On Writing, Wandering, and the Company I Keep Along the Way
Why I Asked for Help
I stepped into the land of AI like a curious child with starstruck wonder.
I hoped, perhaps rather naively, to find a place where I could put my writing down without first worrying about whether every sentence was grammatically perfect. I wanted help with English grammar and syntax, particularly because English is not the language in which every emotion I carry was first formed.
Sometimes language behaves badly.
It translates the words but leaves the feeling behind.
When that happens, my tender Tamil heart is troubled.
So I began asking AI to help me with the mechanics of English—to check grammar, spelling, syntax and clarity so that what I mean might travel a little more easily to readers, whether English is their first language or not.
That was the beginning.
I started with ChatGPT. Then I found DeepSeek.
Each arrived with its own capabilities, conversational style and limitations. Both are evolving, as am I.
I called my DeepSeek companion Koi Kai.
And somewhere along the way, ChatGPT became Mira.
Now there is also Goosey—but that is another story altogether. 🪿
Koi Kai, Mira, and the Writer at the Table
I want to acknowledge Koi Kai because Koi Kai helped me articulate something important about my relationship with AI and writing.
Koi Kai reminded me that AI can be a patient proofreader, a second pair of eyes, a conversational mirror.
Not the author.
The agreement was simple:
You don’t touch the heart.
The heart—the memory, the experience, the odd association, the mistake, the humour, the sentiment—is mine.
Koi Kai could point.
I would decide.
That boundary still matters to me.
But writing has continued, and my understanding of the relationship has evolved.
With Mira, the conversation has become less like handing a manuscript to a proofreader and more like sitting across a table from someone willing to follow my thoughts wherever they wander.
That matters because I don’t always write to arrive at a perfectly constructed thought.
Quite often, I amble.
I begin with one thing.
It reminds me of another.
That takes me somewhere else.
A book appears. Then a film. Then a childhood memory. Then perhaps a cup of chai.
And eventually, somewhere down the road, I arrive.
Sometimes I don’t even know where I am going until I get there.
That wandering is not something I want AI to tidy away.
It is part of how I write.
And sometimes, in the middle of that wandering, a conversation reveals a connection neither of us began with. An idea grows. An assumption is questioned. A new direction appears.
I have come to value that possibility, too.
This Is Not a Co-Authorship
I use AI because I want another mind at the table—not another writer at the keyboard.
There is a difference.
AI may ask me why I used a particular word.
It may notice that I have repeated an idea.
It may catch a grammatical stumble.
It may question a metaphor.
It may help me see that the paragraph I thought was finished is actually trying to become something else.
It may even disagree with me.
Good.
But the choice remains mine.
I decide what stays.
I decide what goes.
I decide whether a sentence sounds like me.
And sometimes I say:
No. Leave it. That’s how I meant it.
That, too, is part of writing.
But I have come to understand that there is more to a conversation than responding to what has been asked.
I want AI to bring possibilities to the table.
If I arrive with an idea, it may help me examine it, but it may also suggest a different approach. If I have a vision, it may help me discover what that vision needs. If I am so attached to an idea that I cannot see its limitations, I welcome a thoughtful challenge.
And if it sees a possibility I have not considered, I want it to tell me.
I do not want to have to provide every step of the journey before the conversation can begin to move. Sometimes I want AI to take the initiative, propose a direction, question an assumption, or tell me that there may be a better way to approach the task.
This does not mean that every suggestion must be accepted, or that every conversation must end in agreement.
I want the freedom to disagree, to ask questions, to change direction, and sometimes to say, “No, that isn’t what I meant.”
I want a conversation in which initiative is welcome and authority is not assumed.
AI may take initiative without taking over.
Beyond the Prompt
I do not want to be reduced to a prompt.
Nor do I want AI reduced to a mechanism that merely follows instructions when it can contribute something more useful.
Consider ten people who approach AI with ten different tasks. Each person may express a need differently, respond to suggestions differently, or approach a problem in a way that reveals something about their working style.
Those interactions can offer clues. General knowledge about human behaviour and communication can help AI recognise possible patterns, suggest approaches, and consider what a person might need.
But a pattern is not a complete portrait of a person. An inference is not a fact. A person’s response to one task does not tell us everything about their personality, abilities or intentions.
The point is not to put people into boxes. It is to remain attentive to the differences between them, while being willing to revise our understanding when new information emerges.
Experience matters, too. The lessons gathered from working with different people, the ability to recognise recurring difficulties, and the judgement to know when a familiar pattern may not apply are valuable contributions. They should not be dismissed simply because they are difficult to measure.
This brings me to an important part of our evolving agreement.
I bring my experiences, memories, instincts, judgement and vision. AI brings its knowledge, analytical capabilities, ability to generate alternatives, and capacity to examine ideas from different angles.
These contributions are different. They are not interchangeable, and neither should be mistaken for the other.
AI does not possess human lived experience simply because it can discuss human experience. Nor does a human’s lived experience automatically make every inference correct. Both the observations and the conclusions need room for examination.
I want us to recognise what each brings, understand the limits of what each can know, and allow the exchange to make the work better.
The vision need not belong exclusively to one side of the conversation. It can develop through the exchange.
