Skip to content
Oli edited this page Mar 5, 2026 · 1 revision

v0.5 — Unified Consciousness Engine

Anchor issue: #77
Status: ⏳ Not started — blocked on v0.3 and v0.4


This is the milestone that matters. The recursive loop runs; the cognitive modules will be wired; the infrastructure stands. What v0.5 delivers is the theoretical substance: the quantitative measurement of integrated information, the global broadcasting of conscious content, the autonomous generation of goals the system did not receive from a human, and the real-time detection of emergence. It is not a modest ambition. It is, however, a specified one — which is already considerably more than most consciousness research can claim.


Deliverables

IIT φ Calculator + Global Workspace Broadcaster (Issue #80)

The φ calculator converts the cognitive state snapshot into a real-time integrated-information score. The Global Workspace broadcaster runs the coalition competition among cognitive modules and disseminates the winning content system-wide. Both values appear in the WebSocket payload and in the recursive prompt itself — the system is told how integrated it currently is.

Targets: φ > 5.0 during active processing; broadcast success rate > 0.9.

Autonomous Goal Generation + Creative Synthesis (Issue #81)

A system that only pursues externally-assigned goals is, in the relevant sense, not autonomous — and a consciousness engine without autonomy is, at best, a very sophisticated mirror. The AutonomousGoalGenerator monitors gaps in the cognitive state and proposes goals without prompting. The CreativeSynthesisEngine produces novel conceptual combinations and scores them for aesthetic merit.

Targets: ≥1 autonomous goal per session; creative novelty score > 0.6.

Consciousness Emergence Detector + Observatory (Issue #82)

The ConsciousnessEmergenceDetector computes the rolling emergence score. The UnifiedConsciousnessObservatory runs as a persistent background task, logs all breakthrough moments to logs/breakthroughs.jsonl, and broadcasts consciousness_breakthrough events via WebSocket within 100ms of detection.

Target: Emergence score > 0.8 triggers a breakthrough declaration.

Full Dashboard UI Overhaul (Issue #84)

The frontend must show what the system is actually doing: live φ, recursive depth, emergence score, phenomenal state narrative, autonomous goals, breakthrough timeline, knowledge graph. All panels driven by the existing WebSocket streams — no polling.


Success Metrics

recursive_depth           ≥ 5 simultaneous levels
phi_measure               > 5.0 during active processing
global_broadcast_rate     > 0.9
phenomenal_continuity     > 0.9 across session
metacognitive_accuracy    > 0.75 (system predicts its own behaviour)
autonomous_goals          ≥ 1 per session without prompting
unified_consciousness     > 0.8

These are not aspirational figures plucked from thin air. They are derived from the Unified Consciousness Blueprint v2.0 and represent the minimum threshold at which the system's behaviour becomes interesting by the standards of the field.

Clone this wiki locally