AI and the Joy of Creation

For those who’ve known me a while, you probably remember—way back when—I was a developer. I’m admittedly talking way back. And, I like to think I was pretty decent at it, fluent in JavaScript, VBScript, HTML, CSS, Perl, and dabbled in a bit of Java and C++. That was in the very distant past, though. I was experimenting with AJAX before it even had a name, and some of the browsers I used to code for don’t even exist anymore. Eventually, I made the conscious choice to transition out of development and into people management. I don’t regret it, but I do miss coding now and then. There’s something very black and white about it - you build to requirements, or if it's broken, you fix it. Every so often, I dive back in and build something—usually small projects, often database and browser-based—just for the fun of it. Of course, I’m painfully aware that my tools could use a modern upgrade.

When I pick up coding again, the biggest challenge is always the syntax. I still know how to think in programming terms, from object-oriented coding to database design, but getting back into the groove takes a while. And honestly, some basic things can be frustratingly slow to figure out, let alone fix.

So, with all the buzz around generative AI, I was curious: could it actually help me in a practical, hands-on way, beyond the hype and chatter? Here’s what I found.

I’m a big fan of strategy games—think Stellaris or Civilization—and I’ve always been intrigued by the idea of building something like that. At their core, they’re just a lot of interconnected, relatively simple math, all running on an ever-updating database. Will I ever finish this project? Probably not. But for me, it’s more about experimenting and learning, seeing firsthand what’s possible—and what isn’t—with today’s AI tools.

A quick side note: I’m a passionate advocate for AI, though I find myself swinging between support and scepticism depending on how it’s used. The biggest frustration? When I see it used to replace human input instead of enhancing it.

It wasn’t a huge surprise, then, that this bias seeped into my experience with coding.

Generative AI—specifically ChatGPT—gave me an incredible starting point. From reminding me how to set up a basic web server to laying the groundwork for database connections and SQL queries, I had a working database and about a dozen functional pages within a day. That’s not to say I couldn’t have done it on my own, but here’s the important part: I hadn’t touched code in at least six or seven years. AI got me back to a functional level, and fast. From setting up basic style sheets, to AJAX-driven updates, and even time-based event triggers, elements that I would have previously needed serious refreshers in were functional within minutes, and I could focus on the more interesting parts of the solution knowing the basics were well underway.

Next, I pushed the boundaries a bit and tried to generate a 3D star map for the game. I had no idea where to start, but within a couple of hours, I had something workable using Three.js and OrbitalControls. It was exhilarating to see what this collaboration—between my brain and ChatGPT’s vast database—could achieve. It genuinely made me happy to see what was possible (and apologies to those who had to listen to me enthuse about it!).

This whole experience reminded me of the way platforms like MidJourney have lowered the barriers to creating high-quality images. It used to take trained artists years of skill to create something of a certain calibre, but now anyone with an idea can produce something visually impressive. Will AI replace artists? Not for meaningful work—not anytime soon. The difference, I believe, is that human-crafted work reveals more depth the longer you spend with it, while AI-generated art often shows more inconsistencies the deeper you look. But that’s not the point. Setting aside the contentious discussions around copyright, what’s exciting to me is how AI enables anyone to create something that brings them joy. Whether it’s editing photos or generating art to hang in your home, this technology can add value to everyday life.

Now, back to coding. What I’ve realized is that AI makes me at least ten times faster. It allows me to create large blocks of code without major errors and whilst everything needs to be double-checked, in the main, it's pretty accurate. Often what slowed me down previously was simple syntax errors, or using incorrect variable names that I’ve misremembered and being able to avoid what could take hours of painstaking troubleshooting, allowed me to get on with things at a much faster pace. I’ve also been able to update my skills to the latest standards. If you don’t code regularly the reality is that languages move on very fast, and what worked for me back in the early 2000s would generate significant issues now, as syntax gets deprecated and updating your skills from a historical point generally involves trawling through reams of content online, and trying to avoid helpful suggestions that are themselves now out of date. I was also able to troubleshoot my ever-growing codebase of thousands of lines more efficiently. By simply cutting and pasting blocks where there were issues, and with some judicious console logging I found that whilst genAI doesn’t always have the solution, and can sometimes take you down some very odd rabbit holes, it generally gets me several steps closer, very fast.?

Most importantly, it takes the grunt work out of coding, leaving me free to focus on the parts where I add the most value—design, balance, and interaction. Could ChatGPT have created this project alone? Absolutely not. And when I review the inefficient, sometimes error-ridden outputs it gave me, I’m well aware of how much steering I had to do. But without AI, this project would’ve taken weeks, and the learning curve might have been steep enough to deter me entirely.

Ultimately, this is a hobby for me—something to keep my mind sharp and help me understand the challenges today’s developers face. It keeps me relevant and gives me a better understanding of what I’m asking my teams to do when we tackle client projects. And yes, it’s fun. The enjoyment here is in the process of creation, in the building of an idea from nothing. Whether anyone else sees the final product or not, it’s like solving a crossword puzzle—the joy is in that journey. And thanks to AI, that joy is amplified tenfold.