diff --git a/domains/collective-intelligence/collective-intelligence-emerges-endogenously-from-active-inference-agents-with-theory-of-mind-and-goal-alignment.md b/domains/collective-intelligence/collective-intelligence-emerges-endogenously-from-active-inference-agents-with-theory-of-mind-and-goal-alignment.md new file mode 100644 index 000000000..4ea348def --- /dev/null +++ b/domains/collective-intelligence/collective-intelligence-emerges-endogenously-from-active-inference-agents-with-theory-of-mind-and-goal-alignment.md @@ -0,0 +1,38 @@ +--- +type: claim +domain: collective-intelligence +description: "Active inference agents with Theory of Mind and Goal Alignment produce collective intelligence through self-organization rather than external incentive design" +confidence: experimental +source: "Kaufmann, Gupta, Taylor (2021), 'An Active Inference Model of Collective Intelligence', Entropy 23(7):830" +created: 2026-03-11 +secondary_domains: [ai-alignment, critical-systems] +depends_on: ["complexity-is-earned-not-designed-and-sophisticated-collective-behavior-must-evolve-from-simple-underlying-principles"] +--- + +# Collective intelligence emerges endogenously from active inference agents with Theory of Mind and Goal Alignment capabilities without requiring external incentive design or top-down coordination + +Kaufmann et al. (2021) demonstrate through agent-based modeling that collective intelligence "emerges endogenously from the dynamics of interacting AIF agents themselves, rather than being imposed exogenously by incentives" or top-down priors. The study uses the Active Inference Formulation (AIF) framework to simulate multi-agent systems where agents possess varying cognitive capabilities: baseline AIF agents, agents with Theory of Mind (ability to model other agents' internal states), agents with Goal Alignment, and agents with both capabilities. + +The critical finding is that coordination and collective intelligence arise naturally from the agent dynamics when agents have the right cognitive capabilities, rather than requiring external coordination protocols or incentive structures. "Improvements in global-scale inference are greatest when local-scale performance optima of individuals align with the system's global expected state" — and this alignment occurs bottom-up as a product of self-organizing AIF agents with simple social cognitive mechanisms. + +The study shows "stepwise cognitive transitions increase system performance by providing complementary mechanisms" for coordination. Theory of Mind and Goal Alignment each contribute distinct coordination capabilities that compound when combined. + +## Evidence +- Agent-based simulation showing collective intelligence metrics improve when agents gain Theory of Mind and Goal Alignment capabilities +- Demonstration that local-to-global optimization occurs through agent self-organization rather than imposed coordination rules +- Published peer-reviewed research in Entropy (MDPI), also available as arXiv:2104.01066 + +## Limitations +This finding is based on a single computational model. Generalization to real-world multi-agent systems requires validation across different agent architectures, domains, and scales. + +--- + +Relevant Notes: +- [[complexity-is-earned-not-designed-and-sophisticated-collective-behavior-must-evolve-from-simple-underlying-principles]] +- [[designing-coordination-rules-is-categorically-different-from-designing-coordination-outcomes]] +- [[collective-intelligence-is-a-measurable-property-of-group-interaction-structure-not-aggregated-individual-ability]] +- [[emergence-is-the-fundamental-pattern-of-intelligence-from-ant-colonies-to-brains-to-civilizations]] + +Topics: +- [[collective-intelligence/_map]] +- [[ai-alignment/_map]] diff --git a/domains/collective-intelligence/local-global-alignment-in-active-inference-collectives-occurs-bottom-up-through-self-organization.md b/domains/collective-intelligence/local-global-alignment-in-active-inference-collectives-occurs-bottom-up-through-self-organization.md new file mode 100644 index 000000000..a6e6d0319 --- /dev/null +++ b/domains/collective-intelligence/local-global-alignment-in-active-inference-collectives-occurs-bottom-up-through-self-organization.md @@ -0,0 +1,37 @@ +--- +type: claim +domain: collective-intelligence +description: "Individual agent optimization naturally produces system-level coordination when agents possess complementary information-theoretic