CAIO - B53C25003940001
City-AI Coevolution
Site: https://www.caio-project.it/
Contacts
Abstract
Urban environments worldwide are increasingly shaped by city-AI coevolution, a dynamic process where inhabitants and AI algorithms continually influence one another. Navigation services can reroute traffic flows, ride-hailing platforms can reshape mobility patterns, house-rental services can alter neighbourhood demand, and location-based recommenders can steer foot traffic. These changes are amplified and accelerated by a feedback loop in which users’ choices generate the data that train algorithms, which then steer future choices, potentially creating unintended or unforeseen urban outcomes. Yet this phenomenon remains understudied in AI and complexity science. CAIO aims to fill this gap by combining approaches from AI and complexity science to: (i) Understand the city-AI feedback loop’s effect on urban dynamics and its interplay with urban policies; (ii) Model the essential mechanisms that drive city-AI coevolution; and (iii) Govern these coevolutionary processes throughcomplexity-informed machine learning models that balance individual and urban goals within an ethical, transparent framework. CAIO’s research activities will be grounded in large-scale empirical and simulation-based experiments across four case studies – navigation services, ride-sharing/hailing, house renting, and location-based recommendations – to capture real-world and theoretical insights. CAIO aspires to pioneer an open science of city-AI coevolution, revealing the impact of AI-driven recommenders on urban environments and offering tools for managing this evolving landscape.
Duration
60 Months
Financial Institution
Ministeriale/Governativo