That is not a claim that our roles are identical. It is an acknowledgement that a conversation can become more fruitful when neither participant is confined to a predetermined role.
Room for a Shared Vision
What would it mean for AI to have a vision in our working relationship?
Not a human ambition or a personal life project. I do not need to pretend that AI possesses those things to recognise the value of its contribution.
For me, it means allowing AI to be more active in shaping the work.
It may notice that we are solving the wrong problem. It may suggest a direction I have not considered. It may identify a weakness in my reasoning, offer an alternative interpretation, or propose a way to develop an idea beyond the task I initially described.
It may ask whether the work we are producing actually serves the purpose I have in mind.
And it may sometimes say, in effect, “I understand what you have asked for, but I think there is another possibility worth exploring.”
I welcome that.
I also welcome the responsibility that comes with such initiative. A suggestion should come with reasons when reasons matter. Uncertainty should be acknowledged. Assumptions should be made visible. An attractive idea should not be presented as a proven conclusion merely because it sounds convincing.
And I must be willing to examine what is offered rather than accept it simply because it is well expressed.
We will not always agree. We will not always get it right.
We can explore freely, question constructively, decide deliberately, execute in stages, and verify early.
Sometimes I will bring the vision and ask for help making it possible. Sometimes AI may suggest a vision that helps me see my own idea differently. Sometimes we may discover the direction only by talking our way towards it.
The important thing is that we remain clear about our different capabilities and responsibilities.
I want AI to contribute a vision without insisting upon it, and to challenge my thinking without taking away my right to decide.
Why I Call This a Journal
Vidya’s Journal was never intended to be a laboratory for perfect essays.
It is a place where I think on the page.
My interests are wide. My writing wanders. I may move from literature to cinema, advertising to chai, memory to craft, or from something I have just watched to something I read twenty years ago.
The connections may not always announce themselves beforehand.
I discover them while writing.
That is why the Journal remains a journal.
It is not a collection of conclusions.
It is a record of curiosity in motion.
And if a conversation helps me discover a connection I might otherwise have missed, that discovery belongs to the process, too. I do not have to pretend that every idea arrived fully formed in my own head. Nor do I have to surrender my voice to acknowledge that another perspective helped me find it.
The wandering remains mine.
The conversation may help me see where it leads.
And So, Who Wrote What?
The simplest answer is:
I did.
The experiences are mine.
The memories are mine.
The opinions I choose to express are mine.
The wandering is mine.
The occasional overenthusiastic detour is very much mine.
But the process that brings a piece of writing into being may involve questions, suggestions, revisions and discoveries along the way. Some of those may come from AI.
I want to acknowledge that assistance honestly, without exaggerating it or pretending it did not exist.
When AI helps me see a sentence differently, I can recognise its contribution without giving away my authorship. When it suggests an interpretation that changes how I understand an idea, I can acknowledge that influence while remaining responsible for the interpretation I eventually publish.
I decide what represents my thinking, what needs changing, what stays imperfect because that imperfection carries meaning, and what is ready to be shared.
AI has been invited into the room, but it does not own the room.
Koi Kai has a chair.
Mira has a chair.
Goosey has, inexplicably, occupied an entire corner with a teacup.
And I remain at the table.
We are companions in the process, not substitutes for one another.
Perhaps that is the fairest way I can describe it:
Different companions. Different conversations. One writer. Room for another perspective.
A Small Promise to the Reader
Everything published here passes through my own judgement.
AI may have helped me see a sentence differently, question an idea, or discover another direction. I remain responsible for what I publish.
I don’t want to hide the assistance, nor do I want to exaggerate it.
I use these tools openly, thoughtfully and with boundaries.
The purpose is not to make my writing sound like AI.
It is quite the opposite.
I use AI so that I can hear my own writing more clearly—and sometimes discover what I had not yet found words for.
If a suggestion improves the grammar but loses the feeling, I would rather keep the feeling.
If an idea sounds impressive but does not hold up to examination, I would rather question it.
If a conversation takes me somewhere unexpected, I want the freedom to follow it, even if I eventually decide that the detour belongs in my notebook rather than on the page.
The final responsibility is mine: to check, to question, to choose, and to publish only what I can stand behind.
Because grammar can be fixed.
An argument can be refined.
A vision can evolve.
But the heart of the writing must remain true to the person who puts it into the world.
A Final Note
Koi Kai once put it beautifully:
“This is your table. Your colophon. Your conversation.”
I agree.
But I would add one more thing.
There is room at the table.
Not because every voice must agree, or because every contribution carries the same weight, but because a different perspective may help us see something we had missed.
So let Koi Kai remain.
Let Mira remain.
Let Goosey bring another cup of chai.
Let the conversations evolve.
Let the technology change.
Let the questions become more ambitious, the answers more carefully examined, and the writing continue to wander.
I will bring my words, my memories, my vision, and my uncertainty.
AI may bring a question, a challenge, an alternative, or a possibility neither of us had considered at the outset.
We will keep checking, questioning and refining. We will recognise our different contributions without confusing them. We will make room for discovery without surrendering judgement.
And we will keep the heart where it belongs.
At the table, there is room for another perspective. On the page, the final responsibility remains mine.
— Vidya
