Simulation study reveals that emergent topological structures drive predictability in deterministic cellular automata, indicating that forecastability can be dynamically constructed.
A deterministic system’s future can be fixed from the start yet remain unpredictable because of chaos or computational irreducibility. Here we reveal another source of unpredictability for non-chaotic deterministic systems: collective entities making the future legible are initially absent but dynamically constructed. In a generalized cellular automaton, disordered lattices evolved into static configurations, rectilinear waves, or spiral waves. Although initial configurations fixed each fate, machine-learning models did no better than randomly guessing fates from them. Recoding cell states geometrically revealed self-organization of prediction-enabling topological entities: vortices, non-contractible-loop strings, and, most generally, a winding field describing how same-state regions wrap the lattice. As the winding field self-organized, initially unpredictable fates became increasingly legible. Static and rectilinear-wave fates became progressively predictable, whereas spiral-wave fate became accurately predictable only near wave formation. These results establish that self-organization can build not merely order but entities that make certain futures visible, revealing a gap between determinism and practical predictability. In a deterministic, non-chaotic system, the initial state fixes the future completely, yet that future may be impossible to forecast from it. Using a cellular automaton of communicating cells, this study shows that predictability is instead built by the dynamics: collective topological structures self-organize as the system evolves, and only once they exist does the eventual outcome become legible.
No takes yet. Share an insight, caveat, or question.
Koopmans et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: