Many Body Dynamics of Living Systems
Open Access Depositedhow mainstream parenting communities drifted toward anti-vaccine narratives before COVID vaccines existed, how networks self-heal after platform interventions, and how topic entanglement creates system-level resilience. The phenomenological framework proves remarkably versatile. My coupled differential equations captured cross-platform hate network synchronization around January 6, 2021, and the evolution of pro/anti/neutral vaccine discourse on Facebook during COVID-19's critical early months. Network analysis reveals that post-pandemic resilience emerges from multi-dimensional topic entanglement across geographic scales, explaining why targeted interventions consistently fail. The framework enables quantitative predictions about onset timing, identifies network signatures that precede polarization, and suggests interventions that work with rather than against network properties. This understanding carries dual implications. While the mapped ecosystem reveals intervention opportunities through heterogeneous group deliberation and network engineering, it simultaneously exposes vulnerabilities. The extensive distrust ecosystem identified through this work represents an attack surface for AI-generated manipulation, transforming the mapped vulnerabilities from diagnostic tools into potential blueprints for automated exploitation. By building understanding from the bottom up---treating online communities as a well-defined system where collective behaviors emerge more robustly than individual pieces---this thesis demonstrates how physics thinking can illuminate the fundamental dynamics governing billions of humans interacting in digital spaces.
Online extremism spreads through social media networks in patterns that appear chaotic yet follow reproducible dynamics. This thesis applies many-body physics to understand how extremist movements emerge and propagate across digital platforms. Building on our research group's derivation of a generalized Burgers' equation for online community dynamics, I developed a phenomenological model that captures the essential physics while enabling practical multi-platform analysis. This approach sidesteps the computational intractability of exact Lambert W solutions, instead using exponential forms that retain the key features---sudden onset and saturation---while permitting coupling terms necessary for modeling interacting movements across platforms and ideologies. The empirical foundation centers on the Facebook vaccine ecosystem
1,356 interconnected communities representing nearly 100 million individuals. I inherited and maintained this dataset from its pre-pandemic state through 2025, documenting its transformation from routine vaccine discourse through COVID-19 polarization into post-pandemic multi-topic entanglement spanning vaccines, climate change, elections, and conspiracy theories. This longitudinal stewardship revealed dynamics invisible in static snapshots
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.