top of page

A Question for ChatGPT! Have We Forgotten How to Write?

Writer: Paul Johns
Paul Johns
Sep 30
9 min read

I walked into Marks & Spencer in West London today and was asked to write down a few simple details. Nothing complicated. The sort of information you should be able to put on a piece of paper without giving it a second thought.

There was a cup of pens on the counter. I picked one up, leaned over the paper and began to write. A familiar task, except that it no longer felt particularly familiar. My hand seemed to be working harder than the occasion deserved. The letters were awkward, the spacing uncertain, and I found myself looking at the result with a slight sense of embarrassment.

When had my handwriting become so bad?

Somewhere along the way, I had conceded cursive to a MacBook. Years of emails, presentations, messages and documents had turned me into a capable typist who had almost forgotten how to write. My thoughts still reached the page, but through a keyboard. Give me a pen, and suddenly something I once did effortlessly required concentration.

It was a small moment, but it stayed with me. A capability I had taken for granted had become less available simply because I had stopped using it. There had been no conscious decision to abandon it. I had found something easier, used it repeatedly, and gradually lost some of the fluency I once possessed.

As I left the store, I began to wonder where else that might be happening.

Over the last two years, AI has found its way into so many parts of our working lives. We use it to summarise, organise, research, draft, refine and respond. It helps us get started when the page is empty and keep going when the diary is full. For those of us working in marketing, communications and strategy, its usefulness is obvious. Much of our work involves turning complicated ideas into something another person can understand.

I use it myself. I appreciate what it makes possible. This reflection comes from being inside that change, enjoying its benefits and beginning to question some of the habits I have developed around it.

What surprises me is how quickly those habits have formed. My handwriting deteriorated over years. This feels as though it has happened almost between meetings. A tool we were experimenting with became something we consulted, then something we relied upon, and sometimes the place we went before we had even tried to form a thought ourselves.

How did we give up so much ground so quickly?

There was no moment when I consciously decided that the difficult business of finding the words, shaping an argument or establishing a point of view should become someone else’s responsibility. There were simply deadlines, busy days and a succession of useful shortcuts. Each one saved a little time. Together, they began to change how I approached the work.

Lately, I have noticed a kind of blurring. I see it in some of the work around me, and I recognise it in my own. The language is competent. The structure is sensible. Everything appears to be in the right place. Yet sometimes I finish reading and struggle to identify what the person actually thinks.

There is a fuzziness beneath the fluency. A paragraph can sound complete before the idea inside it has been properly worked through. A point can arrive beautifully dressed but still have very little to say. Because the words are easy to accept, we move on without noticing that something important remains unresolved.

That concerns me more than an obvious mistake. An obvious mistake invites scrutiny. Smooth language tends to receive a warmer welcome. It looks like progress. In a busy working day, that can be enough to get something approved, shared or added to a presentation.

But writing has always been part of how I work out what I believe. I rarely begin with a perfectly formed thought and simply find the words to describe it. I discover the thought through the writing. I put something down, read it back, realise it is too broad or too easy, and try again.

Sometimes a sentence exposes a weakness in the argument. Sometimes I repeat a point because I have not yet understood what I am really trying to say. Sometimes the paragraph I thought was the beginning belongs halfway through. The process can be untidy, but something useful is happening in that untidiness.

I am making decisions. I am distinguishing between what sounds appealing and what I can stand behind - this is really important and speaks to the integrity of what we convey.

When too much of that process arrives pre-packaged and already completed, I have fewer encounters with my own thinking. I become an editor of a plausible answer before I have developed a position of my own and this has started to make me feel uneasy. The work moves quickly, but I am less certain that I have travelled the full distance with it and for me this is critical - candidly I enjoy my own emotional dilemma with the organic creativity I own.

Perhaps that is the professional equivalent of picking up a pen and discovering that your hand no longer quite remembers what to do.

Over the last twelve months, I have started trying to recover some of that ground. I have become more deliberate about curating my own language, or perhaps re-curating it. Giving an idea time before asking for help. Writing the rough version myself. Allowing the opening to be clumsy while I establish what I actually mean.

There is a temptation to tidy too early. You have a few thoughts, a half-developed argument and a deadline, so you ask for something coherent. Within seconds, there it is. The relief is immediate. Yet coherence can close a conversation with yourself that needed to remain open a little longer.

I want to spend more time in that unfinished place. To hear where my language catches, where an observation leads somewhere I had not anticipated. Some of it will be discarded. Some will need considerable work. But I will know where it came from.

AI should support our thinking, not supplant it. It can amplify an idea, explore its implications and accelerate a concept towards something useful. It can challenge an assumption or suggest a connection we had not considered. But it cannot become a proxy for our reasoning while we retain only the responsibility of pressing send.

There is a difference between recognising a convincing answer and understanding how you reached a conclusion. If I cannot explain why I believe something, what experience informed it or which assumptions hold it together, then I have more work to do. The quality of the wording does not relieve me of that responsibility.

