Sleepy Cat Entertainment

📝 A personal note

I find myself at a crossroads in my professional career. Having been in the gaming industry for nearly a decade, I have settled into somewhat of a comfortable mediocrity. It’s not that I’ve stagnated in my skill or even my passion. These past few years in particular I’ve grown as an technical developer, code architect and have even started to feel comfortable managing and coaching.. With the introduction of AI tools I feel my potential expanding. I find that I am able to overcome larger challenges than ever. While my interests have expanded beyond the realm of gaming, I still feel a deep love and connection to it. With all of that being said, I do feel unfulfilled in my chosen area of expertise.

When I was a young boy I remember dreaming of the epic adventures I would create. I made countless projects in Scratch, simply pouring my creative soul into the multi-colored logic blocks and throwing together my best MS Paint art assets. As I grew older, entering college, I simply wanted to make something interesting and unique. As my understanding of game development increased I envisioned MMOs and massive space exploration simulators. Some of those projects became something, a Unity game folder on some hard drive lost to my boxes of cords and old computer parts. Some live in git repositories I may dust off and share some day.

On one hand they say to “do what you love and you’ll never work a day in your life.” But on the other, we all know the risk of working in your area of passion. Sure, your work becomes an expression of your creative self. But your passion also becomes work. A story many can relate to, I’m sure, nothing profound to be found here. But this is the situation I find myself in nonetheless. So, what do we do about it? Well, let’s start by creating for creation’s sake, and see where our soul takes us from there.

📝 Update Notes: Project Wisp

Project Wisp is a game idea spawned from several of my gaming interests. The primary inspiration was the Pokémon games. (Perhaps I’ll have to write a personal note about gaming patents someday.) Yes, I know how generic that sounds, but I have always been in love with the complicated systems of stats and creature development that have been refined over the years in those games. Furthermore, I was enamored by this concept of raid battles in a JRPG setting. Now on this front, I don’t think anyone has done such a system justice, not that I have seen. The other thing I have been drawn to over the years is idle games. The mobile industry thrives with idle mechanics. Even your premiere titles have cooldown or accumulation mechanics intended to allow players to go about their days and return while still having made progress. This can be used as a cudgel, forcing spend to skip arbitrary wait times, but I also think it can be a low-pressure positive feedback loop. I dream of the perfect mix of ‘always having something to do’ and the low pressure of idle mechanics.

I built a rough initial version of Project Wisp in a few spur of the moment furious coding sessions last year. This initial work included a main screen with a ‘Generate’ button and a roster screen. The wisps had some basic stats and even generated with a ‘quality’ which would introduce Step one was resurrecting the project from this state and starting on some of the other core functionality.

Once I was acclimated to the project, I made a few broad updates and project QoL changes. I added a new script which refreshes the assets on play, and disabled auto-refresh since it was distracting to see Unity constantly refreshing on my other monitor. I also updated the whole UI (bare as it was) to fit portrait mode on iphones. I figure the sooner I commit to a UI scale, the better, since that would be a nightmare refactor in the future.

Then it was on to our new functionality. The system I’m starting with is the ‘ritual’ system. The way you will generate new wisps will be to perform a ritual, using reagents, glyphs, and even wisp helpers in order to summon new creatures. Rather than a debug button to generate a wisp, the game now uses this ritual ‘site’ which is a button with a few states. You kick off a ritual and after a cooldown, using timestamps at the point of the button press, you can collect your reward of a brand new wisp. This system already flows so much better and gives a lot more impact to the experience of collecting creatures.

I see AI in development as an unavoidable reality at the moment. I do understand the moral objections to the use of the technology in general. I do hope for regulations that help limit the negative impacts of it on the environment and society as a whole. But as a developer, the reality is that I would be hamstringing myself by avoiding the tools altogether. For my own personal finances/credit usage, and because I want to stay sharp, I will still write much of the code myself. My architecture will be driven by my own understanding of best practices and my own creativity. However, some problems can be tedious, test-writing is still new to me, and AI is already faster and better than I will ever be at certain tasks. One line that I won’t cross is with AI generated art. I have too many artist friends who I could commission and collaborate with to justify having AI generate something soulless.

So with that being said, I am starting to experiment with AI coding companions for this project. I already had a subscription to ChatGPT so I will be starting with Codex. So far it has helped create a wisp info popup using the existing AI as a baseline. Part of using AI will also be seeing just what it can do. I never want to assume it has an arbitrary limitation, and will likely ask it to take on some large challenges, which I will then review the solutions to and refine into workable game architecture. The dangers of vibe-coding is real and the last thing I want to do is build a house of cards that myself and the agents won’t have a clue how to touch after a few sessions.