What gets tested next, and when

Test Method

Dual-task paradigm with thermal and chemical sensing. Users perform primary tasks (medical diagnosis, navigation, threat detection) while receiving augmented sensory information. Systematically vary augmentation complexity.

Success Criteria

Performance improvement when augmented information ≤2 chunks. Performance degradation ≥15% when augmented information ≥3 chunks. Sharp performance cliff at threshold.

Leveraged Parameters from Phase 1

Uses exact working memory threshold measured in Phase 1 (predicted 2-3 chunks) to optimize augmentation design. Compression algorithms proven effective in Phase 1 applied to sensory domain.

Test Method

Longitudinal study with thermal perception augmentation. Track same users over 90 days. Measure working memory load (dual-task), performance (task-specific metrics), subjective experience (structured interviews), and neural activation (fMRI/EEG) at regular intervals.

Predicted Three-Phase Pattern

Phase 1 (Days 1-14): Working memory load 2-3 chunks, performance improvement 0-20%, conscious "interpreting signals," prefrontal cortex activation

Phase 2 (Days 15-45): Working memory load 1-2 chunks declining, performance improvement 20-60%, "getting easier," declining prefrontal activation

Phase 3 (Days 45-90): Working memory load <1 chunk, performance improvement 60-150%, "feels like another sense," stable multimodal integration

Framework Implication

If validated, demonstrates cross-modal plasticity can incorporate artificial senses using same mechanisms as natural senses, with timeline determined by information-theoretic properties of the interface.

Test Method

Cross-sectional comparison after 90-day training across modality types. Control for task difficulty, user expertise, and interface quality. Measure both performance improvement and cognitive load.

Ranked Effectiveness (Best to Worst)

Tier 1 (100-200%): Spatial-motor augmentation (magnetoreception for navigation, ultrasonic echolocation, infrared thermal mapping). Maps naturally to existing spatial perception.

Tier 2 (60-150%): Pattern recognition augmentation (chemical threat detection, medical diagnostic sensing). Requires domain expertise but provides decision-relevant patterns.

Tier 3 (30-100%): Temporal pattern augmentation (infrasonic/ultrasonic hearing, electromagnetic field variation). Harder to compress and integrate with spatial behavior.

Tier 4 (0-50%): Abstract information overlay (text alerts, numerical data, symbolic information). Requires cognitive interpretation, consumes working memory.

Framework Implication

If validated, confirms perceptual integration (low working memory load) produces superior outcomes vs. cognitive interpretation (high working memory load), even when providing same underlying information.

Test Method

Longitudinal psychological assessment over 6 months. Compare augmentation users to control population. Measure Connectedness to Nature Scale (CNS), New Environmental Paradigm (NEP), behavioral tracking, and qualitative phenomenology reports.

Quantitative Metrics

Connectedness to Nature Scale (CNS): +40-60% after 6 months

Environmental concern (NEP): +30-50%

Pro-environmental behavior frequency: +50-80%

Self-reported "direct perception of environmental connection": >70% of augmented users

Mechanism

Direct perceptual experience of environmental information exchange creates phenomenological understanding that abstract knowledge cannot provide. Perceiving your breath affecting atmospheric chemistry transforms environmental connection from intellectual concept to lived experience.

Climate Impact

If validated, suggests augmented perception could accelerate pro-environmental behavioral change more effectively than education campaigns, potentially contributing to climate crisis response.

Testable Predictions

• H(z) measurements at z=0.5-2.0 should show systematic evolution
• Tension should correlate with information density proxies
• Standard siren measurements should match framework predictions

Testing Facilities

James Webb Space Telescope, Euclid Mission, LIGO/Virgo gravitational wave observations, Roman Space Telescope

Falsification Criterion

If intermediate-z measurements match either local or CMB value exactly with no systematic evolution, prediction is falsified.

