🎄 Merry Christmas Eve, hackers! 🎄
While everyone is unwrapping gifts, I’m sitting here deploying code. I’m an indie dev with a crazy hypothesis:
In the near future, publishing a video game should be as low-friction as posting a Tweet or a TikTok.
You shouldn't need Unity or C# to share a playable idea.
You should just be able to “describe” it or “show” it.
I built GamiCool.com to test this idea.
It’s an AI-native game platform powered by Google Gemini 3.
Currently, it supports three modes:
Text-to-Game
Prompt it like ChatGPT, get a game.
Image-to-Game
Upload a doodle or sketch, it generates the terrain.
Video-to-Game
Upload a clip, it extracts the physics logic.
To be honest, a month ago, I was skeptical if AI could handle real-time game physics.
But after weeks of fine-tuning Gemini 3, it finally clicked.
The moment your static sketch turns into a playable character that jumps and interacts with the terrain…
it feels like pure magic ✨
It’s not just a toy anymore; the gameplay is genuinely satisfying and fluid.
Since the core engine is now solid, I’m going all-in on the “TikTok-style” features next
(Social & Remixing).
I want you to be the first to test the limits of this engine at gamicool.com.
Please try the following:
🎨 Try the Image-to-Game
Draw something on a napkin, upload it, and see if the AI understands your level design.
🔥 Roast the UX
Is the generation flow intuitive enough for non-gamers?
🚀 The Viral Test
If you made a game, would you actually share the link with a friend?
Can’t wait to see what you build.
Happy holidays! 🍻