What the flock knows that the birds do not: exploring the emergence of joint agency in multi-agent active inference

Collective behavior pervades biological systems, from flocks of birds to neural assemblies and human societies. Yet, how such collectives acquire functional properties – such as joint agency or knowledge – that transcend those of their individual components remains an open question. Here, we combine active inference and information-theoretic analyses to explore how a minimal system of interacting agents can give rise to joint agency and collective knowledge. We model flocking dynamics using multiple active inference agents, each minimizing its own free energy while coupling reciprocally with its neighbors. We show that as agents self-organize, their interactions define higher-order statistical boundaries (Markov blankets) enclosing a flock'' that can be treated as an emergent agent with its own sensory, active, and internal states. When exposed to external perturbations (a predator’’), the flock exhibits faster, coordinated responses than individual agents, reflecting collective sensitivity to environmental change. Crucially, analyses of synergistic information reveal that the flock encodes information about the predator’s location that is not accessible to every individual bird, demonstrating implicit collective knowledge. Together, these results show how informational coupling among active inference agents can generate new levels of autonomy and inference, providing a framework for understanding the emergence of (implicit) collective knowledge and joint agency.

💡 Research Summary

The paper investigates how collective agency and implicit knowledge can emerge from a minimal system of interacting active‑inference agents. Each “bird” in the simulation is an identical active‑inference agent that minimizes its own variational free energy by updating a belief about its heading direction based on local observations of neighboring birds. This local inference replaces rule‑based flocking algorithms with a principled Bayesian process.

Using the concept of a Markov blanket, the authors formalize statistical boundaries that separate internal states from external states via sensory and active states. When many agents become sufficiently coupled, their individual blankets nest within a higher‑level blanket that encloses the whole group, which the authors call a “flock.” The flock thus constitutes an emergent agent with its own internal, sensory, and active states, while the constituent birds retain their individual autonomy.

Two simulation experiments are presented. In the first, birds start with random headings and gradually align, illustrating the formation of the higher‑level Markov blanket over successive time steps. In the second, a predator appears at random locations, perturbing the flock. The flock responds more quickly and coherently than isolated birds, demonstrating enhanced collective sensitivity to environmental change.

Crucially, the authors apply synergistic information analysis to quantify how much information the flock holds about the predator’s position that is not present in any single bird’s observations. The flock’s synergistic information exceeds the sum of individual contributions, indicating that the group encodes implicit “collective knowledge” about the predator’s location. This knowledge is operational: it enables the flock to select more efficient escape trajectories than any individual could compute alone.

The study shows that hierarchical Markov blankets provide a rigorous way to describe nested autonomy, and that purely statistical coupling among active‑inference agents can give rise to emergent joint agency without explicit “we‑representations” in the agents’ internal models. The authors discuss broader implications for neuroscience (e.g., how neuronal populations might generate collective predictions), robotics (swarm coordination), and social science (emergence of shared beliefs). Overall, the work bridges dynamical systems, information theory, and active inference to offer a formal account of how joint agency and collective knowledge can arise from simple, locally interacting agents.