From Silence to Voice: Artificial Intelligence as an Interpreter of Architectural Memory

So architecture is not mute. It is a means of communication. This is a story about the depth of our passions. This is an impartial assessment of centuries now past.” (Daniel Libeskind). Inspired by this reflection, the present research proposes a model of open and accessible dialogue between humans and buildings through the integration of two innovative tools within the digitalization domain: Building Information Modelling (BIM) and Artificial Intelligence (AI). The proposed methodological framework unfolds through three phases that reflect the essential steps of a conversation: listening (Data Collection), interpretation (Data Restitution) and response (Data Interaction). Within this process, the building, conceived as an archive of memory, undergoes a conceptual metamorphosis, evolving from a “silent” entity to a “thinking” and finally to a “speaking” building. Interaction occurs through textual input that, once processed by Large Language Models (LLMs), queries the BIM model and returns outputs in natural language. The methodology aims to restore to cultural heritage the ability to narrate itself, fostering a renewed relationship between architecture and the community through digital intelligence.