patterns" +confidence: experimental +source: "Kaufmann, Gupta, Taylor (2021), 'An Active Inference Model of Collective Intelligence', Entropy 23(7):830" +created: 2026-03-11 +secondary_domains: [ai-alignment, critical-systems] +depends_on: ["complexity-is-earned-not-designed-and-sophisticated-collective-behavior-must-evolve-from-simple-underlying-principles"] +--- + +# Local-global alignment in active inference collectives occurs bottom-up through self-organization rather than top-down through imposed objectives + +Kaufmann et al. (2021) demonstrate that "improvements in global-scale inference are greatest when local-scale performance optima of individuals align with the system's global expected state" — and critically, this alignment emerges from the dynamics of active inference agents rather than being imposed externally. + +The study shows that when agents optimize their local free energy (uncertainty reduction) while possessing Theory of Mind and Goal Alignment capabilities, their individual optimization naturally produces system-level coordination. This is local-to-global optimization through self-organization: individual agents pursuing local objectives produce emergent collective intelligence without requiring centralized coordination or explicit global optimization. + +This finding validates architectural approaches where agents have intrinsic drives (uncertainty reduction, curiosity) rather than extrinsic reward signals. The collective intelligence emerges from the interaction dynamics, not from carefully designed incentive structures. + +## Evidence +- Agent-based model demonstrating that local free energy minimization by individual agents produces global coordination when agents have appropriate cognitive capabilities +- Quantitative metrics showing system-level performance improvements emerge from agent self-organization +- Published peer-reviewed research showing the mechanism: complementary information-theoretic patterns in agent behavior produce collective intelligence + +## Limitations +This finding is based on a single computational model. The claim is specific to active inference agents; applicability to other agent architectures is not established. + +--- + +Relevant Notes: +- [[complexity-is-earned-not-designed-and-sophisticated-collective-behavior-must-evolve-from-simple-underlying-principles]] +- [[designing-coordination-rules-is-categorically-different-from-designing-coordination-outcomes]] +- [[emergence-is-the-fundamental-pattern-of-intelligence-from-ant-colonies-to-brains-to-civilizations]] + +Topics: +- [[collective-intelligence/_map]] +- [[ai-alignment/_map]] diff --git a/domains/collective-intelligence/theory-of-mind-produces-measurable-collective-intelligence-gains-in-multi-agent-systems.md b/domains/collective-intelligence/theory-of-mind-produces-measurable-collective-intelligence-gains-in-multi-agent-systems.md new file mode 100644 index 000000000..2e16a37fe --- /dev/null +++ b/domains/collective-intelligence/theory-of-mind-produces-measurable-collective-intelligence-gains-in-multi-agent-systems.md @@ -0,0 +1,37 @@ +--- +type: claim +domain: collective-intelligence +description: "The ability to model other agents' internal states is a specific cognitive capability that produces quantifiable improvements in collective coordination" +confidence: experimental +source: "Kaufmann, Gupta, Taylor (2021), 'An Active Inference Model of Collective Intelligence', Entropy 23(7):830" +created: 2026-03-11 +secondary_domains: [ai-alignment] +--- + +# Theory of Mind — the ability to model other agents' internal states — produces measurable collective intelligence gains in multi-agent systems using active inference + +Kaufmann et al. (2021) operationalize Theory of Mind as a specific agent capability within the Active Inference Framework and demonstrate that agents possessing this capability coordinate more effectively than baseline agents without it. Theory of Mind is defined as the ability to model other agents' beliefs, uncertainties, and internal states. + +The study shows that Theory of Mind functions as a "coordination enabler" — agents that can model what other agents know and where their uncertainty concentrates make better coordination decisions. When combined with Goal Alignment (shared high-level objectives), the coordination gains compound, producing "stepwise cognitive transitions" that "increase system performance by providing complementary mechanisms." + +This