Modeling and Simulation of Artificial Societies to Study Precarity and Inequity

Nokhiz, Pegah

Full Metadata

Overview

Year:: 2024
Contributor:: Nokhiz, Pegah (creator); Venkatasubramanian, Suresh (Advisor); Bach, Stephen (Reader); Patwari, Neal (Reader); Brown University. Department of Computer Science (sponsor)
Genre:: theses
Subject:: Artificial intelligence; Fair Machine Learning; Algorithmic Fairness; Social Computing; Artificial Societies
Extent:: xii, 115 p.

Files

Description

Abstract:: Financial insecurity characterizes modern life. It reflects ongoing instability in employment and earnings which are notably pronounced in the gig economy. This instability is compounded by the widespread use of automated decision-making tools that directly affect employment and income. Over time, this “precariousness” unfolds as a sequence of events for individuals. Thus, to understand and address it, a shift in perspective from decision-makers to individuals is necessary. This requires that we develop “artificial societies” - computational simulations of an agent-based behavioral model capable of capturing various related phenomena simultaneously, including individual consumption responses to financial shocks, the influence of predictive tools on income, and the long-term behavior of individuals striving to maximize utility. This individual-level perspective is one direction to study precarity and inequity in artificial societies with computational simulations or models designed to replicate and investigate the intricate behavior of complex social systems. However, there is also a societal-level viewpoint wherein neither a singular decision-maker nor defined agent behavior rules exist. Consequently, there is a need for a model that does not attempt to describe underlying systems or capture individual actions. Adopting a system-based approach to studying inequity in feedback loops opens avenues to explore social systems that are otherwise challenging to model directly. This dissertation first introduces the concept of latent financial instability, or precarity, to the artificial intelligence community. It develops agent-based computational models embodying realistic human-like behaviors to explore precarity dynamics, drawing from various strands of inquiry in economics. Additionally, we investigate work schedule instability and the impact of foresight on financial security. Next, we present a model from linear systems theory to quantify feedback in social systems holistically, enabling the examination of long-term policy effects even without individually characterized feedback mechanisms. Our frameworks facilitate the examination of precarity dynamics, the development of mitigation strategies for precarity, policy investigations, and the production and sustainment of long-term equity.
Notes:: Thesis (Ph. D.)--Brown University, 2024

Content

Citation

Nokhiz, Pegah, "Modeling and Simulation of Artificial Societies to Study Precarity and Inequity" (2024). Computer Science Theses and Dissertations. Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:xextjmwj/

Relations

Collection:

Computer Science Theses and Dissertations

Theses and Dissertations for the Computer Science department.
...