Realism of social interactions: AI in the game - the key to bringing the entire inzoi game world to life

Wishlist
, , , ,
1

Hello, development team! :blush:

Screenshot-20260819-025651
Screenshot-20260819-0256511114×723 120 KB

First, I want to say - the game inzoi looks absolutely stunning. The graphics and level of detail are just incredible. But there’s one thing that already sets your game miles ahead of any competitor, including The Sims. It’s your “icing on the cake” - artificial intelligence (AI) at the core of the game. This is your main trump card, and I really want you to use its full potential to bring the social aspect of the game to life :rocket:

You know what’s most frustrating about life simulators? When you spend hours in the editor, fine-tuning every facial feature, picking the perfect style, creating a unique character type for your zoi… and then you go out into the city and realize the game world doesn’t care about you. Other characters walk past without noticing your efforts. It feels like you’re playing “just for yourself” in a void. The purpose of creation is lost because your work goes unnoticed within the game itself :pensive_face:

Your advanced AI can fix this. Make it so the AI in the game makes characters feel almost “alive”! And here’s how it could work:

  1. Personal Taste System (AI as the “eye” of a character) :eye:
    Let the AI analyze the tags that already exist in the game. Each zoi should have their own preferences: height, body type, clothing style, tattoos, or accessories. When the AI of other characters starts “reading” the appearance of my zoi and matching it with their own taste, the game gains depth. Creating a character becomes not just a waste of time, but an important part of your strategy for life in the city.
  2. “Living” First Impression :sparkles:
    I’d love to see that my efforts are noticed! If a zoi walks by who matches another character’s “type”, let the AI produce some instant reaction, for example: a smile, a lingering gaze, or a thought bubble: “Wow, what style!” or “This is my ideal!”. When you see that the world reacts to your character, interest in the game skyrockets.
  3. Intelligent Boundary Protection and Fixing the Jealousy System :shield:
    Right now the jealousy system in the game works “awkwardly”. Zoi can start autonomously flirting with someone else right in front of their partner or spouse. As a result, the partner gets upset immediately, turns on jealousy, and is ready to break up because of actions the player didn’t even plan! This is a serious logic flaw.
    The AI core should first check relationship status and the presence of a partner nearby before allowing autonomous flirting. Relationships should feel like something important that the AI recognizes and protects. Instead of “breaking” relationships, the AI should make the partner react to the outside threat: intervene in the conversation, hug their significant other, or set boundaries with the stranger.
  4. Smart Autonomy: Logic Instead of Random Scripts :brain:
    Thanks to AI at the core, characters can behave meaningfully. A zoi in love or in a family shouldn’t start flirting with everyone because of a mood. The AI should “understand” the value of current relationships and block inappropriate behavior. Flirting should be a rare and targeted event based on genuine attraction, not a random blind action.

It’s the AI core that will allow you to create not just a game, but a living world where every zoi feels, sees, and reacts logically with understanding of their own actions. Use this advantage to make the game world and characters feel almost “alive”!:fire:

Thank you for considering my ideas!

2 Likes
2

I actually like parts of this idea, especially having attraction preferences, first impressions and Zois taking their current relationship into account before autonomously flirting. But I’m a little confused why all of this would need AI at its core.

If things like tattoos, clothing style, body type, height etc. may already have tags, couldn’t the game simply compare those tags with a Zoi’s preferences and calculate an attraction/compatibility score? The same with relationships: if a Zoi is married, has a loyal personality and their partner is nearby, the chance of autonomously flirting could simply be much lower. That sounds more like normal game logic/autonomy to me than something that needs generative AI.

I would actually be concerned about performance if every Zoi around us constantly had to use AI to analyse every other Zoi they encounter. Smart Zoi is already quite demanding and still has stability/optimization issues, so expanding that kind of processing across normal social interactions sounds very expensive to run.

Personally I would rather see a good attraction + preference + relationship-awareness system that uses existing traits/tags and only calculates things when necessary. It could still create exactly the kind of natural reactions described here without needing AI for everything. :four_leaf_clover:

Also… I’m genuinely confused by the VR picture. Is VR supposed to be part of the suggestion? Because I couldn’t find anything about it in your post…

3 Likes
3

Thank you for the feedback! I’m glad my ideas about the preference system and jealousy in the game resonated with you :handshake:

To answer your question about AI: “ordinary game logic” is exactly what makes other life simulators predictable and boring after just a few hours of gameplay. Tag comparison is just dry mathematics, but real appeal is always about context :brain: The AI core can process complex combinations of character traits, moods, and environment, creating living, nonlinear behavior. A simple “compatibility score” can’t do that. Since inZOI already has an AI-based architecture built in (Smart Zoi), it makes much more sense to leverage that power for deep social mechanics than to layer outdated scripts on top :rocket:

As for performance: if Krafton is building a game around an AI core, they’re optimizing the engine for such tasks from the start. Using an existing core to create social “chemistry” is strategic resource allocation, not extra strain on the system :gear:

About the VR picture — it’s simply a visual metaphor for that “living” world I’m trying to describe. The image, so to speak, embodies the moment when a virtual world, thanks to AI, becomes so emotionally deep and realistic that the boundary between game and reality seemingly disappears :sparkles: This isn’t a suggestion to add VR headset support, but rather a symbol of that “genuine” feeling I’d like to see in the relationships between Zoi characters. It’s the AI core that’s the bridge connecting player creativity and a living, responsive game world :globe_with_meridians:

4

And one more important point I’d like to add. It seems to me that if we, as players, preemptively limit our ideas and dreams, worrying that it will be difficult for the developers to implement something or that they might not want to do it — then the very meaning of the “Suggestions” section simply gets lost.

