The Freeciv AI seems to have done this for as long as I can remember, which is before version 2.0. However, the AI does seem to be less aggressive with less players total -- like about five players. I tend to play with many AI players so my memory may be biased towards that situation.
I know the AI has always been pretty aggressive, but I was sure I'd seen them be quite competitive with building the spaceship in the past. I remember one good game where the world superpower (I was like third or something) built and launched a spaceship, and the other AIs all ganged up on them and took their capital, causing the spaceship to be destroyed. They ended up being completely wiped out. Perhaps it was because I was on a lower difficulty and with fewer civs, I wonder if the increased aggression on higher difficulties causes them to lose focus on technology.
I ran a few other AI only games for testing just out of curiosity (on normal difficulty, varying map sizes and number of civs), and every time one AI becomes the overwhelming superpower. They are the only one to reach the end of the tech tree, and sometimes they made limited attempts to build the spaceship, but it always ended in them destroying all other civs.
If you make a new rule set, and then add a rule like this:
Thanks so much for this The Square Cow, this sounds like exactly what I want to play around with. Is the value a percentage? Does it only affect AIs love towards humans? I'll have a look for some documentation in a minute.Edit: Found it.
Looks like it adds x points of love (presumably per turn) with the AIs, but it doesn't say if this is toward everyone or just humans. Anyone know? Otherwise I'll just have a play with the setting.
Perhaps the most simple way to make a first step amendment is to offer a switch where AI players have to play by the rules, without advantages.
As far as I know, the AI actually doesn't cheat on any difficulty level (except Cheating of course!), except that they can set their tax rate to anything and they can see the whole world.
One thing that make it seem like a challenging opponent is that it plays with extreme risk (especially in versions before 2.5)
This makes sense, and applies in some way to the commercial games as well. The aggressive AIs tend to perform better than the peaceful ones, and so this is a way to make the AI competitive. It would just be nice if so many games weren't a permanent war state.
I'd like to help improve the AI if I could, I've read through the code in advdiplomacy.c many times and I think I understand it pretty well, I just don't know where to start and what changes are desirable to the project!