Our ingenuity and individuality need room to operate. They emerge through the connections we make, the questions we ask and the things we notice that someone else might pass by. I do not want to surrender so much of that process to a large language model that my own contribution becomes difficult to find.

I remember going to hear Vivaldi’s The Four Seasons at St Martin’s with my partner. Sitting in the church, watching the musicians settle into the performance, I became aware of the physical effort involved. The movement of the bows, the concentration on their faces, the attention passing between the players. We were close enough to see that the music required something of them.

As the familiar passages unfolded, I found myself listening to the small variations. The breath before a phrase. The tension in a quieter moment. The way a change in pace seemed to move through the musicians before reaching the room.

I enjoy electronic music, but that evening gave me something I would not have received from a perfectly produced track. You could see people making the music. You could feel their judgement in the performance. A composition I knew became particular to those musicians, that room and that moment.

The human presence was part of what we had come to experience.

It made me think about how much we value the evidence of a person in what we encounter. We hear it in a voice. We recognise it when someone speaks with conviction about something they have lived through. The experience carries more than information or technical accomplishment. It carries the person who made it.

That matters in business too, although we sometimes behave as though professionalism requires us to remove every trace of individuality. We polish presentations until every sentence has the same finish. We soften opinions until nobody could reasonably object. We reach for familiar language because it feels safe, even when familiarity is what makes it forgettable.

In marketing, that creates a particular problem. We spend our working lives trying to establish differentiation, then describe it using language that could belong to almost anyone. We talk about trust, innovation, simplicity and customer focus without explaining what those things mean in the experience of buying from us.

AI can accelerate that habit because it readily produces language we recognise. Give it a broad brief and it can return a broad promise, expressed with considerable confidence. The result may reassure the people approving it while leaving the customer with little reason to pay attention.

The harder work involves making a choice. What do we believe? What do we understand about the customer that changes how we serve them? Which problem are we prepared to name clearly? What can we demonstrate? Where are we willing to be specific enough that someone might disagree?

Those questions require judgement and a degree of exposure. Once you say something precise, it becomes possible to challenge it. That is part of what makes it valuable.

I think this is where imperfection begins to matter. I mean the small irregularities that reveal an individual mind at work: an unexpected observation, a phrase you would actually say aloud, a story that could only have come from your experience.

A piece of writing can be carefully edited and preserve those things. But they are easily lost when every unusual turn is treated as something to smooth away. We can end up removing the qualities that made the work worth reading.

Perfection has become remarkably easy to approximate. Producing something recognisably yours still asks something of you. It requires a point of view, a willingness to make choices and enough confidence to leave your character in the work.

Which brings me to a question I think we should keep asking ourselves.

If AI were suddenly taken away tomorrow, could you still sit down with a blank page and create something original?

Could you put pen to paper, or fingers to a keyboard, and write honestly and authentically about something that matters to you? Could you build an argument, find your language and arrive at a conclusion without first asking a model to show you what it might look like?

I ask myself this because I want to know how much of that capability I am still exercising. Could I tolerate the slow beginning? Could I work through the uncertainty without immediately reaching for a finished paragraph? Would I trust that I had something worth saying before a machine helped me say it?

The question becomes more useful if we return to it regularly. Confidence in our independence can persist long after our habits have changed. I would have said my handwriting was perfectly adequate until someone handed me a pen.

I have no desire to shy away from AI. I want to use it well, with a clearer understanding of what I bring to the exchange. It should help amplify my ingenuity, extend an idea and improve its expression. My own thinking needs to remain active enough to guide that process.

Before I reach for a tool, I am trying to establish the observation, the argument and the reason I care. Sometimes that means a rough paragraph. Sometimes it means talking an idea through or writing notes after a meeting. The form matters less than making the first decisions myself.

Then I can ask for help with a clearer sense of what needs protecting: a particular phrase, a tension that should remain visible, a point that needs strengthening. I can judge the output against an intention I have already formed.

There is a responsibility here for leaders too. If we reward only speed and volume, we should expect people to reach for whatever produces acceptable output fastest. If we value judgement and originality, we need to make room for the work that produces them, and remain interested in the reasoning behind the polished slide.

I left Marks & Spencer thinking I should write by hand more often. A modest ambition, admittedly. But the moment reminded me that ease becomes habit, and habit can quietly change what we are capable of doing for ourselves.

I want to remain involved in the making of my work. To feel the resistance of an idea before it becomes a sentence. To recognise my judgement in what I publish. To leave enough of myself on the page that another person can tell there was someone there.

Perhaps the awkward handwriting did me a favour. It reminded me how easily we can lose touch with something we assumed would always be ours.

And that picking up the pen again is still a choice.

For the record - no LLM was overused in the creation of this article!

Originally published by Paul Johns on LinkedIn.

Comments


bottom of page