Thanks for playing. ML vs ML was interesting concept to try out for this Jam but it needs a lot more work.
This was a prelude to a project that would involve teams of agents and player among them, but now I am not convinced it will actually end up being fun. 

The gimmick was supposed to have neural network control the character but it turned out to be very hard to train that in 2 days. :) 

Glad you like my visuals. 

Thank you! Glad you liked  it, I try to bring new trick to every new game jam. There is more "dynamic" version of Pepet Master coming tonight or tomo morning since I have figured out how to train them better.