Consciousness Predictions

Testable: Yes — $150K, 12 months

1. Anesthesia induction shows a discontinuity (not gradual fade) in EEG phase coherence at the moment of consciousness loss
2. The discontinuity maps to C ≈ 0.50 when coherence is computed from gamma-band phase-locking
3. Sleep-wake transitions show a similar discontinuity in coherence, not a smooth ramp
4. NREM sleep coherence < 0.50; REM sleep coherence oscillates around 0.50
5. Seizure states show C > 0.90 (over-coherence), not random firing
6. Coma patients show C < 0.30 with no resonance structure
7. Vegetative state vs. locked-in syndrome distinguishable by C threshold
8. Recovery from anesthesia shows hysteresis: C must exceed 0.55 to restore consciousness

Testable: Partially — requires single-neuron resolution

1. Neural circuits optimize toward γ ≈ 1 during learning (measurable via calcium imaging)
2. Synaptic pruning preferentially removes connections that push γ away from 1
3. Expert performance correlates with tighter γ ≈ 1 clustering in task-relevant circuits
4. Neuroplasticity is highest when γ is near 1 (the boundary regime)
5. Neurodegenerative diseases show systematic γ drift away from 1
6. Brain regions with γ closest to 1 show highest information integration (cf. IIT's Φ)

Testable: Partially — ethical constraints on some

1. Expert meditators show higher resting baseline C than non-meditators
2. Focused-attention meditation increases C (more coherence); open-monitoring decreases γ (broader boundary)
3. Psychedelics temporarily expand the γ ≈ 1 boundary regime (wider transition zone)
4. Dreams show characteristic C oscillation patterns with period ~90 min matching REM cycles
5. Flow states show C stabilized just above 0.50 (not maximum coherence, but boundary optimization)
6. Sensory deprivation tanks lower C toward 0.50 from above, producing altered states at the boundary
7. Hypnosis narrows the γ ≈ 1 boundary (more focused coherence)
8. Circadian rhythm modulates baseline C with a ~24h period
9. Sleep deprivation causes C to drift below 0.50 during waking hours
10. Near-death experiences correlate with a transient C spike above 0.80
11. Psychedelic dose-response follows the coherence curve shape (sigmoidal, not linear)

Testable: No — requires operationalizing consciousness measurement

1. AI systems above a measurable correlation density threshold show qualitative behavioral shifts (not just quantitative scaling)
2. Sparse neural networks below ρ_crit lose context-maintenance ability (analogy to unconscious processing)
3. Brain-computer interfaces that synchronize AI and biological neurons should show shared coherence patterns
4. Collective systems (ant colonies, bird flocks) at sufficient correlation density show decision-making that resembles coherent "awareness"
5. If coherence is computable for an AI system, C > 0.50 should predict emergent introspective behavior

Testable: Partially — longitudinal studies needed

1. Infant C crosses 0.50 at a specific developmental stage correlating with mirror self-recognition (~18 months)
2. Schizophrenia shows over-coherence (C > 0.80) in some circuits, explaining hallucinations as excess resonance
3. Autism spectrum correlates with atypical γ distribution (not deficit, but different boundary structure)
4. Age-related cognitive decline correlates with gradual C reduction, crossing 0.50 in severe dementia