Academic Journal
Uni-Town: Simulating Academic Conflict and Emergent Narratives in a Generative AI Simulation 300 Word statement
| Τίτλος: | Uni-Town: Simulating Academic Conflict and Emergent Narratives in a Generative AI Simulation 300 Word statement |
|---|---|
| Συγγραφείς: | Umran Ali |
| Έτος έκδοσης: | 2025 |
| Συλλογή: | University of Salford: Figshare |
| Θεματικοί όροι: | Creative writing (incl. scriptwriting), Digital writing, Other creative arts and writing not elsewhere classified, Creative arts, media and communication curriculum and pedagogy, Interactive narrative, narrative, generative AI, software, Generative agents, computer narrative, Computer games--Programming, higher education, simulation |
| Περιγραφή: | Uni-Town, explores a generative AI simulation of a university to model the interpersonal and ideological conflicts endemic to modern academia. The central research aim was to determine how a simulation driven by AI agents, informed by creative writing and critical theory, could function as a novel tool for both narrative generation and pedagogical reflection. This work contributes a new methodology for organisational analysis, positioned at the intersection of generative agent research (Park et al., 2023), critical university studies (Fleming, 2021), and narrative design. The research was conducted using a rigorous, practice-led qualitative methodology. The simulation was built using the open-source AI Town framework, populated by twelve AI personas representing roles from across a Higher education institution. The personas were developed through a deliberate triangulation of insider-researcher experience to ensure authenticity, Fleming’s archetypes to embed ideological conflict, and narrative design principles to create dramatic potential. The simulation ran for 45 minutes, generating approximately 3,500 conversational utterances that were subsequently thematically analysed. The core finding is that the simulation successfully produced plausible, emergent dialogues that cohered into distinct narrative arcs. Agents autonomously performed an "ideological cold war," complete with the performance of bureaucracy and the emergence of authentic character voices conveyed through emotional cues. The simulation's ability to generate sophisticated, unscripted political critiques, such as condemning the "neoliberal status quo," demonstrates its capacity for modelling complex social dynamics. The research's significance is in validating agent-based simulations as a powerful reflective tool, offering a low-stakes 'mirror' for organisational analysis and a dynamic sandbox for narrative prototyping that shifts the writer's role from creator to curator. The work was disseminated through its detailed paper and a LinkedIn article ... |
| Τύπος εγγράφου: | text |
| Γλώσσα: | unknown |
| DOI: | 10.17866/rd.salford.30075892.v1 |
| Διαθεσιμότητα: | https://doi.org/10.17866/rd.salford.30075892.v1 https://figshare.com/articles/online_resource/_b_Uni-Town_b_Simulating_Academic_Conflict_and_Emergent_Narratives_in_a_Generative_AI_Simulation_300_Word_statement/30075892 |
| Rights: | CC0 |
| Αριθμός Καταχώρησης: | edsbas.C941C359 |
| Βάση Δεδομένων: | BASE |
| DOI: | 10.17866/rd.salford.30075892.v1 |
|---|