Finding the Right Role for AI in Your Writing Workflow

Most writers fall into one of two camps when it comes to AI: they either expect it to do everything, then feel let down when the results fall flat, or they avoid it entirely. And honestly? Both reactions make sense. The conversation around AI tends to focus on replacement—AI writing novels, AI taking jobs, AI making human creativity obsolete. For writers who love the craft, who want to write, that framing feels threatening at best and insulting at worst.

But here's the thing: AI doesn't have to be about replacement. It doesn't have to write for you to be useful to you.

Think of AI as a Newbie Personal Assistant

The key to making AI work in your writing life is adjusting your expectations. Your AI assistant isn't a seasoned collaborator who intuits your needs. It's more like an eager intern on their first day—capable, willing, but entirely dependent on your direction.

It needs specific instructions. It needs feedback. When it gets something wrong, it needs to understand why it's wrong and how to do better. The more precisely you communicate, the more useful it becomes.

This might sound like extra work, and at first, it is. But once you've trained your AI on a particular task, you've essentially built yourself a reusable tool.

Putting It to the Test: The Cover Brief Example

Let's make this concrete. Say you need to write a brief for your cover artist. This is exactly the kind of task where AI can shine—not because it's creative work, but because it's detailed, structured work that benefits from thoroughness.

Here's the catch: if you just ask ChatGPT or Claude or Gemini to write you a cover brief, you'll get something generic and probably useless. The AI doesn't know your genre. It doesn't know your audience. It has no idea what makes a cover work in your specific corner of the market.

So you teach it.

Start with your comp titles—those books with covers that capture the vibe you're going for. Feed them to your AI one by one and ask for detailed analysis. Push for specifics: composition, color palette, typography, imagery, mood, how the model is posed or styled, what themes the design communicates to browsers. If the analysis comes back shallow or off-base, say so. Explain what's missing. Ask it to try again.

This back-and-forth is the work. But by the end, you'll have taught the AI—within this conversation, at least—what you mean when you talk about a good cover.

Now you can bring in your own book. Describe the story, the tone, the audience, the emotional hook. Ask the AI to draft a brief based on everything you've discussed together.

Will the first draft be perfect? Probably not. But don't just fix it yourself and move on. Tell the AI what's wrong. Guide it toward a better version. That's how you build something you can actually use again.

Once you have a solid brief, give it a final pass to make it sound like you, then send it off to your artist. Better yet, save the conversation or the final result as a template. Next time you need a cover brief, you won't be starting from scratch.

Why This Works

AI has a few advantages over human assistants for this kind of task. It won't forget the details you mentioned three messages ago. It won't make assumptions based on information you never actually provided. And it's available at two in the morning when you suddenly remember you need that brief by tomorrow.

It's also faster than doing the whole thing yourself, once you get past the initial learning curve.

Finding Your Own Use Cases

Cover briefs are just one example. The real skill is learning to notice where AI might help in your own workflow.

Pay attention to the tasks that make you groan—the ones that feel tedious, repetitive, or administratively necessary but creatively draining. Formatting. Research compilation. Drafting newsletter copy. Summarizing your own series bible when you can't remember a character's eye color. Brainstorming titles or taglines when your brain is fried.

Not every experiment will work. Some tasks won't translate well; some will take more effort to set up than they save. That's normal. There's a learning curve to working with AI, just like there's a learning curve to any new tool.

But once you find the spots where it clicks, you might wonder how you managed without it—not because AI is writing for you, but because it's handling the stuff that was never really writing in the first place.