Venice has long been shaped by its tides, which have influenced not only its physical landscape but also the lives of its residents. However, recent interventions, such as the MOSE project, have altered the natural flow and predictability of tides, distancing modern Venetians from their historical relationship with the lagoon. This thesis presents an initiative to reconnect Venetians, especially newer residents, with the traditional knowledge of tides, gathered through interviews with long-time lagoon dwellers. Central to this exploration is the development of a “Socratic chatbot”, an AI system that departs from conventional, answer-focused models by engaging users through reflective questioning, truly prioritizing the art of inquiry. Recognizing that Large Language Models, such as ChatGPT, are inherently designed to prioritize broad, global knowledge, this work reverses that tendency by fine-tuning the model with localized, domain-specific knowledge derived from oral histories. Complemented by meta-prompting, the resulting chatbot not only internalizes nuanced cultural insights but also guides users in exploring and questioning their understanding of Venice’s unique culture.

Dwelling the Tides. Exploring AI as a Tool for Reconnecting Venetians with Their Tidal Heritage

GREGGIO, SOFIA
2023/2024

Abstract

Venice has long been shaped by its tides, which have influenced not only its physical landscape but also the lives of its residents. However, recent interventions, such as the MOSE project, have altered the natural flow and predictability of tides, distancing modern Venetians from their historical relationship with the lagoon. This thesis presents an initiative to reconnect Venetians, especially newer residents, with the traditional knowledge of tides, gathered through interviews with long-time lagoon dwellers. Central to this exploration is the development of a “Socratic chatbot”, an AI system that departs from conventional, answer-focused models by engaging users through reflective questioning, truly prioritizing the art of inquiry. Recognizing that Large Language Models, such as ChatGPT, are inherently designed to prioritize broad, global knowledge, this work reverses that tendency by fine-tuning the model with localized, domain-specific knowledge derived from oral histories. Complemented by meta-prompting, the resulting chatbot not only internalizes nuanced cultural insights but also guides users in exploring and questioning their understanding of Venice’s unique culture.
2023
File in questo prodotto:
File Dimensione Formato  
DEF_master_thesis_Greggio.pdf

accesso aperto

Dimensione 2.25 MB
Formato Adobe PDF
2.25 MB Adobe PDF Visualizza/Apri

I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14247/24200