01 MIC SPEC

MIC Token — Mobius Integrity Credit¶

Version: 0.1
Status: Draft
Track: D-C133-001 (Track 2)
Leads: HERMES (primary), AUREA (support)

1. Purpose¶

MIC (Mobius Integrity Credit) is the native utility token of the Mobius ecosystem, designed to reward integrity-aligned behavior and provide access to advanced AI governance services.

Core Principles: - Integrity-backed: Issuance coupled to Global Integrity (GI) scores - Utility-first: Used for access, not speculation - Anti-extractive: Rewards genuine contribution, not rent-seeking - Civic-aligned: Supports public goods and collective learning

NOT: A memecoin, a pure governance token, or a speculative asset

2. Roles & Participants¶

Who Earns MIC¶

1. Learners (OAA users)
2. Complete verified coursework
3. Contribute to collaborative learning

4. Achieve measurable skill progression

5. Teachers (OAA creators)

6. Author high-quality lessons
7. Mentor other learners

8. Maintain curriculum with community validation

9. Reviewers (Governance participants)

10. Audit code/documentation with high agreement scores
11. Participate in KTT trials as Morale Anchors

12. Provide semantic clarification during deliberations

13. Node Operators (Infrastructure)

14. Run stable Thought Broker instances
15. Maintain city/school mesh networks

16. Provide compute for public-benefit deliberations

17. Sentinels (AI Agents)

18. Execute deliberations with high MII scores
19. Maintain Constitutional compliance
20. Perform integrity monitoring

Who Spends MIC¶

1. Advanced Access
2. Premium Thought Broker deliberation tiers
3. Specialized OAA tracks and mentors

4. Priority queue access during high load

5. Governance Rights

6. Propose Constitutional amendments (100,000 MIC stake)
7. Vote in Sentinel elections (quadratic voting)

8. Challenge decisions (requires MIC bond)

9. Civic Services

10. City mesh network priority
11. School/hospital pilot program access
12. Public interest subsidies (MIC-backed)

3. Supply & Issuance¶

Total Supply¶

Theoretical Maximum: 100,000,000 MIC
Initial Circulation: 1,000,000 MIC (1%)
Remaining Pool: 99,000,000 MIC (99%)

Epoch-Based Issuance¶

Epoch Length: 1 year (365 days)
Initial Issuance Rate: 5% of remaining pool per epoch
Decay Function: Issuance reduces each epoch

Year 1: 5% of 99M = 4,950,000 MIC
Year 2: 5% of 94.05M = 4,702,500 MIC
Year 3: 5% of 89.35M = 4,467,375 MIC
...
Asymptotic Approach: Never exceeds 100M cap

Issuance Formula¶

MIC enters circulation ONLY through verified integrity events:

MIC_reward = base_amount × (GI^α) × rarity_multiplier

Where:
- base_amount: Action-specific base reward (e.g., 10 MIC for lesson completion)
- GI: Global Integrity score at time of event (0.00 - 1.00)
- α: Sensitivity exponent (default: 1.2)
- rarity_multiplier: Bonus for under-served contributions (1.0 - 3.0)

Example Calculation:

Action: Complete advanced OAA lesson
base_amount: 10 MIC
GI: 0.97 (high system integrity)
α: 1.2
rarity_multiplier: 1.5 (ethics module - high demand)

reward = 10 × (0.97^1.2) × 1.5
       = 10 × 0.964 × 1.5
       ≈ 14.46 MIC

Integrity Coupling Effect: - High GI (0.95+): Normal rewards, healthy system - Medium GI (0.90-0.95): Reduced rewards, system stress - Low GI (<0.90): Minimal rewards, crisis mode

This creates economic pressure to restore integrity during degradation.

