In the paper Computation at the edge of chaos, Chris Langton was able to find interesting (turing complete) patterns by varying the distribution of state transition to a specific "root" state.
Inspired by that, I tried something similar once, with interesting results: https://imgur.com/gallery/sRUrI. Maybe you could try it as well - more variation, and competitive selection. Core Wars on steroids :)
It could also be interesting to find out what happens if the machines only have partial transition tables (and can cooperate/sabotage each other). Also: an energy cost.
There was an ios game called survivium a while ago, where players could program the things themselves: https://youtu.be/0WOJuZIlVfY?t=70
In the paper Computation at the edge of chaos, Chris Langton was able to find interesting (turing complete) patterns by varying the distribution of state transition to a specific "root" state.
Inspired by that, I tried something similar once, with interesting results: https://imgur.com/gallery/sRUrI. Maybe you could try it as well - more variation, and competitive selection. Core Wars on steroids :)
It could also be interesting to find out what happens if the machines only have partial transition tables (and can cooperate/sabotage each other). Also: an energy cost.
There was an ios game called survivium a while ago, where players could program the things themselves: https://youtu.be/0WOJuZIlVfY?t=70