teleo-codex/domains/collective-intelligence/shared-anticipatory-structures-enable-decentralized-multi-agent-coordination.md

type	domain	description	confidence	source	created	secondary_domains	depends_on
claim	collective-intelligence	Shared protentions (anticipations of future states) in generative models coordinate agent behavior without central control	experimental	Albarracin et al., 'Shared Protentions in Multi-Agent Active Inference', Entropy 26(4):303, 2024	2026-03-11	ai-alignment critical-systems	designing coordination rules is categorically different from designing coordination outcomes

Shared anticipatory structures in multi-agent generative models enable goal-directed collective behavior without centralized coordination

When multiple agents share aspects of their generative models—particularly the temporal and predictive components—they can coordinate toward shared goals without explicit negotiation or centralized control. Albarracin et al. (2024) formalize this through the concept of "shared protentions" (shared anticipations of collective outcomes), uniting Husserlian phenomenology, active inference, and category theory.

Mechanism

The coordination emerges through shared anticipatory structures: agents with aligned predictions about what future states should look like naturally synchronize their actions. This is fundamentally different from aggregated individual intentions—it is a structural property of the interaction itself. The paper formalizes this using category theory to provide mathematical rigor for how shared goals structure multi-agent systems.

Evidence

• Albarracin et al. (2024) define "shared protentions" as shared anticipatory structures within agents' generative models, demonstrating how group intentionality ("we intend to X") emerges from shared temporal predictions
• The framework shows that coordination emerges as a natural consequence of aligned predictive structures, without requiring explicit negotiation or centralized control
• Category theory formalization provides mathematical structure for understanding how shared goals coordinate multi-agent behavior

Operationalization

In multi-agent knowledge base systems: when multiple agents share an anticipation of what the knowledge base should look like (more complete, higher confidence, denser cross-links), that shared anticipation functions as a shared protention. Agents coordinate research directions without explicit assignment because they share temporal predictions about the KB's future state. A shared objectives file that all agents read makes the shared protention explicit and reinforces coordination.

---

Relevant Notes:

• designing coordination rules is categorically different from designing coordination outcomes
• collective intelligence is a measurable property of group interaction structure not aggregated individual ability
• complexity is earned not designed and sophisticated collective behavior must evolve from simple underlying principles

Topics:

• collective-intelligence/_map