Autonomous worlds built with Dojo with players & characters powered by on-chain AI
The project attempted to develop an on-chain ai powered game using Dojoengine autonomous worlds. The aim of the app was to modify the react-phaser-app example from Dojoengine, to allow for ai powered decision from external systems via contracts deployed on Starknet.
The primary aim is to create moves based on suggestions provided by the machine learning model.
The core technologies that were used while developing the ai powered on chain game are
Dojo - The game developed on Dojoengine. Once a world is selected. The smart contract fires an event. The event listener in javascript then forwards the state of the world to the model. The deployed model then suggest the next move. The model under development is in SKlearn Sklearn for creating a model for making moves in the react-phaser-app game.