Our task as a community is to share our vision of the ideal game and inspire the development team to achieve new heights. As for questions of technical implementation, optimization, and choice of tools — it’s better to leave those to the professionals at Krafton. After all, it’s often the bold and sometimes “impossible” ideas from players that become that very spark which drives the industry forward and creates something truly great! :sparkles:

5

I get what you mean, but I still don’t really agree that AI is necessary for this.

I also don’t think a preference/tag system would automatically be predictable or boring. We as players wouldn’t necessarily have to know another Zoi’s preferences from the beginning. We also don’t immediately know every trait of a Zoi we meet now. We can ask about traits and discover more about them over time.

Attraction preferences could work the same way. Maybe my Zoi meets someone and I notice that person seems especially interested in him, but I don’t immediately know why. Later I could discover that this Zoi likes tattoos, a certain clothing style, certain personality traits etc. The calculations happening underneath can still be hidden from the player.

And “normal game logic” also doesn’t have to mean one simple compatibility score. It can take personality, mood, relationship status, memories, previous interactions, attraction preferences, the current situation and even some randomness into account. Games have been creating complex and unpredictable NPC behaviour with systems like this for a very long time.

I also still disagree regarding performance. Just because Smart Zoi already exists doesn’t mean using AI for more and more decisions suddenly has no additional cost. If many Zois around us have to use AI to evaluate other Zois, their appearance, relationships, mood, environment and context, that processing still has to happen somewhere. Using an existing AI system much more frequently and for many more characters is still additional work for the system.

And if those calculations are heavily limited, cached or only triggered at certain moments to save performance, then we are already getting quite close to the kind of optimized game logic I was talking about in the first place.

I also don’t completely agree that we shouldn’t discuss limitations in the Wishlist because developers should figure everything out. Of course we don’t need to be programmers to suggest something, and I absolutely think people should be creative and suggest ambitious ideas. But discussing performance, scope and possible alternatives doesn’t kill an idea. Sometimes it actually helps finding a more realistic way to achieve the same gameplay.

And if technical implementation and the choice of tools should be left to KRAFTON, then I think the important suggestion is the desired result: deeper attraction, hidden preferences, better first impressions, jealousy and relationship-aware autonomy. KRAFTON could then decide whether AI, traditional simulation systems or a combination of both is the best way to achieve that.

I like the actual relationship ideas here. I just don’t think every complex behaviour automatically needs AI behind it. :four_leaf_clover:

1 Like
6

The proposal is interesting, but @Minthyuyu is absolutely right: everything you’ve described is achievable using scripts. There’s really no need for AI here at all — a relatively complex system would be enough — so that the reaction doesn’t happen every time but occurs with some probability and depends on Zoe’s mood (if he’s feeling sad after some events, he might simply “not notice” our, say, lady, no matter how beautifully made up and dressed she is, but in a normal state — react to her), and of course, all of this would depend not on an additive but on a multiplicative combination of attractive features.

Anyway, it’s a good idea, and it would be great if they implement it. But it’s possible without any AI whatsoever, and with a competent approach you won’t even notice how exactly it was done.

1 Like
7

Thank you for continuing the discussion! :blush: But I’m getting the impression that this forum thread is turning into not “Suggestions”, but some kind of competition — who will propose a better idea or who will find a flaw in someone else’s idea faster. This seems a bit strange to me, because we’re here to share dreams about the perfect game, not to compete with each other.

Moreover, it seems to me that you’re overlooking how the inzoi core is actually structured. The game already has an “AI Thinking” feature that can be enabled in the settings. And, as we can see from the instructions, this AI becomes prioritized over regular scripts. It generates thoughts and determines the behavior of zoi based on a text description (prompt) that we can write ourselves.

So, my main argument is: why would developers waste time creating a “complex script system” for jealousy or attraction when they already have autonomous AI that can “think”?

All I’m proposing is an expansion of this already existing system. If the AI can generate thoughts based on a character description, then adding “preferences” and “reactions to a partner” parameters to it — that’s the most logical and simple path. The AI will itself understand how to react to a tattoo or a stranger’s flirting based on the zoi’s character description. This is much more efficient and modern than writing thousands of “if… then…” conditions.

As for performance: since the AI activation feature is already implemented and works autonomously, it means the developers have already solved the optimization question. Adding social logic to this same AI process won’t create “new” load, it’s just a more efficient use of what’s already built into the game.

