This article describes research on using local interactions to generate intricate global patterns and emergent urban forms. An agent-based system (ABS) is used to optimize an urban network and construct the micro-level complexity within a simulated urban environment. The author focuses on how agent-driven emergent patterns can evolve during the simulation in response to the “hidden hand” of macro-level goals. The research extends to the agents’ interactions driven by a set of rules and external forces. An evaluation method is investigated by combining network optimization with the space syntax. The multi-phase approach starts with defining the self-organizing system, which is created by optimizing its topology with ABS. A macro-level “attraction map” is generated based on the space syntax analysis, where the map is used to control various construction operations of an urban model.