By Tracy Brown, 5 January 2026
‘I write to know what I think.’ (Flannery O’Connor)
O’Connor’s insight captures a truth at the heart of writing: the act itself is a journey of discovery. Each sentence we wrestle with, each paragraph we revise, brings clarity, not just to our readers, but to ourselves. Writing is thinking made tangible. But what happens when artificial intelligence enters the picture? Can AI become a partner in this process, or does it risk short-circuiting the very mechanism through which writers discover their own thoughts?
AI has undeniable appeal. For experienced writers, it can generate ideas, suggest phrasing and help navigate the occasional bout of writer’s block. But for new writers, particularly those who have never written without it, AI can be more of a crutch than a catalyst. The difference lies in the relationship between thinking and writing, which is a relationship AI, however sophisticated, cannot replicate.
The value of writing without AI
At its core, writing is a process of self-clarification. When we write unaided, we confront our ideas in their raw, unfinished form. Struggling to articulate a thought forces reflection: we wrestle with ambiguity, untangle contradictions and confront gaps in understanding. This struggle is where voice is born. Style emerges not from polish, but from persistence, from returning to the page again and again until the words sound like us.
Consider a beginner drafting an essay or story. They pause, scratch out sentences, reconsider word choice and sometimes abandon an idea entirely. This friction, the mental resistance encountered when shaping thought into language, is essential. It is not just about grammar or flow; it is about discovering what we think and how we feel. Writing teaches us our own minds.
When AI steps in too early, it smooths over this friction. It offers fluency without struggle. And while that can feel productive, it can also bypass the very work that makes writing meaningful.
How AI can help writers
This is not an argument against AI. Used consciously, AI can be a genuinely useful tool, especially for experienced writers who already have a sense of voice, perspective and purpose. In those cases, AI functions less as a replacement for thinking and more as a support for execution.
Some of the ways AI can help include:
- Outlining and structuring ideas
AI can help organize complex material, suggest logical flows or surface gaps in an argument. This is particularly helpful when a writer already knows what they want to say but needs help shaping it. - Editing and revision
AI can identify awkward phrasing, repetition or unclear sentences. Crucially, this only works if the writer approaches its suggestions critically. Without that critical stance, AI becomes a mirror that simply validates whatever is already on the page. - Organizing scattered thoughts
For drafts that exist as notes, fragments or rough paragraphs, AI can help cluster related ideas and propose a clearer structure. - Ensuring consistency of voice and tone
For longer projects, AI can flag inconsistencies in tone or terminology, helping writers maintain coherence across chapters or sections.
In all these cases, AI works best as a secondary tool. The thinking still originates with the writer. The judgment still belongs to the writer. The writer remains in control.
What AI cannot do
What AI cannot do is give you insight into yourself.
It cannot tell you what you actually believe, or why a particular idea matters to you. It cannot help you arrive at a position you did not already hold. It cannot replicate the internal shift that happens when, halfway through a paragraph, you realize you were wrong or that the real point is something else entirely.
Finding your own voice is not just about sounding distinctive. It is about discovering your own ideas, your own opinions, your own way of seeing the world. That discovery happens through effort. Through uncertainty. Through writing sentences that don’t quite work and staying with them anyway.
AI also cannot give you the thrill of a breakthrough, the moment when something clicks, when a vague feeling crystallizes into a clear thought. Those moments are not incidental to writing; they are the reward. And they are intrinsic.
Intrinsic vs. extrinsic rewards
This is where the deeper risk lies, especially for new writers.
Writing offers intrinsic rewards: discovery, clarity, the quiet satisfaction of understanding something more deeply than you did before. These rewards emerge slowly, through effort.
AI, by contrast, offers extrinsic rewards: speed, completion, polish. A finished paragraph. A clean draft. The sense of being ‘done’.
When writers rely too heavily on AI, the balance shifts. Completion replaces discovery. Output replaces insight. The work may look finished, but the writer has learnt less from it.
For beginners, this matters enormously. If you skip the struggle, you skip the growth. You may produce text, but you do not develop the muscle of thinking through writing. Over time, that loss compounds.
Tool or crutch?
For experienced writers, AI can be a powerful tool ‒ one that accelerates, supports, and occasionally challenges their thinking. For new writers, especially those who have never written without it, AI risks becoming a crutch that dulls curiosity and replaces the hard but necessary work of self-discovery.
The distinction is not about technology. It is about intention.
If writing is merely a means to an end, AI may be enough. But if writing is how you come to know what you think, how you find your voice, refine your ideas, and understand your own position in the world ‒ then no machine can do that work for you.
In the end, the writer still has to write.
|
Blog post by: Tracy Brown |
