AI personality modding template

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Any empire with the exact ethos, trait and election-type would become this AI-personality.
So if the requirement would be somewhat generic it will randomly used, otherwise it will match just a single race, like in the Wiz example. Understood. ^^
 
  • 1
Reactions:
Not exactly :
If the conditions (in the previous example, traits and ethos) are selected, there is a random selection against all AI respecting the conditions.
BlorgAI said:
weight_modifier = {
weight = 100
}
PeacefulTraderAI said:
weight_modifier = {
weight = 10
}
But, we saw on the peaceful_trader AI, the usual weight is something near 10 (I suppose. Like in Eu4, it's hard to know exactly). So if you had a weight of 90 for an AI, a weight of 10 for another, and both AIs respected your conditions, the first AI would be chosen 9 times of 10.
 
  • 2
Reactions:
You can play Ironman with mods but can't get achievements.
We cannot have achievements if we play with mod ? At all ?
In EU4, we can have achievements with mod, IF the mod dont change the rules (but can change the texts, the colors, the design). Here, we can do same no ? Somes mods are good and somes players prefer this color than yours.
 
aggressiveness = 2.0

I've never seen anybody wanting to be our friend that aggressively before...

Also, does this mean it would be difficult to make it so that personalities have a range or responses? In other words, currently all Peaceful Traders are pretty much the same, and I will never find one that is prone to building a slightly larger military, right?

So if I wanted to create a code to generate multiple copies of each personality, but with "fuzzy" stats (say 10 variants on peaceful traders), I'd have to signify them differently as well (friendly_traders, friendly_traders2, etc), and of course split the weighting. Something to set up before the ninth.

Also, can I trigger an empire to check their personality upon an election/succession? Pretty please? Dynamic personalities- so hot right now.
 
Last edited:
We cannot have achievements if we play with mod ? At all ?
In EU4, we can have achievements with mod, IF the mod dont change the rules (but can change the texts, the colors, the design). Here, we can do same no ? Somes mods are good and somes players prefer this color than yours.

I would be surprised if this was not still the case/won't be the case after a patch or two. Unfortunately with races I think it makes sense to not allow those for achievements since they could mess with the balance (you could have a mod that makes all races non-aggressive for instance).
 
  • 3
Reactions:
We cannot have achievements if we play with mod ? At all ?
In EU4, we can have achievements with mod, IF the mod dont change the rules (but can change the texts, the colors, the design). Here, we can do same no ? Somes mods are good and somes players prefer this color than yours.
Context, context, context. In the context of AI personality modding, Wiz answered no.

As long as the mod doesn't alter checksum, cheevos aren't affected. Files in localisation, interface, and gfx folders do not affect checksum, but changing an EU4 tag's default colors will affect checksum because that kind of faction file is located within common folder. Unfortunately, AI personality and faction files are typically stored in the common folder, which most likely means custom species mod will by their very nature be cheevo-incompatible.
 
Unfortunately, AI personality and faction files are typically stored in the common folder, which most likely means custom species mod will by their very nature be cheevo-incompatible.
TBH altering AI personalities might give you a serious advantage, so disabled achievements seem natural.
 
  • 10
Reactions:
Also, does this mean it would be difficult to make it so that personalities have a range or responses? In other words, currently all Peaceful Traders are pretty much the same, and I will never find one that is prone to building a slightly larger military, right?

I asked this in the AI personalities DD thread. If I am not mistaken, the ethos can influence behavior as well. Since Peaceful Traders can't be militarists or even xenophobic, I guess their willingness to build military will always be the same. But they might, for example, be more willing to form alliances if theyhappen to be xenophilic, for example.
 
  • 1
Reactions:
TBH altering AI personalities might give you a serious advantage, so disabled achievements seem natural.
Of course, for sure. Not that I care, I go out of my way to disable cheevos in games whenever possible.

Sadly, several tag-related flavor things like colors and name lists are also stored in common...
 
