Writing with AI – Is It Still My Voice?
AI is a seductive tool. It’s helping most of us work more efficiently with copy, marketing, ideas, and research. These tools can now produce entire blog posts in seconds. They can mimic tone if you train them (though not quite, right?) they’ll sort out structure and generate content.
I get why the time-saving aspect is so appealing. But I think if you’re saving too much time, you’re probably not putting enough of you in. And the result will likely feel a little hollow.
I’ve been experimenting. The more I use ChatGPT, the more I realise how much I still have to write and re-write to make something truly mine. I want to be able to talk about what I post, stand by the ideas, and know I’ve shaped the words. I don’t always get the balance right, but I know I could sit with a reader and discuss anything I’ve written with passion and some knowledge. This piece, for instance, took me a reassuringly long time to get the tone, the shape, and the conversational feel just right.
Where’s the Line?
If I use AI to help me write, is it still my writing? When can I use it and still feel in integrity?
To me, writing with AI can be thoughtful and intentional or it can easily slide into mindless copying and pasting. That’s the grey area where a lot of people are quietly outsourcing their thinking and calling it creative work. I’m still figuring out how and when to acknowledge that help, and how to stay transparent.
How I Use AI
Here’s what works for me:
-
I’ll ask AI for a rough structure, based on a topic I’ve already thought about.
-
I might include certain research or angles I want in the mix.
-
Then I copy the draft into Word and start shaping it properly.
This process helps me get started, especially because I procrastinate. But I still put the time in. I don’t let AI think for me, and I don’t publish anything I couldn’t explain in a real conversation.
If I read an opinion piece, I want to know the author could speak about it beyond the screen that they’ve thought about it, lived it a little, and can expand beyond what they’ve typed. Otherwise, what are we doing? At worst, it’s not just lazy & it’s dishonest.
You Can Tell When It’s Not Real
The following section? Is AI-generated. I left it as-is, so you can hear the shift in tone:
People underestimate readers. We know when something feels authored vs assembled.
– The very short sentences. For impact. You know the ones.
– The same voice again and again: polished, generic, eager-to-please
– Phrases like “Let’s break it down”
– No specifics. No detail. Just vibes.
There’s no depth. No lived experience. No voice. It’s functional but soulless. I quite enjoy the irony of AI warning us about itself using the very voice that gives it away.
That’s not the writing I want to publish.
My Process (When It Works)
-
I start with an idea or story I care about.
-
I tell AI roughly what I’m aiming for, include any relevant detail or points.
-
I get a structure, copy it into Word, and begin my own writing from there.
-
I reference my own learning, reading, courses, books.. and if I quote something I haven’t fully read, I acknowledge that.
Even with help from AI, it still takes me hours to shape something that feels honest and thought through.
What About Short-Form Content?
Not everything I publish takes hours. Some updates, newsletters or event blurbs are quick. Sometimes they’re AI-assisted, or I’ll co-create them with Emma. They’re functional brand-awareness, reminders, marketing and they serve a different purpose. They’re still me, but not trying to be personal essays.
