Actually, all of this is moot. These people have been programming games and game AIs for a very long time. I seriously doubt that anyone here could suggest some new trick that they haven't thought of. Multi-threading is definitely not a new trick. Scenarios, units, civs, game concepts, etc., sure, people here can and have come up with lots of ideas. When it comes to implementing those, though, we can tell them nothing useful. They know it's important. They know the AI has been bad and slow before. I have full confidence that the developers will be able to provide the best AI with the best performance that is possible, or pretty close to it.