Let’s face facts: Krafton created inzoi as a next-generation game precisely because of AI. Using old methods (scripts) where intelligence already works — that’s a step backward. I’m all for using the game’s potential at full power! :rocket::sparkles:

8

I’m honestly a little confused by the “competition” part. I don’t think either @Carnel or I are trying to propose a “better idea” or find flaws faster than someone else.

We both actually said that we LIKE the idea of deeper attraction, preferences and relationship-aware behaviour. The only thing we are discussing is whether AI specifically is necessary to achieve it. To me that is just normal discussion of a suggestion, not a competition.

I also still don’t agree with the idea that because Smart Zoi already exists, the optimization question is basically solved and expanding its use would create no additional load.

Kjun himself recently talked about this. He said that Smart Zoi currently has bugs where it fails to respond properly as existing content expands, and that the responsible team is still discussing improvements:

In the very same post he also explains that even the normal autonomy system can become expensive when many Zois perform decision-making at the same time. Because of that they are optimizing the search order, limiting the search scope and reducing unnecessary calculations.

So I don’t think “the system already exists” automatically means using it for many more calculations has no performance cost. How often something runs and how many Zois use it still matters.

There is another interesting Kjun post regarding this as well:

Kjun specifically says that the LLM-based NPC storytelling system they were developing is not the current Smart Zoi. He also talks about improving things such as betrayal, jealousy, rumors and Zois recognizing relationships through multitasking, group behaviour and the event system.

So even KRAFTON themselves don’t seem to treat this as “Smart Zoi already exists, therefore everything social should use Smart Zoi.” They appear to use different systems depending on what they are trying to achieve.

That is why I also don’t really agree that traditional systems are simply an outdated step backwards. A good game can use AI where AI actually provides an advantage, normal simulation/game logic where that works better, and possibly combine both.

And “normal logic” doesn’t mean developers have to manually write thousands of simple “if this → then that” conditions either. There can be weighted probabilities, traits, moods, relationships, memories, context and randomness involved.

For example, the preference system we are talking about could still be completely unpredictable to the PLAYER because we may not know another Zoi’s preferences when we first meet them. We could discover those preferences gradually, similar to how we discover traits now.

I’m not against AI in inZOI at all. I just don’t think “AI exists in the game” automatically makes AI the best or cheapest solution for every complex behaviour.

And this is also why I mentioned earlier that if the technical implementation should be left to KRAFTON, as you said, then I think we can suggest the result we want — deeper attraction, first impressions, jealousy, hidden preferences and better relationship awareness — and let KRAFTON decide which combination of systems is actually best for achieving it. :four_leaf_clover:

2 Likes
9

Thank you so much for such a thoughtful response and for the links to K-jun’s words! :blush: It really does help to better understand the challenges the development team is facing right now.

I want to clarify one important point: my proposal is absolutely not a demand for the immediate implementation of all these features. I understand perfectly well that the game is currently in an active phase of polishing and bug fixing. I’m viewing my ideas as a vector for steady, step-by-step development of inzoi in the future. It’s more of a long-term vision of where the game could be in a year or two, once the main optimization issues are resolved.

I agree with you that the tools — whether pure AI, traditional logic, or a combination of both — are a choice for the professionals at Krafton. My task as a player is not to limit myself by fear of current technical challenges, but to inspire the team to create something truly unique from a long-term perspective. Traditional systems are an excellent foundation, but it’s the potential of AI, in my view, that will allow the game to “breathe” and remain alive years from now.

I’m glad that on the main point — the need for emotional depth, jealousy, and meaningful relationships — we are completely in agreement. And as for exactly how this will be implemented technically and on what timeline, let K-jun’s team decide. The important thing is that in the end, the world of zoi stops feeling empty and becomes truly responsive to us and our creativity :sparkles:

Let’s just believe in the potential of this game and support the developers in their bold endeavors! :four_leaf_clover:

2 Likes
10

I hope that it is a bit more clear that I really don’t mean to “diminish” or “talk down” on any ideas. And I absolutely don’t want to compete and be like “my idea is the better one!”. That isn’t what I want.

I just try to think of how something like this can be implemented to be enjoyed by all players (in regards of performance) plus also see it from the devs’ standpoint.

But in conclusion, as we both agreed on and I really agree 100%, is the end result that we get deeper relationships. And since kjun and the team are already cooking something up in that regard, alongside with the “fundamentals first” principle, I’m sure similar mechanics will be implemented one way or another​:four_leaf_clover:

2 Likes
11

Thank you for understanding, and I’m glad we heard each other :blush:

Ultimately, we’re all in the same boat and want inzoi to become the game of our dreams. I completely agree — the choice of tools remains with the professionals, but we really care about the development of this project. It’s great that we can not just watch from the sidelines, but actually help the game become better by sharing our thoughts and experience. I wish luck and inspiration to all of us — both players and, of course, the development team in creating this amazing world! :four_leaf_clover::sparkles:

2 Likes