Based on a short discussion on twitter the other day about AI opponents in computer games I was wondering how often Machine Learning is used for such a job? Probably in most cases either an overkill or way too complicated. But there must be a (small) group of conflict simulations where this could be easily implemented.
This might be even a project for an interested student: writing an ML opponent for a wargame. Maybe for one of the COIN titles? Or as the player in ComancherÃa?
P.S. Uh oh, going down a rabbit hole. Just found this blog.
Gringogamer 
I think most of AI is rather scripted, not using ML. It is easier for developers, less prone to errors (you need to put a lot of effort to implement good ML). I would love to see this implemented in boardgames.
LikeLiked by 1 person
Yes, I also thought that this is nothing a software company would throw personpower at. There’s no money to be made from. But it might be something a group boardgame fans with programming skils can get behind.
LikeLike