Game Design and Interactive Narrative

Two articles that explore the oldest dream in game design: build a complex simulation, and fascinating emergent stories will write themselves.

Social Simulation In Games traces the Simulation Dream from Ultima Online's failed ecology through Dwarf Fortress to ChatGPT-powered OSRS bots. The key insight: the whole value of a game is in the mental model it projects into the player's mind, and players project emotion and meaning onto mechanical systems through apophenia — seeing intention in bouncing colored balls. Dwarf Fortress works not because its AI is sophisticated but because it generates situations that put human values at stake, and the player's imagination fills in the rest. The ChatGPT bots in RuneScape represent a new frontier: when NPCs can hold conversations, does it matter if they're automated?

Interactive Narrative picks up where social simulation leaves off: the quest for stories that respond to player choice with the coherence of authored fiction. From TALESPIN's Aesop fables through Facade's marriage counseling to Versu's gossiping AI characters, each attempt reveals something about why it's hard: invisible complexity (players don't notice the simulation), communication gaps (rich behavior that can't be perceived), and the fundamental tension between authored depth and emergent freedom. Disco Elysium's dialogue design shows that the answer might be constraint, not expansion — finding the exact shape of player agency that makes a specific character sing. The AI Dungeon Master analysis suggests that LLMs could eventually fill the role, but as one component in a composite system, not as a standalone intelligence.

Both articles connect to Predictive Processing through the brain's pattern-completion machinery, to Emergence through the question of how simple rules produce complex behavior, and to Simulators And Simulacra through the parallel between game personas and LLM simulacra.