An A.I. cannot learn at our present moment, at least not like we learn, so therefor every possible strategy and tactic needs to be programmed and planned for ahead of time. Since they cannot learn, they do not have human-like situation awareness and judgement calls.
Until we get true A.I., every past and current game will have roughly the same level of A.I. (i.e. not true A.I.).
The good thing about space sims is that even though at it's surface is complex, it's not terribly hard to program decent A.I. due to the lack of complexity knowing the strategies and tactics are few (fewer then others).
Even GTA 5 or Fallout 4 is nothing more then lines of programming telling the AI what to do in any particular situation, all known ahead of time and coming to the conclusion using math and variables, not human intuition.
Until we get true A.I., every past and current game will have roughly the same level of A.I. (i.e. not true A.I.).
The good thing about space sims is that even though at it's surface is complex, it's not terribly hard to program decent A.I. due to the lack of complexity knowing the strategies and tactics are few (fewer then others).
Even GTA 5 or Fallout 4 is nothing more then lines of programming telling the AI what to do in any particular situation, all known ahead of time and coming to the conclusion using math and variables, not human intuition.
