How I Learned to Stop Worrying and Love AI — Experiment 1
If you’ve ever thought to yourself, "Wouldn't it be cool if there was an app that did..." and then shrugged that thought off wistfully because you assume (rightfully) that building an app or website requires technical knowledge, funding from investors, and a team of developers — none of which you have at your disposal, then you know exactly how I felt at the beginning of 2024. But then, last year, armed with nothing but unbridled curiosity and an inability to write code, I stumbled into the world of Large Language Models. What I discovered was transformational: I could actually build things. Proper things. Digital things that work!
For the past year, I've been lurking in the shadows of various AI beta launches, fumbling my way through technologies I barely understood yet somehow figuring them out as I went. It’s been exhilarating and terrifying, like being handed the keys to a Formula 1 car after having only ever driven bumper cars at the fair. I was there when DALL-E first opened its digital doors, when Claude was still learning to count properly, and when MidJourney was still trying to figure out how many fingers humans have.
—
The One Where AI Does The Prompting
My first proper venture into the brave new world of “AI-assisted development” was born from a rather specific domestic challenge: my wife's creative block. Now, for creative professionals like her and I, a creative block isn't just an inconvenience — it's a barrier to actual work. When your livelihood depends on consistently producing creative work, these moments of creative paralysis can be absolutely nerve-wracking (which, ironically, only makes the creative block worse).
I'd read that sometimes the best way to break through a creative block is to create something small, quick, and entirely inconsequential. It's the creative equivalent of putting on proper trousers when working from home — a small action that somehow tricks your brain into the right mode.
Despite having virtually zero coding proficiency, I thought: What if we could get AI to help with this? But here's where it got interesting — instead of having humans prompt the AI (which was all the rage at the time), what if we flipped the script and had the AI prompt *us*?
Enter the Creative Block Buster — my first GenAI experiment, built more with enthusiasm than actual expertise. Think of it as a creative personal trainer, except instead of counting your reps, it's counting your minutes of creative writing. It throws a writing prompt at you (generated by AI — unique and random and very fun), starts a 5-minute timer, and off you go.
The beautiful thing about it is its complete lack of consequence. There's no pressure to create something brilliant or meaningful. It's just you, responding to an often slightly absurd prompt, for a couple of minutes, kickstarting your creative flow.
And surprisingly (or perhaps unsurprisingly, if you understand the psychology of creative blocks), it worked. My wife and I started using it whenever we felt stuck, and gradually, these little five-minute creative sprints began unlocking bigger ideas.
This little project marked the beginning of what would become a year-long adventure in building with AI. And I’ve actually learned so much about development now. From helping non-sports types fake their way through sporty conversations to creating an app that recommends movies based on your current emotional state, each project has been a delightful experiment in seeing what's possible when you combine human creativity with artificial intelligence — and a healthy dose of a "how hard can it be?" attitude (spoiler alert — it’s often very hard and very frustrating!)
Given the charged atmosphere around “AI” in the creative industry today, I should mention that these are entirely personal experiments — little weekend projects born from curiosity rather than commerce. No client work was harmed in the making of these experiments, and I haven’t “monetised” any of them.
If you’re curious, you can try out the Creative Block Buster yourself at this link.
Fancy a chat about how I managed to build these things despite my coding skills being rubbish? Drop me a message — I'd love to share my process and perhaps convince you that you, too, can build silly things with serious technology.
And stay tuned — I'll be sharing more about my other AI experiments soon, including an app that helps you pretend to understand cricket. Because sometimes the best innovations come from solving very specific, slightly ridiculous problems.
