Ask a writer who is not worried about AI what they are currently working on, and they will tell you something oddly specific. Not a topic. A particular question they cannot stop turning over. A tension between two things they have experienced and cannot yet reconcile. A moment they keep returning to because it still has not given up its meaning.
They are not performing calm. They are oriented toward something AI cannot replicate: themselves. Not their style, not their niche, not their output volume. The one characteristic these writers share has nothing to do with writing ability.
What “specific” actually means
Not specific in the way writing teachers mean it when they tell students to use concrete nouns. Specific in a deeper sense. Specific in the sense that the writing could only have come from one person’s accumulation of experience, observation, contradiction, and way of seeing. The kind of writing where, if you tried to replicate the voice without replicating the life, something essential would be missing.
This is different from having a recognizable style. Style can be imitated. What cannot be imitated as easily is a particular psychology made visible on the page. A particular relationship with grief or ambiguity or place or memory. A particular way of noticing things and then deciding what the noticing means.
Writers who have spent years developing this kind of specificity, whether consciously or not, tend to be less worried about AI for one simple reason: they understand that the most compelling thing about their work is not the writing itself. It is the writer behind it.
The confusion between skill and self
A lot of anxiety about AI and writing conflates two things that are worth separating: writing as a skill set and writing as an act of self-disclosure. Both are real. Both matter. But they are not the same thing, and they are not equally replicable.
Writing as a skill set involves grammar, rhythm, structure, clarity, argumentation, and craft. These are learnable, teachable, and yes, increasingly automatable. If a writer’s value proposition rests mainly here, on producing clean, readable, well-organized prose, then the anxiety makes sense. That ground has shifted.
Writing as self-disclosure is something different. It involves the writer bringing something of their actual interior life into the work. Their uncertainty. Their contradictions. Their specific history with the subject. Their willingness to be seen thinking something through, rather than presenting a finished position. This kind of writing creates a different relationship with the reader. It is less about information transfer and more about recognition. The reader does not just learn something. They feel accompanied.
AI can produce the first kind of writing competently. It has much more difficulty producing the second, not because it lacks language, but because there is no self behind it to disclose.
What these writers do differently
A food writer who has spent twenty years thinking about what hunger means culturally and personally is doing something different from a food writer producing content about trending ingredients. A travel writer who explores why certain places make them feel temporarily more alive is doing something different from one producing itineraries. A psychology writer who examines their own emotional patterns alongside the research is doing something different from one summarizing studies.
The distinction is not genre. It is depth of investment in the self as source material.
The relationship between specificity and staying power
There is a deeper structural reason why this quality matters, beyond AI. Writing that comes from a genuine self tends to have staying power in ways that technically proficient but self-absent writing does not. Readers return to writers, not just to writing. They develop a relationship with a particular sensibility. They want to know how this specific person makes sense of something.
This was true long before AI. Writers who treated their own inner life as raw material, who were willing to bring their confusion and contradiction into the work alongside their insight, tended to build readerships that followed them across formats, topics, and years. The work becomes a body of work. It has coherence not because the topics are consistent but because the consciousness behind them is recognizable.
Joan Didion writing about grief, James Baldwin writing about race and self, Annie Dillard writing about attention itself — none of these are defined by a consistent topic. They are defined by a consistent consciousness. Readers follow them not because they know what the next piece will be about, but because they know whose mind they will be inside.
What AI has done is not so much threaten this kind of writing as clarify its value. In a content environment where the production of competent, well-structured prose is increasingly cheap, the things that AI cannot produce cheaply, namely a genuine self, a particular history, a specific way of being in the world, become more valuable, not less.
The characteristic, stated plainly
Writers who are least worried about AI share this: they have invested in themselves as the source of their work, not just in their ability to produce work.
They may not have framed it that way. They may simply have been writing honestly for a long time, following their genuine curiosity rather than optimizing for reach, staying close to what they actually find interesting or confusing or worth examining. But the result is the same. Their work contains something that cannot be extracted from them and placed into a prompt.
This is not a guarantee of commercial success. It is not a simple instruction. You cannot decide to have a rich interior life and publish it by the end of the week. But it points toward what has always been true about writing that lasts: the work is only as interesting as the person behind it is willing to be.
What this means for working writers now
It is worth asking, honestly, what proportion of your writing currently contains something only you could have written. Not only in the sense of personal anecdote, but in the sense of genuine perspective. Ideas you have actually sat with. Positions you have arrived at through real uncertainty. Observations that came from your specific accumulation of experience.
If the answer is a small proportion, that is not a reason for despair. It is a direction. Concretely: write one thing this week that you would not have written if you weren’t the one writing it. Not a personal anecdote dropped into an otherwise generic piece. Something where the angle, the uncertainty, or the conclusion could only have come from your specific accumulation of experience. Do that consistently, and the gap between your work and what AI can produce begins to widen in the only direction that matters.
AI has made a great deal of writing easier. What it has made harder to replicate is a writer who has genuinely shown up in their own work.
