Answers are grounded in this paper and cite their sources.
Given the complexity of cities today and the deepening debates about urbanization, developing new approaches in education and research has become increasingly important. This study explores the potential of metaverse-based educational and research environments by developing an Urbanist Metaverse (UM) platform. The UM integrates generative agents modeled after prominent urban theorists, who operate autonomously within a virtual environment built on the AI Town software. Eight agents were instantiated using the Meta:llama3 language model (without further fine-tuning) and the mxbai-embed-large embedding model; their behaviors were defined through prompt engineering based on each theorist’s core concepts and objectives. Methodologically, the study focuses on analyzing theoretical coherence, socio-epistemic interaction patterns, and pedagogical affordances within the simulated environment. Analyses indicate that discussions on key urban concepts evolve throughout agent interactions, resulting in the formation of a dynamic and interconnected knowledge ecosystem. In contrast, authentic user-agent interactions reveal the platform’s potential to provide educational scaffolding, as demonstrated by agents offering well-supported academic explanations and referencing relevant literature. Although the UM platform features immersive elements that may enhance the sense of presence, the current study did not empirically assess this aspect. This research emphasizes the capacity of metaverse platforms to innovate educational practices in urban planning by fostering immersive, interactive settings that promote critical analysis, cross-disciplinary engagement, and theoretical exploration.
No takes yet. Share an insight, caveat, or question.
Mete Korhan Özkök (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: