[PaulRobinson]

I've got a lot of hope for the Google DeepMind attempt to solve Mornington Crescent to add creativity back into the game, like they did for Go.

The problem is that Google seem to be struggling to get the computing power and algos together to deploy nib effectively and to square out the down flows when dealing with Lancashire Pushback - which you must have in a modern game playing against a top player. It's one of those things that seems still out of grasp of ML...

[stoneman24]

It was rumoured that an early form of the game was played and researched by AM Turing in the 1930’s. Work had been nearly completed to show that good play was undecidable, but it was abandoned to focus on the much easier ENGIMA problem. This was before the modern rules and the addition of the Elizabeth line, leading to much expanded game space.