I think it would be interesting if the AI could keep up with this suggestion and all types of victory points were equally their goal. Now adays it seems like the AI is very personality and civ based. Shaka goes for domination whether he's isolated or not or on a sea based map or not. I haven't noticed so far if they are aware of their surroundings or just playing as a personality.
It would be interesting though if the AI was more adaptive to the different types of victory points. Like if a warrior civ that ignored religion conquered the holy site a of a really wide spread religion and decided: "Hey If I can get the rest of the city states in my religion and 5 of my own cities I get 5 VPs I'm going to spam faith and get great prophets." Or If a scientific civ gets to astronomy first and though it never built much of an army found another continent with a single low tech civ on it would invade it for some kind of other victory point condition.
It seems like the hardest part of this suggestion would be programming an adaptive ai aware of all the VPs but one that's not to hard that wipes you out too quickly. Either the AI quickly adapts from one VP to another with lazer accuracy or it kind of vaguely builds up while the Human gains easy VP after easy VP. That would be difficult to make work.
It would be interesting though if the AI was more adaptive to the different types of victory points. Like if a warrior civ that ignored religion conquered the holy site a of a really wide spread religion and decided: "Hey If I can get the rest of the city states in my religion and 5 of my own cities I get 5 VPs I'm going to spam faith and get great prophets." Or If a scientific civ gets to astronomy first and though it never built much of an army found another continent with a single low tech civ on it would invade it for some kind of other victory point condition.
It seems like the hardest part of this suggestion would be programming an adaptive ai aware of all the VPs but one that's not to hard that wipes you out too quickly. Either the AI quickly adapts from one VP to another with lazer accuracy or it kind of vaguely builds up while the Human gains easy VP after easy VP. That would be difficult to make work.