How My Conversations With ChatGPT Have Evolved (Cute Cat Pic Inside)
- Caroline Warnes

- Nov 20
- 3 min read
Every now and then I run little experiments with ChatGPT to see how well it understands me. Not in a mystical way, but in an editorial-maturity way. And because I am a bit of a geek at heart.
Today I asked a blank chat window: “How have my conversations with you evolved over time?”
Its response was actually pretty cool and even a bit of a trip down memory lane (ah, late 2022, how far away you seem now). It captured the arc of how I now use AI across my work and life. And honestly, the evolution resembles what I see in a lot of organisations as they move from dabbling in AI to embedding it properly.
Here’s the condensed version of what it told me.
From quick tasks → to actual workflows
I started with copy rewrites and tidy-ups. Now it’s structured editorial systems, product builds, content diagnostics and multi-step processes. The tool has become an engine room rather than a paragraph polisher.

From business-only → to human reality
As the trust built, my chats moved outside work and into the real stuff: parenting, neurodivergence (sometimes you've just got to ask ChatGPT to interpret that person's facial expression for you), values, life logistics and, critically, pets.
The details of my more personal chats don't need to be revealed here, but my cat Bowie is an exception, because everyone loves him.
ChatGPT has solved all sorts of cat-related conundrums for me over the years, like this classic: "Does Bowie really think I can't see him when his tail fluff is hanging around the door?". (ChatGPT's answer: he knew.)
From “fix this draft” → to “build this with me”
The biggest shift for me has been using AI to help me develop the OGC Content Store, tighten product positioning and shape entire logic systems. It’s become a co-architect, rather than a supporting tool.
From single threads → to multi-layered thinking
Now our conversations run across investing and options trading, narrative design, client strategy, product development, personal processing and the occasional feline chaos. It’s a thinking partner across work and life.
From simple prompts → to an editorial operating system
Tone rules. Cadence. Banned words. AI-tell removal. LinkedIn voice. Brand logic. Over time I’ve essentially trained a second version of my editorial brain - critically, one that doesn’t rely on those distinctly human weaknesses such as lack of caffeine or sleep.
From support → to strategic sparring
These days, I rarely bring a fully formed idea. I bring the messy version, the half-thought or the dilemma. ChatGPT helps me shape it, test it and work out what actually matters. It's less “do this” and more “think this through with me.” (And sometimes, "help me interpret the tone of this email" - see earlier point about neurodivergence.)
And here's the real point
For me, the real evolution has been using ChatGPT as supporting infrastructure behind a very human brain - one that knows my clients, my products and, apparently, my cat.
If you're still stuck at the “rewrite this paragraph” stage with AI, it might be time to rethink how you're working with it. The more context, boundaries and intent you bring, the more useful it becomes.
That's part of the work we do best at Only Good Content: helping teams move from dabbling with AI to building content systems that actually support how they think, write and decide. Bowie’s involvement is optional, but highly recommended.