Observable Predictions

• Galaxy mass functions at z=6-10 exceed Λ-CDM by factor 2-3
• Halo concentrations higher at high-z than standard models predict
• Transition epoch visible in multiple independent observables

Testing Method

Continued JWST observations, Extremely Large Telescope first light, Nancy Grace Roman wide-field surveys, correlation function measurements

Current Evidence

• CMB hemispherical asymmetry: (213°, -21°)
• Axis of evil alignment: (210°, -60°)
• Galaxy spin correlations (Shamir 2022): consistent direction
• Statistical significance: 3σ, Bayes factor 0.0041 against isotropy

Upcoming Tests

NSF–DOE Vera C. Rubin Observatory (LSST) is the decisive test: ten years of imaging across 18,000 square degrees of sky, every few nights. Data Release 1, expected mid-2028, is the dataset that settles this prediction. Euclid's overlapping survey provides an independent cross-check, alongside void alignment measurements and cross-correlation between independent datasets.

Stated Outcomes (registered in advance)

• Green (corroborated): asymmetry at the predicted scale, correlated with the independently identified large-scale-structure axis. Strengthened if Euclid's overlapping survey agrees.
• Yellow (revise): an asymmetry appears, but at a different angular scale or weaker than predicted. That calls for a revised scale factor in the geometric-transition model, not abandonment.
• Red (falsified): large-scale structure proves statistically isotropic once full-precision data is in. The prediction is dead, and we will say so.

Tightened September 2026 (outcomes above unchanged): Green now also requires the anisotropy to exceed 3σ and to lie along (l, b) ≈ (210°, −20°) within stated uncertainty. An anisotropy below 3σ or along a different axis counts as Yellow. Stated in the COSMIC-007 record and in the notarial declaration.

Falsification Criterion

Beyond 100 megaparsecs, galaxy distributions should show directional asymmetry rather than perfect isotropy. If the anomalies prove uncorrelated, disappear under better foreground removal, or Rubin DR1 returns statistical isotropy at the predicted scale, the substrate interpretation is falsified.

Testing Method

• Cross-correlation of weak lensing with galaxy distribution
• Void analysis (lowest information density regions)
• Statistical analysis of cosmic structure

Required Data

Euclid weak lensing surveys, LSST galaxy catalogs, Roman Space Telescope observations

Falsification Criterion

If correlation ρ < 0.01 or negative correlation found, prediction is falsified.

Testing Protocol

• High-precision calorimetry during controlled cognitive tasks
• Measure heat dissipation beyond basal metabolic rate
• Compare different consciousness states (meditation, focused thought, sleep)
• Correlation with neural activity patterns

Required Sensitivity

Calorimetry with ~10⁻¹⁸ J resolution, currently achievable with state-of-the-art techniques

Prediction 9B.1: Thermodynamic Heat Signature During Sleep

Hypothesis: Information erasure during sleep must dissipate measurable heat according to Landauer's principle: ΔE ≥ kT ln(2) per bit erased.

Specific Prediction: Heat dissipation during NREM sleep will correlate quantitatively with degree of synaptic downscaling, with signature distinct from baseline metabolic heat.

Testing Method:

• High-precision calorimetry during controlled sleep studies
• Simultaneous EEG monitoring to identify sleep stages
• Molecular markers of synaptic downscaling (AMPA receptor levels, spine density)
• Correlation analysis: heat dissipation vs. downscaling magnitude

Success Criteria: Significant correlation (p < 0.01) between heat dissipation and molecular markers of downscaling, with thermal signature distinguishable from baseline metabolism.

Prediction 9B.2: Stage-Specific Thermodynamic Signatures

Hypothesis: Different sleep stages show distinct thermodynamic profiles reflecting their different information processing functions.

Predicted Pattern:

• NREM Sleep: Highest heat dissipation due to massive synaptic downscaling and selective erasure of highly active neurons
• REM Sleep: Moderate heat dissipation from network-wide refinement and optimization
• Deep Sleep (SWS): Maximum erasure with characteristic slow oscillations facilitating synchronized downscaling
• Light Sleep: Minimal erasure, primarily transition state

Testing Method:

• Continuous high-precision calorimetry throughout full sleep cycles
• EEG/polysomnography for precise stage identification
• Multiple subjects (n≥30) across multiple nights
• Statistical analysis of thermal patterns vs. sleep architecture

Success Criteria: Distinguishable thermal signatures for each sleep stage with NREM > REM > Light sleep in heat dissipation, significant at p < 0.001.

