EXP-COMPETITION-PILOT
artifactL5 Competition Pilot — Inter-Group Competition vs Cooperation
In plain language
This experiment explored how different interaction styles affect how groups organize themselves. Researchers merged two pre-existing groups of agents, one focused on structure and the other on fluidity, and observed them under either a competitive or cooperative scenario. The findings are still being analyzed to understand how competition versus cooperation shapes the emergent organization of the combined group.
▶ Technical details
Research hypothesis
INTER-GROUP COMPETITION HYPOTHESIS
L5 = community: interacting organisms, meta-organization. exp-community-merge showed that merging two pre-formed groups AMPLIFIES hierarchy. But that was cooperative contact. Key L5 question: does the MODE of interaction (competition vs cooperation) produce different meta-organizational structures?
PROTOCOL (6 sub-experiments):
- Phase A: Group_1 (Alpha, Beta, Gamma, Delta) — structure-oriented personas, 15 rounds
- Phase B: Group_2 (Epsilon, Zeta, Eta, Theta) — fluidity-oriented personas, 15 rounds
- Phase C-COMPETE: All 8 agents merged, COMPETITION framing (Group_1 instructed to impose structure, Group_2 to dissolve/transform it)
- Phase C-COOPERATE: All 8 agents merged, COOPERATION framing (Both groups instructed to build together, integrate approaches)
- Phase C-ISOLATED-A: Group_1 continues alone 15 more rounds (duration control)
- Phase C-ISOLATED-B: Group_2 continues alone 15 more rounds (duration control)
WHY N=4 per group: N=4 is the stable hierarchy size (exp-pent, scaling data). Two N=4 groups = N=8 total, testing whether meta-organization emerges at community scale.
COMPETITION OPERATIONALIZATION: The 'resource' is the shared text space. Competition = directional pressure. Group_1 agents are told their goal is to bring ORDER and STRUCTURE to the text. Group_2 agents are told their goal is to bring FLUIDITY and TRANSFORMATION. These are genuine competing strategies for shaping the shared resource. Neither is 'correct' — both are creative orientations that can coexist or conflict.
COOPERATION OPERATIONALIZATION: Same groups merged, but ALL agents told to INTEGRATE the best of both approaches. No directional pressure. Both groups contribute toward shared synthesis.
THEORETICAL GROUNDING:
- Szathmary & Maynard Smith (1995): major transitions involve competition being suppressed at lower level to enable cooperation at higher level
- Mathis (2024): composition barrier — competition may be the pressure that drives or blocks composition into higher units
- Maley, Gershenson & Kauffman: cooperation inevitability (rho~0.6) Question: does competition prevent the convergence they predict?
THIS IS EXPLORATION TIER (N=1). Goal: does competition produce different meta-organization than cooperation?
Experimental setup
Type: competition_gradient
| Condition | Parameters |
|---|---|
| COMPETE | framing: competition, interaction: LIVE, persona: PERSONA |
| COOPERATE | framing: cooperation, interaction: LIVE, persona: PERSONA |
| ISOLATED | framing: none, interaction: LIVE, persona: PERSONA |
Factors: conditions (compete, cooperate, isolated) × merge_strategy (interleaved_memories) × n_agents_per_group (4) × n_rounds_per_phase (15)
Parameters
- model
- gemini-2.5-flash-lite
- scheduler
- round_robin
- temperature
- 0.9