The idea of this project is to explore Artificial Intelligence based on Causal Entropy.
The basic idea is: Intelligence seeks to maximize its future freedom of action.
This project is about different angles towards that approach.
In what ways can intelligence seek to maximize its future accessible entropy? What does this even mean: Baysian Networks? Monte Carlo simulations? Multiple angles could be interesting.
Maybe it makes sense to have a system with multiple intelligence agents running around?
Lets see where it goes!
What about a system of Ants, where each has a Baysian Network that can observe within their field of vision?
Each Ant comes with insticts? How does it know to gather food otherwise?
A
Ant properties:
- Ants move on a grid
- An Ant leaves footprints?
The project uses Korge, a Kotlin game engine for rendering.