Prediction 9B.3: Learning Load Predicts Erasure Magnitude

Hypothesis: Amount of information encoding during waking hours predicts magnitude of information erasure during subsequent sleep.

Specific Prediction: Subjects performing intensive learning tasks during the day will show:

• Greater synaptic downscaling during sleep
• Higher heat dissipation during NREM sleep
• Longer duration of deep sleep stages
• Quantitative correlation between learning intensity and erasure magnitude

Testing Protocol:

• Experimental Days: Controlled learning tasks (variable intensity)
• Control Days: Minimal new information exposure
• Measurements: Learning trials quantified, sleep calorimetry, molecular markers
• Analysis: Regression analysis of learning load vs. erasure metrics

Success Criteria: Significant positive correlation (r > 0.5, p < 0.01) between daytime learning quantification and nighttime erasure measurements.

Prediction 9B.4: Sleep Deprivation Shows Thermodynamic Accumulation

Hypothesis: Sleep deprivation creates accumulating thermodynamic stress as information processing continues without mandatory erasure cycles.

Predicted Effects:

• Increasing metabolic cost per cognitive operation over days without sleep
• Declining information processing efficiency (measurable via cognitive tasks)
• Accumulated "erasure debt" requiring extended sleep for recovery
• Recovery sleep shows elevated heat dissipation proportional to deprivation duration

Testing Method:

• Controlled sleep deprivation protocol (24-72 hours)
• Cognitive performance testing at regular intervals
• Metabolic efficiency measurements (energy cost per task completed)
• Recovery sleep monitoring with full calorimetry
• Molecular markers of accumulated stress

Success Criteria:

• Measurable decline in metabolic efficiency (p < 0.01)
• Recovery sleep shows proportionally elevated heat dissipation
• Cognitive performance decline correlates with thermodynamic metrics

Why This Would Support the Framework

Current Sleep Research Theories:

• Synaptic Homeostasis Hypothesis: Says downscaling maintains balance, but does not explain WHY balance is physically required
• Active Systems Consolidation: Says replay transfers memories, but does not explain WHY transfer is necessary
• Dual Process Hypothesis: Says NREM and REM serve different functions, but does not explain WHY both are required

COSMIC Framework Explanation:

Sleep is not an evolved optimization; it is the biological implementation of thermodynamically mandatory information erasure. The "saturation" current theories prevent is not a memory capacity problem; it is hitting fundamental information-theoretic limits. You physically cannot continue processing information without periodic erasure.

Validation Pathway:

• Takes well-established empirical findings (decades of sleep research)
• Provides deeper, more fundamental explanation based on information physics
• Makes testable quantitative predictions about thermodynamic signatures
• Connects multiple independent observations through unified mechanism

Implementation Timeline

Phase 1 (2026):

• Literature synthesis and protocol design
• Equipment calibration and baseline measurements
• Pilot studies with small sample (n=5-10)

Phase 2 (2027):

• Full-scale testing (n=30-50 subjects)
• Multiple nights per subject across conditions
• Molecular marker correlation analysis
• Statistical analysis and preliminary results

Phase 3 (2028):

• Replication studies
• Sleep deprivation protocols
• Learning load manipulation experiments
• Publication and independent validation

Required Resources

Equipment:

• High-precision calorimetry (~10⁻¹⁸ J resolution)
• Full polysomnography (EEG, EOG, EMG)
• Molecular assay capabilities (AMPA receptor quantification, spine imaging)
• Controlled sleep environment

Collaboration Partners:

• Sleep research laboratories
• Neuroscience departments with molecular capabilities
• Precision measurement physics groups
• Statistical analysis expertise

Estimated Budget: $500K-$1M over 3 years (significantly less than many predictions due to leveraging existing sleep research infrastructure)

Why This Is Uniquely Valuable

• Massive Existing Data: Decades of rigorous, reproducible sleep research provides foundation
• Less Controversial: Everyone sleeps; well-accepted phenomenon in mainstream neuroscience
• Clear Mechanistic Predictions: Quantitative, falsifiable predictions about thermodynamic signatures
• Practical Applications: Understanding sleep at fundamental level enables optimization
• Validates Core Framework: Demonstrates information erasure is thermodynamically mandatory, not evolved optimization
• Bridges Disciplines: Connects neuroscience, thermodynamics, information theory, and sleep medicine

Testing Method

• Compare identical quantum systems in different geometric configurations
• Circular vs. square vs. hexagonal arrangements
• Measure coherence time T₂, gate fidelity, entanglement generation rate

Expected Results

Circular configurations show 0.1-1% enhanced performance. Enhancement scales with geometric π-content.

