Automation Didn't Start With AI
When people talk about automation today, the conversation usually revolves around AI, APIs and modern software tools.
For me, though, automation has always been part of the way I work.
Long before ChatGPT existed, and before many websites even offered APIs, I was looking for ways to make computers do repetitive tasks instead of doing them myself.
The technology has changed dramatically over the years.
The mindset hasn't.
Spotting an Opportunity on CafePress
One of the earliest examples came while my cousin and I were building products for CafePress.
He handled the design work.
I handled the development.
At the time, personalised products were becoming increasingly popular. The problem was that creating them was painfully slow.
Every personalised product had to be created individually, and there was no simple way to generate thousands of variations automatically.
It was obvious that the manual approach wasn't going to scale.
Building Our Own Workflow
Since the platform didn't offer the tools we needed, we built our own.
The designs were converted into SVG files, allowing names to be inserted automatically without redesigning every product from scratch.
Alongside that, I built a database containing thousands of names.
The software combined the designs with every name, generating all of the personalised variations automatically.
Instead of creating products one at a time, the system created them for us.
When There Was No API, We Built Around It
There was still one major problem.
CafePress didn't provide an upload API.
Even though we could generate the products automatically, we still had no automated way of getting them onto the website.
So we found another solution.
I used Firefox macros to control the browser itself.
The software opened pages, filled in forms, uploaded images and repeated the process over and over again.
It wasn't elegant by today's standards, but it worked.
Sometimes automation isn't about having the perfect tool.
It's about finding a way around the limitation that's in front of you.
Building Scale Instead of Repetition
The result was that we weren't uploading dozens of products.
We were uploading thousands.
Every name represented another product.
Every product represented another opportunity for someone to discover us.
The system kept expanding our catalogue without requiring us to repeat the same manual work over and over again.
That was the real advantage.
We weren't working harder than everyone else.
We were building systems that worked on our behalf.
The Technology Has Changed. The Thinking Hasn't.
Looking back now, I realise how little my approach has actually changed.
Today I use AI to automate very different kinds of work.
I use it to research products, generate ideas, write content, create images, help produce videos and speed up software development.
The tools are completely different from the ones I had back then.
But the principle is exactly the same.
Whenever I find myself repeating a task, my first instinct is still to ask whether the computer could do it instead.
Every New Technology Creates More Leverage
One thing I've noticed throughout my career is that every generation of technology creates new opportunities for people willing to embrace it.
Years ago it was browser macros, databases and custom-built workflows.
Today it's AI.
Tomorrow it'll be something else.
The specific technology almost doesn't matter.
What matters is recognising when a repetitive task can become a system.
That's where the real leverage comes from.
What I'm Focusing on Next
As AI continues to evolve, I find myself approaching it in exactly the same way I approached those early automation projects.
Not as a replacement for creativity, but as another tool for removing repetitive work.
The platforms, programming languages and technologies may keep changing, but I'm still chasing the same goal I've always had: building systems that do more than I ever could on my own.