichés. And used well, it's like having a very fast, slightly nitpicky editor sitting next to you. But you do have to remember that it doesn't know your long-term vision, it doesn't know your readers the way that you do or at least the way you should do. It doesn't care about preserving your unique quirks. So, when you use it to critique your writing, you're not saying tell me what's right and wrong, you're saying show me things that I might want to look at. Then you sort that feedback into buckets, things like okay, yes, this is a real clarity or pacing issue that I should look into or that's not quite right, but it's pointing at something useful or that suggestion would make my writing more generic, so pass. Uh essentially, you want to think of it like watching a game tape if you're an athlete. The footage doesn't tell you who you are, it just shows you what happened so you can make better decisions next time. You remain the coach and AI is just another camera angle.
Number eight is to treat AI like a writing tutor and not a ghostwriter because there's a big difference between AI do my homework and AI show me what I did wrong and how to fix it. The first one makes you dumber and more dependent over time. The second one accelerates your learning like crazy. As a tutor, AI can help you break down why a passage from a great author works. It can compare your writing to that passage and highlight the concrete differences. It can suggest specific micro exercises to practice the skills that you're missing. Instead of randomly writing and hoping you get better, you're actually doing deliberate practice with AI as your coach. Focused repetition, doing dialogue exercises, scene transitions, description, pacing, etc. Gives you immediate feedback on each rep. And the goal is not to end the session with perfect pages. The goal is to end the session as a better writer who can produce more good pages tomorrow and next week and next year.
Number nine, use AI as your outlining and reverse outlining partner. Organizing ideas into coherent structure is where a lot of people get stuck. They can write decent paragraphs, they just don't know how to turn a big pile of thoughts into a book. AI can help from both directions. So first, there's the outlining. You can collect all your ideas and stories and concepts in one place, then you have AI explore different ways of arranging them into a logical progression. And you end up with several possible outlines. Maybe one that starts with your personal story, one that starts with the reader's problem, one that's purely chronological, and you can decide exactly what that would look like because when you're using AI, you essentially are choosing between options instead of just staring at chaos. Then there's reverse outlining. This is when you feed in a chapter from a book you admire or one of your own, and AI breaks it down into its underlying structure. What each part is doing for the reader. And with when you do this, patterns start to pop out. You might learn how long the setup is before something happens, how often tension strikes, how they enter and exit scenes. And once you see those patterns, they're kind of hard to unsee. You start building them into your own work on purpose instead of by accident.
And this actually leads me into tip number 10, which is to tear down comparison books and competitor books without copying them. When you're writing in a specific genre, there are a few unspoken rules. There's expectations about pace and tone. There are tropes that readers expect to see, and if they don't, they will leave you bad reviews. There's types of twists that feel kind of satisfying, and normally you'd have to read tons of books and slowly piece that together. But with AI, you shortcut a lot of that analysis. You can take a successful book in your genre, you can have AI walk you through like what's happening in each chapter, what changes, what's promised, what's paid off, etc. It pulls out recurring patterns in plot, character arcs, emotional beats. And the important part here is that you're not trying to copy their plot. You're instead learning the grammar of the genre. It's like a musician learning chord progressions. Lots of songs use the same basic chords, but they sound completely different because of melody and rhythm and lyrics. So, you can do the same thing with story. You understand the core moves that make readers in that genre happy, and then you invent your own version of those moves with new characters, settings, and details.
And that brings me to tip number 11, last one, use AI as a brutal beta reader panel. But do it carefully. Most writers either get no feedback, or they get feedback that's too vague to be useful. AI can simulate a room full of readers with different attitudes. A die-hard fan of your genre who notices every trope, for instance, or a casual reader who gets bored easily, or a harsh critic who lives to find flaws. You can ask each of these voices to react to your chapter. This allows you to understand where they get bored, where they were confused, what moments really landed, what would make them excited to keep reading, etc. Now, of course, if you can, get real human beta readers, but they need to be beta readers that love your genre already so they can give you accurate feedback. And you need to have enough of them that you can start to see the trends. Most authors, especially when they are starting out, don't have that. They don't have the resources to find those people. And so, having AI stand in for that is great. If you ask for brutal feedback, AI will always find something wrong, no matter how good your chapter is, and that's the role because that's the role you're giving it. So, you want to be careful that you're not taking everything it says absolutely literally, but it can absolutely be a great tool to give you an idea of where you are falling short. You want to treat this as a generator of possible problems to investigate, a way to see your writing from multiple angles, a tool to sharpen your own editorial instinct, but you remain the person who decides if this is a real issue or if it's just nitpicking that wouldn't actually make my book any better.
AI is not the enemy of writers. What kills writers is avoiding hard thinking, outsourcing judgment, and hoping a tool will save them from doing the reps. If you use AI the way we just walked through, including narrowing your jobs, structuring your context, getting feedback, all of these things, then AI becomes the thing that lets you do more reps faster without losing yourself in the process. And if you want to see exactly how all of this fits into a start-to-finish system, here's your next step. I've put together a full course video here on YouTube that walks through my entire process. And in that process, AI does the typing, the human does the thinking. So, if you're thinking, "Okay, Jason, these 11 ideas are great, but how do I string them together into a real workflow that gets me from zero to published?" That's exactly what this full course video is for. So, click over to that next, watch it, take notes, and then pick one small part to implement this week. And I'll see you in the next video.