Professional Practice and Decision-Making in the Age of Generative Artificial Intelligence: A Critical Approach to Large Language Models for Planning and Research Design




November 7th 2025


In addition to becoming part of individuals' daily lives, generative Artificial Intelligence tools are also impacting disciplinary methods and decision-making processes. Specifically, significant developments in natural language processing have led to the creation of large language models (LLMs) for both general purposes and specialized applications across various disciplinary domains—from mathematics to finance, healthcare to climatology, legal fields to traffic engineering.  


 


In the past two years, we have also witnessed the prototyping of a first language model (PlanGPT, Zhu et al. 2024) and a visual language model (PlanGPT-VL, Zhu et al. 2025) in the field of urban planning. However, there are numerous pros and cons, and between supporters and detractors, a third way is emerging—one more oriented toward hybridizing innovation and tradition (e.g., supervising LLMs through fine-tuning, text mining control, etc.). These aspects will be opened for debate, and several ongoing trends will be analyzed.  


 


With Dr. Vito Garramone, IUAV University of Venice  


 


Discussion:  


Barbara Pizzo, Sapienza University of Rome  


Giorgio De Ambrogio, Ph.D. Candidate, Sapienza University of Rome

© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma