About the project
We started with an idea and no idea how to build it.
Industrial Excitement is an independent game-development project built around one question: what happens when AI-generated content meets intentional hand-crafted design? We're learning by doing — building atmospheric, systems-driven games and releasing them free while we figure out what works.
Where this began
No prior experience. On purpose.
There was no studio, no game-design background, and no development experience — just an idea and a stubborn curiosity about whether modern AI tools could close the gap between "I want to make this" and "I can actually make this."
So the project became the classroom. Every system, every mechanic, every mistake is part of learning the craft in public. We'd rather show the seams than pretend we arrived knowing what we were doing.
The pipeline
AI-assisted, hand-directed.
The two halves aren't in competition. AI gives us speed and volume; hands give it intention. Here's how a piece of a game actually comes together, step by step.
Idea
A feeling, a place, a question. We name what we want the player to feel before we touch a tool.
AI generation
We generate environments, assets, and narrative fragments quickly and in bulk. It's raw material, not a finished thing.
Manual design
By hand, we build mechanics and level flow, then curate the generated pieces — keeping what serves the idea, cutting the rest.
Design documents
What works gets written down. Rules and systems become a foundation the next idea can build on.
Iteration
Play, break, change, repeat. The real game is found here, in the slow unglamorous middle.
User feedback
You play it free and tell us the truth. That signal decides what we keep, fix, or throw away.
The public mistake log
Five things we got wrong.
Learning in the open means owning the misses. These are real lessons from building Escape — kept here because they were useful, and because hiding them would defeat the point.
Early on we let procedurally generated layouts ship straight into the game. They looked fine and played terribly — dead ends, unreachable ledges, no rhythm. Generation now produces candidates; a human approves every one.
We loved the dark, smoky look so much that players couldn't see where to go. Mood is worthless if the player is lost. We added a guiding light source and pulled the darkness back.
We spent weeks on inventory and stats while the core jump still felt mushy. Now the rule is: one small thing has to be fun before anything gets built on top of it.
A tiny editor bug — the music play button registering only on the second click — ate days. The lesson wasn't about the bug; it was about how long we let "minor" annoyances sit before fixing them.
We polished in private for too long, convinced we knew what was good. The first outside playtest corrected more in an hour than we had in a month. Now feedback comes early, and it's free for anyone to give.
What we hold to
Values, kept short.
Play free. Tell us what you think.
The whole project runs on honest reactions. If you play something, we'd love to know how it went.