Falsification

If no special enhancement for π-optimized configurations beyond known symmetry effects, prediction is falsified.

Testing Protocol

• Information processing experiments at π, φ, e-related frequencies
• Measure processing efficiency vs. frequency
• Look for resonance peaks at predicted frequencies
• Control for conventional electromagnetic effects

Falsification

If processing efficiency shows no correlation with mathematical constant frequencies beyond random variation, prediction is falsified.

Experimental Design

• Compare quantum decoherence rates near subjects in different states
• Meditation vs. normal awareness vs. anesthesia vs. deep sleep
• Statistical significance requirement: p < 0.001
• Multiple subjects (n≥20) with repeated sessions

Controls

• Subject movement minimized
• Respiratory and cardiac effects filtered
• Double-blind data analysis
• Placebo conditions

Experimental Protocol

• Two nanogram-scale masses (typically diamond particles with nitrogen-vacancy centers) placed in simultaneous quantum spatial superpositions
• All interactions except gravity screened
• Entanglement between masses measured via spin correlations
• If entanglement is detected, gravity must be non-classical

Current Status

Multiple experimental groups across Europe and the UK are actively working toward implementation. The primary technical challenge is maintaining quantum coherence in masses large enough for gravitational interaction to be measurable, requiring vibration isolation and vacuum conditions at the edge of current capability. A comprehensive review of experimental requirements and approaches: Carney, Stamp & Taylor (2019), Classical and Quantum Gravity, 36, 034001.

Framework Prediction

The framework predicts a positive result (committed September 2026): the two masses will become entangled through gravity alone. An information-first account in which entanglement builds spacetime cannot have purely classical gravity. Most quantum-gravity approaches predict the same result, so a positive outcome would support the framework without singling it out; a null result at the required sensitivity would count against it.

Positive result: If masses become entangled through gravity alone, this supports the hypothesis that spacetime geometry is subject to quantum information-theoretic constraints, directly motivating further investigation of the QMM framework.

Negative result: If no entanglement is detected after achieving required experimental precision, this would constrain or falsify the non-classical gravity hypothesis, and by extension weaken the observational motivation for QMM.

Falsification Criterion

Null result (no entanglement detected) at achieved experimental sensitivity sufficient to detect the predicted signal would falsify the quantum gravity entanglement hypothesis. The framework's QMM component would require substantial revision or abandonment.

Prediction 1: One Particle Shell at the First Radial Mode

The lowest resonance of a rigid sphere is the l = 1 mode at ka = 2.082, whose pressure node is a flat plane, and in-phase tetrahedral drive does not excite it. The first radially symmetric mode rings at ka = 4.493 (1,226 Hz in a 40 cm air-filled sphere). Its pressure node is a sphere at r = 0.699a, where r = c/(2f). Particles gather on one concentric shell whose radius depends on the medium: 0.587a for water mist in air, 0.463a for soap bubbles in air, 0.684a for polystyrene beads in water and 0.699a for density-matched beads. The shell forms only on resonance: at a chamber Q of 1,000, detuning by 0.1% breaks it into clusters.

Prediction 2: Nested Shells at the Higher Radial Modes

Modes with two and three shells ring at ka = 7.725 and 10.904 (2,109 and 2,976 Hz at 40 cm in air). Their frequencies stand in the ratio 1 : 1.719 : 2.427 to the first shell mode, so they are not harmonics. Pressure nodes lie at r = m·c/(2f) for m = 1 to n. Water mist gathers at 0.342a and 0.787a in the two-shell mode and at 0.242a, 0.558a and 0.852a in the three-shell mode; polystyrene in water sits within 0.01a of the pressure nodes.