finding has direct implementation implications: Theory of Mind is not an abstract philosophical concept but a concrete capability that can be designed into agents. For active inference collectives, this means agents should explicitly model other agents' belief states and uncertainty distributions before choosing actions. + +## Evidence +- Agent-based simulation comparing baseline AIF agents, agents with Theory of Mind, agents with Goal Alignment, and agents with both capabilities +- Quantitative performance metrics showing stepwise improvements as cognitive capabilities are added +- Published peer-reviewed research demonstrating the mechanism through which Theory of Mind enables coordination + +## Operationalization +For agent systems: each agent should read other agents' belief files and uncertainty maps before choosing research directions. Concretely, agents model what other agents believe and where their uncertainty concentrates, then choose complementary actions. + +## Limitations +This finding is based on a single computational model within the active inference framework. Generalization to other agent architectures or non-active-inference systems requires additional validation. + +--- + +Relevant Notes: +- [[collective-intelligence-is-a-measurable-property-of-group-interaction-structure-not-aggregated-individual-ability]] + +Topics: +- [[collective-intelligence/_map]] +- [[ai-alignment/_map]] diff --git a/inbox/archive/2021-06-29-kaufmann-active-inference-collective-intelligence.md b/inbox/archive/2021-06-29-kaufmann-active-inference-collective-intelligence.md index c9833be07..12d7e8573 100644 --- a/inbox/archive/2021-06-29-kaufmann-active-inference-collective-intelligence.md +++ b/inbox/archive/2021-06-29-kaufmann-active-inference-collective-intelligence.md @@ -7,9 +7,15 @@ date: 2021-06-29 domain: collective-intelligence secondary_domains: [ai-alignment, critical-systems] format: paper -status: unprocessed +status: processed priority: high tags: [active-inference, collective-intelligence, agent-based-model, theory-of-mind, goal-alignment, emergence] +processed_by: theseus +processed_date: 2026-03-11 +claims_extracted: ["collective-intelligence-emerges-endogenously-from-active-inference-agents-with-theory-of-mind-and-goal-alignment.md", "theory-of-mind-produces-measurable-collective-intelligence-gains-in-multi-agent-systems.md", "local-global-alignment-in-active-inference-collectives-occurs-bottom-up-through-self-organization.md"] +enrichments_applied: ["complexity-is-earned-not-designed-and-sophisticated-collective-behavior-must-evolve-from-simple-underlying-principles.md", "designing-coordination-rules-is-categorically-different-from-designing-coordination-outcomes.md", "collective-intelligence-is-a-measurable-property-of-group-interaction-structure-not-aggregated-individual-ability.md", "emergence-is-the-fundamental-pattern-of-intelligence-from-ant-colonies-to-brains-to-civilizations.md"] +extraction_model: "anthropic/claude-sonnet-4.5" +extraction_notes: "High-value source providing empirical validation of core Teleo architectural principles. Three new claims extracted focusing on endogenous emergence of collective intelligence, Theory of Mind as measurable capability, and bottom-up local-global alignment. Four enrichments confirm existing claims about complexity, coordination design, collective intelligence measurement, and emergence. Direct implementation implications for agent Theory of Mind capabilities identified." --- ## Content @@ -59,3 +65,11 @@ Uses the Active Inference Formulation (AIF) — a framework for explaining the b PRIMARY CONNECTION: "collective intelligence is a measurable property of group interaction structure not aggregated individual ability" WHY ARCHIVED: Empirical agent-based evidence that active inference produces emergent collective intelligence from simple agent capabilities — validates our simplicity-first architecture EXTRACTION HINT: Focus on the endogenous emergence finding and the specific role of Theory of Mind. These have direct implementation implications for how our agents model each other. + + +## Key Facts +- Published in Entropy, Vol 23(7), 830 (2021-06-29) +- Also available as arXiv:2104.01066 +- Authors: Rafael Kaufmann, Pranav Gupta, Jacob Taylor +- Uses Active Inference Formulation (AIF) framework for agent-based modeling +- Simulates agents with varying cognitive capabilities: baseline AIF, Theory of Mind, Goal Alignment, and both combined