Individual contributions to collective intelligence: spontaneous role specialization in synchronous fireflies

Intelligent behavior in biological systems can emerge from the collective dynamics of many simple agents. However, understanding the mechanisms that facilitate distributed computation remains a central challenge in many fields, with progress being limited by experimental design. Capturing collective behavior demands: 1. observation of every individual and 2. capturing every communication signal; but most experiments can capture only fragments of either. Here we introduce a new experimental paradigm that overcomes these challenges using synchronous fireflies as a model organism, which communicate through bioluminescent flashes. By spatially isolating and recording 81 fireflies, we have constructed the most detailed view of swarm dynamics ever obtained, capturing all social signaling, with unbroken tracking of each individual. Preliminary analysis reveals clear evidence of leader-follower dynamics, spontaneous role switching during and between experiments, and insights into how the collective synchrony stabilizes over time. Topological analysis reveals parallels to neuronal populations, with connection probabilities decaying as a power law, and flash activity resulting in traveling waves. Our goal is to leverage this dataset to form a more complete understanding of the local signaling mechanics that allow collective computation.