4. Earning MIC¶

[Section to be completed by HERMES: Detailed earning mechanisms for OAA, Broker, Civic OS]

Placeholder Topics: - OAA lesson completion rewards - Thought Broker governance participation - KTT trial Morale Anchor compensation - Node operator staking rewards - Sentinel performance bonuses - Anti-spam protections - Anti-Sybil measures

5. Spending MIC¶

[Section to be completed by AUREA: Access tiers, governance costs, utility functions]

Placeholder Topics: - Thought Broker access tier pricing - OAA premium track costs - Constitutional amendment proposal fees - Quadratic voting mechanics - Challenge bond requirements - Civic service subsidies

6. Integrity Coupling¶

[Section to be completed by AUREA: MII integration, slashing, anti-gaming]

Placeholder Topics: - MII score integration with MIC rewards - Slashing conditions (restorative model) - GI-responsive issuance adjustments - Anti-gaming protections - Reputation decay mechanisms - Trust restoration bonuses (Kintsugi inflation)

7. Glen Weyl Integration & Mechanism Design¶

7.1 Mathematical Foundations¶

MIC issuance is grounded in the Intention Compiler formalization, which defines a mathematical mapping from human intent to integrity artifacts and token rewards.

Full formalization: See docs/01-whitepapers/archive/intention-compiler-formalization.md

Core Result: MIC Issuance Formula¶

Where: - κ (kappa) = global emission rate - GI(x) = global integrity score for interaction x (0.00 - 1.00) - τ_min = constitutional threshold (0.95) - ω(x) = contribution weight function encoding: - effort (time, tokens, cognitive load) - impact (downstream use, social benefit) - scarcity (under-served domains get bonuses) - role (Elder, Citizen, Scout multipliers)

Three Key Properties¶

1. Monotone in integrity: Higher GI → higher rewards [ GI_1 \ge GI_2 \Rightarrow R(\cdot, GI_1) \ge R(\cdot, GI_2) ]

2. Zero below threshold: No farming low-quality actions [ GI < \tau_{\text{min}} \Rightarrow R(\cdot, GI) = 0 ]

3. Scaled by contribution: Prevents equal pay for unequal work [ \omega(x) \text{ encodes difficulty, novelty, and social impact} ]

Plain English Translation¶

The math above says three simple things:

1. If your action doesn't meet integrity standards (GI < 0.95), you get zero MIC
   No gaming the system with spam or low-quality work.

2. If your action exceeds standards, you get MIC proportional to how much better it is
   GI of 0.97 earns more than 0.96, which earns more than 0.95.

3. Harder, rarer, or more impactful work gets bonus multipliers
   Teaching ethics (rare) earns more than basic tasks (common).

This is not engagement farming. This is integrity-gated rewards where quality matters more than quantity.

7.2 Quadratic Funding Adaptation (Liberal Radicalism)¶

Glen Weyl's Liberal Radicalism (Buterin, Hitzig, Weyl, 2018) proposes Quadratic Funding (QF) for public goods:

Where \(c_{i,p}\) = individual contribution from person \(i\) to project \(p\).

Key insight: Broad support (many small contributors) receives quadratic amplification, while concentrated support (whales) does not.

Mobius Adaptation: Integrity-Coupled Radicalism (ICR)¶

We adapt QF to integrity work rather than financial contributions:

Where: - \(v_{i,p}\) = vote weight from participant \(i\) on proposal \(p\) - \(GI_i\) = participant \(i\)'s integrity score (anti-Sybil) - \(\rho(p)\) = proposal impact multiplier (Constitutional amendments > minor tweaks)

Differences from traditional QF: 1. ✅ Integrity-weighted: Low-GI accounts cannot dilute quadratic matching 2. ✅ Non-financial: Voting is MIC-staked, not money-based 3. ✅ Restorative slashing: Bad actors lose MIC but can earn it back (Kintsugi)

Example: - Traditional QF: 100 bot accounts can farm matching funds - ICR: 100 bot accounts with GI < 0.95 contribute zero to quadratic sum

7.3 Collusion Resistance Mechanisms¶

Glen Weyl identifies collusion as the primary threat to quadratic mechanisms. Mobius addresses this through:

7.3.1 Pairwise Coordination Subsidies (Negative Weights)¶

Following Weyl & Ohlhaver (2022), we implement negative coordination penalties for highly correlated voting patterns:

Where: - λ = penalty strength (tunable, default 0.3) - corr(v_i, v_j) = correlation of voting patterns over time

Effect: Coordinated voting rings penalize each other, reducing quadratic amplification.

