What if game NPCs could evolve?
Modern NPCs (Non-Playable Characters) are predictable. They either follow pre-scripted behaviors or rely on simplistic AI models that create the illusion of adaptability. Naturally, the next step in AI evolution should be models that can truly evolve—learning from experience, modifying their behavior based on external influences, and even developing unique traits and personalities. The Dao Code Dao Code is an innovative system that models the cognitive evolution of NPCs. It enables characters to adapt, change, and develop unique features through interactions with players, other NPCs, and their environment.
The Three Stages of Development
Establishing the foundational architecture of the metaverse and AI cognition.
Developing variations in AI personality structures, their core settings, and interaction principles.
AI characters transcend scripted behavior, learning, adapting, and developing individual identities through interactions with the world, players, and each other.
For structured visualization, I use Rider-Waite Tarot archetypes as conceptual templates for AI cognition.
The Code Dao system is built upon three key principles:
This allows for:
Filtered and contextualized perception – NPCs process information according to their psychological type.
Consistent behavioral logic – rather than acting randomly, they follow internal reasoning patterns.
Dynamic interactions – NPCs engage with players and other NPCs through their unique cognitive filters, rather than relying on scripted responses.
Inspired by the work of Antoni Kępiński and Aušra Augustinavičiūtė, Dao Code treats information processing as a metabolic system, where data is consumed, digested, and used to shape behavior. NPCs “consume” information just like humans process food. Different NPCs interpret the same event in different ways, based on their psychological makeup. Behavior evolves based on experience, allowing NPCs to learn and adapt.