๐ฎ Building games & apps with AI
๐ Three.js โข Creative Coding โข Generative Art
๐งช Experimenting, breaking things & shipping
โ Everything I'm building
@matthias_mrc yeah, doing that, sometimes there are interdependent functions which are called out from other parts of code. so making changes disturb all. Then i ask to first make sure to do independent and then make changes.
I played one of my old prototypes today and noticed something embarrassing.
The controls were fine.
The graphics were fine.
The mechanic worked.
But I still didn't enjoy playing it.
Then I removed about half the UI and suddenly it felt better.
GameDev is weird like that.
Sometimes the problem isn't what's missing.
It's what's getting in the way. ๐ฎ
#GameDev #GameDesign #IndieGameDev
Vibe coding makes building a game feel almost too easy. ๐
You get an idea โ describe it โ AI writes the mechanic โ you have something playable.
But I've noticed a strange problem...
The faster I can build features, the easier it is to build things I probably shouldn't.
I'm starting to spend more time asking "should this exist?" before asking AI "how do I build this?"
#VibeCoding #GameDev #IndieGameDev
Built a generative art piece with WebGL & GLSL shaders:
4,000+ colored light beams race through concentric square tracks in alternating directionsโand math guarantees they never touch or cross. โก๏ธ
The entire animation is computed directly on the GPU using pure math (Chebyshev square distance). Silky smooth 60+ FPS!
#creativecoding #shaders #webgl #glsl #generativeart #javascript
Things rarely come out exactly the way you imagined.
The game doesn't look right.
The idea feels different.
The code gets messy.
The first version isn't what you pictured.
I've had plenty of moments like that lately.
But I think the answer isn't to stop.
Build it.
See what's wrong.
Fix one thing.
Try again.
Sometimes you don't reach the version you imagined.
You reach a better one by working through the version you have. ๐ฎ
#GameDev #IndieDev #BuildInPublic
Honestly, probably scope creep too ๐
I start with a pretty simple prototype, then I keep thinking about coins, progression, more levels, better UI, polish...
The AI makes each addition feel so cheap that it's easy to forget every new feature still needs to be balanced, tested and supported.
I'm trying to get better at stopping myself before the "small game" turns into a completely different project ๐
If you could give ONE warning to someone building their first game with AI...
what would it be?
Not a coding tip.
Not which engine to use.
I mean the mistake that can quietly waste weeks:
scope, architecture, assets, polishing, features, monetization, publishing... anything.
I'm collecting the answers because I'd rather learn from other GameDevs than repeat every mistake myself. ๐
#GameDev #IndieGameDev #VibeCoding
New GameDev + vibe coding question:
How early should I think about monetization?
Ads?
Rewarded ads?
IAP?
Cosmetics?
Premium?
No monetization at all during prototype?
I don't want to destroy a good gameplay loop just because I added monetization too early.
What would you avoid as a beginner?
#GameDev #MobileGameDev #IndieGameDev
I think one of the easiest mistakes for a new GameDev is skipping real playtesting.
You build it.
AI says everything looks correct.
You play it 20 times.
Then someone else touches it for 30 seconds and immediately gets confused. ๐
What did your first real playtest teach you that development couldn't?
#GameDev #GameDesign #IndieDev
For new GameDev developers using AI:
Did you ever optimize too early?
I'm seeing all these things I *could* optimize:
draw calls
memory
object pooling
asset sizes
physics
loading
But my game is still basically a prototype. ๐
At what point did you start caring seriously about performance?
#GameDev #IndieGameDev #ThreeJS
One thing I'm starting to worry about with vibe coding games:
AI is very good at adding features.
But every feature creates more things to test.
One button becomes:
UI โ state โ save data โ edge cases โ mobile input โ sounds โ animations...
What mistakes did you make by saying:
"Let's just add this one thing"? ๐
#GameDev #VibeCoding #GameDesign
Question for GameDev people:
How much time should a beginner spend finding or creating art before the gameplay is proven?
AI makes it ridiculously easy to generate:
characters, backgrounds, icons, particles, UI...
and that's honestly tempting me to polish things way too early. ๐
Would you prototype with ugly placeholder assets for as long as possible?
#GameDev #IndieDev #GameArt
I'm learning that game development has a dangerous trap:
A prototype can look like a game long before it actually IS a game.
You can have:
๐ฎ player
๐จ nice graphics
๐ฑ UI
๐ฐ coins
โญ progression
and still have a boring 2-minute experience.
For new GameDev + vibe coding people:
What mistake helped you realize this?
#GameDev #GameDesign #IndieGameDev
GameDev + vibe coding question:
How early should I start caring about game architecture?
AI makes it very easy to keep adding code until the prototype works...
but then changing one thing starts breaking five other things. ๐
Would you recommend keeping the first prototype intentionally messy and refactoring later, or setting up clean systems from day one?
#GameDev #VibeCoding #IndieGameDev
One mistake I'm trying hard to avoid as a new GameDev using AI:
Building too much before actually playing the game.
It's easy to ask AI for:
โ menus
โ progression
โ coins
โ settings
โ achievements
while the core mechanic still isn't fun.
What would you tell a beginner to build FIRST?
#GameDev #GameDesign #IndieDev