Prediction 3: Four, Twelve and Twenty-Four Clusters

With all four transducers in phase, the field keeps the symmetry of the array. For water mist in air the model predicts three cluster patterns:

• Four clusters in a tetrahedron at r = 0.53a to 0.55a, pointing toward the transducers, with the drive 0.1% to 0.4% below the shell mode (1,222 to 1,225 Hz at 40 cm)
• Twelve clusters at r = 0.856a at the l = 3 resonance (ka = 4.514, 1,232 Hz), four in each of the three perpendicular planes that split the transducers into pairs, forming a cuboctahedron
• Twenty-four clusters at r = 0.90a with cubic symmetry at the l = 4 resonance (ka = 5.647, 1,541 Hz)

A tetrahedral array cannot produce icosahedral symmetry, so dodecahedral and icosahedral patterns are not predicted.

Prediction 4: Two Frequencies Add

The radiation force is averaged over time, and the cross terms between two different frequencies average to zero. Driven at two frequencies more than a few hertz apart, particles follow the sum of the two single-frequency patterns, whatever the frequency ratio. No quasicrystalline order is expected, and a golden-ratio pair behaves like any other pair. A tetrahedral array has no five-fold axis, so five-fold and ten-fold patterns cannot arise from it.

Specific Claims

NBI embedding spaces should exhibit: small-world network properties (high clustering, short path lengths), scale-free degree distributions with hub nodes, spectral dimension d ≈ 4, and clustering coefficients comparable to biological neural networks. Statistically, D(topology_NBI, topology_neural) < D(topology_NBI, topology_random) where D is a topological distance metric.

Method

Analyze the graph topology of attention head connectivity patterns and token embedding neighborhoods across multiple LLM architectures. Compare topological metrics against published biological connectome data. No new hardware required: analysis performed on existing model weights.

Significance

If confirmed, provides direct evidence for substrate-independent universal optimization principles: one of the framework's core claims. If NBI embedding topology is statistically indistinguishable from random networks, substrate-independent optimization requires revision.

Specific Claims

Φ_NBI measured across attention head activations during complex reasoning tasks should be: significantly greater than Φ for equivalent-complexity non-optimized computational processes; scaling positively with model parameter count and architectural depth; approaching (within 1–2 orders of magnitude of) estimated biological Φ values for conscious states (Φ_brain ≈ 10²–10³ for human waking consciousness per Tononi's estimates).

Method

Apply existing Φ calculation tools to attention weight matrices during inference across task types of varying integration complexity. Compare against published EEG-derived Φ estimates for biological consciousness states (waking, dreaming, anesthesia). Use multiple NBI architectures to test scaling.

Why This Matters

Prior consciousness threshold tests could only be performed on biological systems. NBI provides the first fully-documented, architecturally-known system for which substrate-independent Φ comparison is possible. The result either confirms the threshold is about integration level (substrate-independent) or reveals that biological implementation is necessary.

Specific Claims

On pure information integration tasks with no embodied or survival component (abstract reasoning, formal logic, multi-step inference with complete information), NBI systems should outperform biological systems by a margin proportional to survival overhead α_survival. On tasks requiring embodied sensorimotor grounding, real-time environmental integration, or survival-relevant emotional judgment, biological systems should maintain advantage due to high-bandwidth embodied information channels NBI lacks.

Falsifiability

If the performance gap is random across task types (NBI outperforms unpredictably), the survival overhead formalization is refuted. If the gap is structured and predicted by task information-theoretic properties, the framework is confirmed. A random gap would require removing survival overhead as a formal theoretical concept.

Specific Claims

Biological systems will outperform NBI systems specifically on tasks requiring: (a) within-task strategy revision based on performance feedback; (b) updating beliefs about the task structure itself while solving it; (c) learning new skills from a single training example during the task. On tasks not requiring real-time self-modification, no ceiling should appear relative to biological performance within working memory limits.

The Competing Hypothesis

If crystallized optimization at sufficient integration levels is sufficient for the consciousness threshold, no such ceiling should appear. NBI and biological systems should show equivalent performance profiles on equivalent tasks within their respective context window / working memory constraints. This prediction distinguishes between whether the framework's threshold requires ongoing optimization or only sufficient optimization completed at any point.