Why I'm co-writing with machines

It was some time ago, maybe a year ago, when I realised that I no longer use AI. I work with it.

That shift sounds subtle, but it changed everything — how I write, how I think, even how I judge the originality of a sentence. The more I collaborated with these models, the more I began to see them not as tools, but as co-authors in an unfolding experiment: an exploration of what happens when human intention meets machine probability.

Subscribe to Cowritten.ai — where humans and machines learn to create together.

Most writing about AI sits at the extremes. It’s either apocalyptic (machines replacing us) or evangelical (machines liberating us). The truth is quieter and stranger: we’re entering a period of co-authorship. How we learn to write, think, and create in that space will shape not just our work, but our sense of self.

The myth of automation

When people talk about “AI writing,” they often imagine automation. Faster output, fewer humans. But the real story, at least for me, has been the opposite. Working with language models slowed me down. It forced me to articulate what I actually meant before asking the machine to extend it.

In the same way a good editor or mentor does, the model mirrors my intent. Sometimes clumsily, sometimes insightfully. And that reflection helps me clarify my own thinking. The machine drafts; I interpret. I prompt; it surprises. Somewhere in that oscillation, a rhythm appears that feels less like automation and more like conversation.

Writing with a mind that isn’t yours

Every collaboration reveals a certain asymmetry. The model doesn’t know anything, yet it has read almost everything. It produces meaning statistically, not experientially. Still, when I read back a paragraph it wrote — shaped by my intent but not quite mine — I feel something close to recognition.

The systems behind the scenes

In my day-to-day work, co-writing extends beyond a single model.

I build and orchestrate systems that chain multiple large language models together — one model researches, another generates, yet another critiques, one more refines tone and brand voice.

These automated workflows aren’t designed to remove humans; they exist to preserve consistency and depth at scale. They help maintain brand identity across thousands of pieces of content while leaving space for craft and human editing.

Sometimes I think of these chains as miniature creative ecosystems. Conversations happening between models, each tuned to play a different role, while I conduct from above. That’s where things become philosophical again: you’re not simply writing with a model, but rather engineering the conditions under which meaning emerges.

Beyond text

But Cowritten.ai isn’t only about words.

Generative AI now stretches across every creative domain (images, video, sound, and motion). I sometimes use video models to prototype campaigns, image systems to visualise brand worlds, and voice synthesis to test how tone changes emotion.

These tools expand the sketching process. They allow you to explore creative territory that used to require an entire team or studio. For those of us working in digital marketing or content strategy, they reshape what it means to iterate. We can now test, combine, and rethink ideas faster than we can brief them.

So part of this publication will look at those emerging workflows: how we experiment responsibly with these new media, how orchestration changes creative direction, and what craftsmanship looks like when part of it runs on silicon.

What Cowritten.ai is

This publication is a record of that experiment. It’s where I’ll explore what it means to co-create with machines — as a (co)writer, editor, and strategist designing systems that blend human creativity with algorithmic precision.

Expect essays about:

how AI alters the craft of writing and thinking,

how image and video generation reshape creative work,

what “originality” means when intelligence is distributed across systems,

how orchestration and automation can strengthen, rather than dilute, creative identity,

and how our sense of authorship shifts when our collaborators are models, not colleagues.

Some posts will be reflective, others practical. All are written from inside the transformation, not from the sidelines.

The human in the loop

If you’re reading this, you’re probably already experimenting with AI. Maybe for notes, brainstorming, or to test an idea against something that answers back. You’ve felt the friction of trying to stay human in a loop that keeps getting faster.

My hope is that Cowritten.ai becomes a place to slow that loop down, notice the small shifts in our relationship with these systems, and ask better questions.

Not “Will AI replace writers or designers?” but “What kind of creative do you become when you no longer create alone?”

Thanks for reading Cowritten.ai — subscribe for more experiments in human–machine creativity.