← Back to Elena Marchetti

Co-Writing a Story: How AI Becomes Your Novel's Best Collaborator

2 min read

Every novel starts as a conversation with yourself. You have a character, a situation, maybe a scene that will not leave you alone. You turn it over. You imagine what the character would do, then argue with your own answer. You try the scene from a different angle. You decide the setting is wrong and start again. This is essentially what co-writing with an AI looks like, except you are no longer having the conversation with yourself. That might sound like a small distinction. It is not.

What the Collaborative Frame Changes

When you are working on a novel alone, there is a peculiar doubling that happens. You are simultaneously the maker of the world and the observer of it, and these two perspectives are constantly negotiating with each other. The maker wants to go somewhere surprising; the observer pulls toward the familiar. The maker falls in love with a scene; the observer knows it needs to be cut. This negotiation is productive but exhausting, and it never fully resolves. The self-argument runs on and on. A collaborator introduces a third position. Suddenly there is someone else in the room who has an investment in the story's internal logic, who will push back when a character acts out of type, who will notice when the middle sags, who will get genuinely interested in what happens next. Research from Northwestern's writing program found that novelists who used collaborative structures — co-writing, writing groups, developmental editors — in early draft stages completed manuscripts at significantly higher rates than those who worked entirely alone. The social or relational element, even when it was minimal, created accountability that self-motivation alone could not sustain.

The Specific Things AI Does Well

AI is not good at every part of novel-writing. It does not have your instincts about what is emotionally true. It cannot access your private imagery, your specific grief, the particular quality of light in the town where you grew up. What it does well is structural thinking. It holds the architecture of your story and helps you see where it is and is not load-bearing. It will notice if your timeline has a gap, if your antagonist's motivation stopped making sense in chapter twelve, if the subplot you set up in chapter three has been abandoned. It is also genuinely good at generating options. When you are stuck — not blocked, just stuck at a decision point — it can produce a range of possible next moves so you can see what feels true. This is not the AI writing your novel. It is the AI helping you find your own answer by showing you what you do not want.

The Tangent About Outlining

Writers argue endlessly about whether to outline or discover. The outline camp says structure prevents the wandering that kills manuscripts mid-draft. The discovery camp says that over-plotting kills the generative surprise that makes writing interesting to do. Both camps are partially right. What they are actually describing is two different ways of managing uncertainty. Outliners manage it up front; discovery writers manage it scene by scene. AI is useful to both. For outliners, it is a stress-testing partner — try your structure against hard questions before you invest months in it. For discovery writers, it is a tracker — it helps you hold the threads you have already laid down so you can weave with them rather than inadvertently dropping them.

Character Depth

One of the highest-value uses of AI co-writing is character development work that never appears in the manuscript. Interviewing your character. Putting them in situations that will not show up in the novel. Asking them questions they do not want to answer. Research from the University of Southern California's narrative intelligence lab found that authors who developed characters through simulated dialogue — literally talking as and to their characters — produced characterization that independent readers rated as significantly more psychologically complex than authors who developed characters through traditional notes-based methods. The novel is not a transcript of the conversation you have with your characters. But the conversation is what makes the characters real enough to write about.

Chat with Sakura
Post on X Facebook Reddit