collaborative-emergent-intelligence

Mainstream AI optimizes for accuracy and systematically suppresses low-probability outputs. Yet those suppressed directions are precisely where novel cognition tends to emerge. Collaborative Emergent Intelligence inverts this structure: the algorithm stops filtering and instead exhaustively diverges candidates, leaving final selection to the human who adjudicates by intuition and personal goal.

The plugin ships 13 tools, five-level resonance probing (from manual to four-track parallel auto with temporal resonance), a tri-state phase machine, and three adjudication pop-ups. Code-level safety valves enforce the workflow: no condensation without blindspot surfacing, no overwriting newer memory with older archives, and a breadth floor gate against pseudo-depth. Nine data archives are dual-written for resilience; overflow only silences into an archive zone, never splicing away data.

Best suited for thinkers navigating ambiguous directions, adjudicating between mutually exclusive paths, or resisting premature closure. Not intended for millisecond fact lookups or contexts that do not benefit from expanded search space.