Rip the head off structure.

// The Alarm

I was halfway through sharing my screen when I felt a little ashamed - they were nodding, I think that means that they were a little impressed.

I was walking them through the way I use ChatGPT - the frameworks, the prompt structures, the custom GPTs, the systems I’ve built to make thinking cleaner and output faster - all in special chats, projects and neatly organised.

Hook, different angles, expansions of topics and final revisions - it was all a little efficient and replicable - it was all very smart.

And as I was explaining it, I had this strange thought - I sound like a machine describing how to become one - the complete antithesis of my philosophy - this writing is human, abstract and wanders quite often (like now, I've started in the middle of the story - done on purpose to show what I mean)

Last week was an online session ran by Stewart Perrett where he wanted to get feedback on how he uses Chat GPT and then hear from others on how they use it - Stew stepped up first and was showing how he uses the AI agent to write some SEO laden blogs for his clients to get their first steps on the 'content ladder'.

I could see that he wasn't using projects and he hadn't started any custom GPT development - he was just using ChatGPT by asking in individual chats and building up his history in various conversations.

That's ok - we all use tools in their own way and if it completes the job in the way we want it, then why not use it in that way.

I mean, I watch countless Youtube videos on how to edit videos, what workflows look like for creators and I pick up little hints and tips here and there - but that doesn't stop me from editing in different ways, even this weekend, I've learned another new way to edit videos and I'll slowly get faster and faster as they are produced.

// The Framework

I was happy to show off my way of working in ChatGPT - I have everything looking good and very well organised - I think it's important at this point to point out that this blog is NEVER written by ChatGPT - I don't use it to write anything, I have a project in my ChatGPT called 'The Clique' and the only thing it's used for is the rubbish little tasks after the blog is written, I want email subject lines, preview text and anything else which I hate doing.

Anyway, I digress (again) - my GPT is laid out wonderfully and I have been working on a system to create a webinar structure for Yammayap - the landing page copy, the emails associated with the launch and social posts - everything is systemised and everything hangs together in a framework.

I want to be clear - I don't think there is anything wrong with having a framework and a structure and I don't ever want to use the 'content' produced by ChatGPT as an output, everything has to be 'humanised' but it does give you a baseline to work from, especially if you're creating a stable and repeatable model.

But, as I was explaining this - I felt two things.

The first was that I was slightly enslaved to the structure as the structure was doing it's job - frameworks and structured outcomes feel safe - they control the output and when you build them well enough, you can build an identity on them.

A repeatable voice. A reliable outcome.

But somewhere in that session - somewhere between showing the logic and explaining the layers - I realised something uncomfortable - I felt like it was also editing me.

The pauses, the messy starts, the “I’m not sure yet." - essentially, the human bits.

Structure makes you safer but it also limits you - It makes you a smaller piece of the puzzle.

The second part was that the framework (and the framework that AI has created) gives us comfort.

// Comfort

When you're creating - businesses want to have the safe boundaries to operate within - it's why people talk like corporate robots on Linkedin - they are all talking the same and operating under the same 'safe' corporate language which doesn't take risks.

Having a repeatable, stable and consistent output from ChatGPT feels safe - if you're creating blogs, social posts or just using it to respond to emails, it's trained to not be offensive. It won't jeopardise any client relationships, it won't say anything which will cause you to offend anyone and no-one will feel aggrieved.

You stay in the middle lane if you use ChatGPT.

There is comfort in that - my mum (when she drove) would drive consistently in the middle lane because she didn't really have to do anything else and you don't have to deal with junctions, overtakes or tailgaters - it's safe.

That's what I felt last week - it was like warm blanket knowing that the output of the CustomGPT's I'd created could replicate the same output over and over again. But that was also at odds to what I like to do with my social content.

There is a place for comfort - there is safety in knowing that you're getting something which won't upset or disrupt because you can generate more and more content - it's repeatable.

But why? Why not be different? Why don't we talk like humans and avoid that corporate jargon?

// Rip the head off.

Here’s what’s been sitting with me since that session online with the You Are the Media team - the outputs were good.

That’s the uncomfortable part - They were clear, structured, commercially sound - you could definitely build campaigns on them. You could even build a business on them because there was no chaos, no rambling, no loose ends.

Just clean thinking packaged well.

If the framework is doing that much of the work - if it’s shaping the flow, controlling the tone, smoothing the argument - what exactly is left of me? Structure doesn’t just organise ideas, it edits them. It trims the hesitation and it removes the “I’m not sure about this bit.”

It rounds off the sharp edges before they’ve had a chance to cut through and neat just doesn't hit the same (in my opinion)

ChatGPT didn't unsettle me - It wasn’t the technology, it was how easily I could see myself leaning on it and getting 'what I needed quickly' - it's weird to see how quickly structure becomes a crutch and how tempting it is to let the framework carry the weight when you’re tired, busy, or trying to be consistent.

How many times have you thought - oh go on then, I'll quickly use ChatGPT for that.

You start by using it to create a baseline, then you use it to save time, then you use it because it works. And before long, you’re shaping your thoughts to fit the structure instead of shaping the structure around your thoughts - you become the middle lane driver.

That’s the danger.

Structured outputs don’t wobble, they don’t hesitate.. They stay in the middle lane.

Safe. Competent. Forgettable.

And I don’t want to be forgettable.

So maybe the answer isn’t rejecting structure. I’m not anti-system, I help our team build systems for a living and I believe in clarity. I'd even go so far to say that I believe in frameworks - I said that earlier, I believe in repeatable thinking when it serves a purpose but I don’t want the framework to become my identity.

Sometimes the work is taking the polished output and deliberately breaking it - swerving into the fast lane. because swerving is human.

If everything I produce is safe, scalable and structurally perfect, then I’ve become the template I claim to resist and my mind craves the mess.

The doubt.
The wandering.
The slightly uncomfortable line.

Because that’s where the voice lives.

And if I ever lose that, what exactly am I building?