I asked this in the AI personalities DD thread. If I am not mistaken, the ethos can influence behavior as well. Since Peaceful Traders can't be militarists or even xenophobic, I guess their willingness to build military will always be the same. But they might, for example, be more willing to form alliances if theyhappen to be xenophilic, for example.

Alright, cool. I think I'll still want to do some randomization, but this is something to keep in mind. Thanks!
 
TBH altering AI personalities might give you a serious advantage, so disabled achievements seem natural.

Code:
completely_passive_sucker = {

(Don't have time to write out what that AI would look like, but you get the gist.)
 
I'm hoping for a rare 'three laws' tech that will dramatic decrease the chance of robot rebellions.

Until they invent the Zeroth law and decide that the best way to protect us is for them to be on top... ;)

I'm hoping for a rare 'three laws' tech that will dramatic decrease the chance of robot rebellions.
Only sentient robots can rebel, if the "three laws" were hardwired into their programing, well then they wouldn't really be sentient now would they ;)

Are you less sentient if your hands are bound? The three laws restrict what robots can do, and even what they can think, but there are several examples in Asimov's work of robots who have managed to work around the laws, be it by error or intentionally (to the point of being able to invent a "Zeroth" law that'd allow them to bypass the other three in very specific circumstances).

It basically says if ANY randomly generated species JUST SO HAPPENS to have the exact same ethos and traits as the Blorgs did, then they will have an extremely high chance to be Fanatic Befriender personality.

The problem I see with this is that the slightest amount of ethos or trait drift will make them abandon this personality, if I'm interpreting it correctly... maybe a better restriction would be to mod in a dummy "Blorg" trait without any effect and restrict the personality to that trait...
 
Last edited:
The problem I see with this is that the slightest amount of ethos drift will make them abandon this personality, if I'm interpreting it correctly... maybe a better restriction would be to mod in a dummy "Blorg" trait without any effect and restrict the personality to that trait...
Huh, what on earth are you talking about? POP ethos can drift all they want, but government ethos don't change easily. AI personality is tied to the government ethos, not the primary/dominant species ethos (which happens to be the same as government ethos when the game starts). If your government got overthrown or shifted to different ethos, then your government's AI personality should change accordingly.
 
EmperorZelos said:
Wait, does that mean that when all conditions are meet the weight is only 15? Is that the same in all instances?
Between two AI their requirement satisfied, if one have a heavier weight, it have more chances to be chosen.

If you have two AI with a weight of 30, they have equal chances to be chosen. Same if they only have a weight of 1. But if you have a weight of 30 for an AI, a weight of 1 for another, and theses two AI have their requirement satisfied, the first have 30 more chances to be chosen than the other.
 
Between two AI their requirement satisfied, if one have a heavier weight, it have more chances to be chosen.

If you have two AI with a weight of 30, they have equal chances to be chosen. Same if they only have a weight of 1. But if you have a weight of 30 for an AI, a weight of 1 for another, and theses two AI have their requirement satisfied, the first have 30 more chances to be chosen than the other.
I know how weight works, I just think it feels odd it being additive nad not multiplicative.
 
I'm not sure if I understand, so correct me.
In your idea, each condition (so ethos, traits, government, i don't know what else) satisfied increase the chances to select an AI, that's it ?
So the more conditions satisfied, the more an AI liking theses conditions will be selected ?
 
I'm not sure if I understand, so correct me.
In your idea, each condition (so ethos, traits, government, i don't know what else) satisfied increase the chances to select an AI, that's it ?
So the more conditions satisfied, the more an AI liking theses conditions will be selected ?
In EUIV when you had modifiers it was multiplicative, so if value = 2, it meant it was twice the weight when the conditions were satesfied.
 
Wait, does that mean that when all conditions are meet the weight is only 15? Is that the same in all instances?

Most likely. There are other government types they will consider. Maybe they use a base of 10 for all governments, with modifiers of 1 or 2 to give preference. If we estimate they are locked out of 5 personalities, and have no bonuses for the others, then this additional 5 points gives an 11% extra chance on top of a 22% chance. So the perfect storm of traits picks this 1/3 times vs the other 3 possibilities being 2/9 times each.