Speakers

Abstract:  Many complex systems consist of simple units that follow local rules yet collectively produce large-scale patterns. A long-standing challenge is understanding when the future behavior of such systems becomes predictable. In many deterministic systems, the rules are fully known but the outcome of a particular initial configuration remains difficult to anticipate. In this talk, I will describe a computational “experiment” based on a generalized cellular automaton of non-locally communicating cells that self-organize into distinct macroscopic patterns, including static arrangements, traveling waves, and spiral waves. Although the dynamics are entirely deterministic, the final pattern cannot be reliably inferred from the initial configuration. By recasting the system in terms of a discrete phase field, hidden structures become visible: localized vortices that move, interact, and annihilate over time. These vortices form a set of collective dynamical objects whose interactions progressively constrain the system’s future. As the dynamics unfold, the system passes through stages where the remaining trajectories become increasingly restricted, revealing when and how predictability arises. These results suggest a general principle: in some high-dimensional deterministic systems, predictive structure is not encoded in the initial state but instead emerges through the system’s evolving collective dynamics.

Bio:  Hyun Youk is an Associate Professor of Physics at UIUC. His research combines experiments and computational “experiments” to study how simple interacting units—such as communicating cells—give rise to collective behaviors. His group develops computational and mathematical models alongside wet-lab experiments to uncover dynamical principles underlying self-organization and predictability in biological systems, including how living cells maintain—or irreversibly lose—the capacity to self-organize during slow or suspended states of life.