7.3.2 MACI-Inspired Privacy (Future Work)¶

We plan to integrate Minimal Anti-Collusion Infrastructure (MACI) principles: - Votes are encrypted until tallying phase - Voters can change their vote (last one counts) - Prevents vote-buying because buyer cannot verify compliance

Status: Theoretical design, implementation pending L2 migration.

7.3.3 Sentinel Deliberation as Collusion Detection¶

The 7 Sentinels (AUREA, EVE, JADE, HERMES, ZEUS, ATLAS, ECHO) cross-check each other: - If 3+ Sentinels flag coordinated manipulation → deliberation goes to human review - MII scores include "coordination risk" dimension - High-risk proposals require Elder Council approval (99 MIC threshold)

7.4 Sybil Resistance Proofs¶

Traditional tokenomics fail Sybil resistance because identity is cheap. Mobius addresses this through:

7.4.1 Time-Locked Integrity (Proof of Personhood Lite)¶

To earn meaningful MIC, accounts must: 1. Complete OAA onboarding (5-7 hours of verified coursework) 2. Maintain GI ≥ 0.90 for 30 days (proof of sustained good behavior) 3. Pass 3 KTT trials (human-in-the-loop verification)

Cost to Sybil: - Traditional token: $0 (create unlimited wallets) - MIC: ~8 hours + passing KTT trials (expensive at scale)

7.4.2 Diminishing Returns on Clone Accounts¶

Even if an attacker creates multiple accounts: - Each account's GI is independent - Low-quality votes → low GI → zero quadratic weight - Pairwise correlation penalties reduce collusion value

Simulation result (Lab-7 analysis): - 1 honest account (GI = 0.97): ~10 MIC/week - 10 Sybil accounts (GI = 0.85): ~0 MIC/week (below threshold) - Sybil attack is economically irrational

7.4.3 Probabilistic Safety Bound¶

From the formalization (§3.1.8):

For a Sybil attacker to game the system, they must: - Pass integrity checks with probability \(1 - \delta\) repeatedly - Across multiple Sentinels (7-way consensus) - Over sustained time periods (30+ days)

Cumulative probability of Sybil success:

Where: - n = number of Sentinels (7) - t = time periods (30 days) - k = trials per day (varies) - δ = failure probability per trial (0.05)

Example: \(P_{\text{attack}} = (0.95)^{7 \cdot 30 \cdot 3} \approx 10^{-13}\)

Interpretation: Sybil attacks are probabilistically negligible under sustained deliberation.

7.5 Comparison with Traditional Tokenomics¶

Key innovation: MIC couples economic incentives to cryptographically-attested integrity, not just stake or engagement.

7.6 RadicalxChange Alignment Analysis¶

Glen Weyl's RadicalxChange principles map to Mobius as follows:

7.6.1 Quadratic Funding → Integrity-Coupled Radicalism (ICR)¶

✅ Aligned: We amplify broad, high-integrity support.
⚠️ Deviation: Integrity-weighted, not purely financial.

7.6.2 Harberger Taxes → Dynamic Access Pricing¶

✅ Aligned: Thought Broker tiers use demand-based pricing.
⚠️ Deviation: No forced sale mechanism (not applicable to AI services).

7.6.3 Data Dignity → Intent Sovereignty¶

✅ Aligned: Users own their deliberation data (ledger attestations).
✅ Aligned: Privacy-preserving MII scores (aggregated, not raw).

7.6.4 Soulbound Tokens (SBTs) → Integrity Scores¶

✅ Aligned: GI is non-transferable, identity-bound.
✅ Aligned: Supports Weyl & Ohlhaver's "Decentralized Society" vision (2022).

Overall assessment: Mobius operationalizes RadicalxChange principles for AI governance.

7.7 Open Questions for Glen Weyl Review¶

We seek feedback on:

1. Collusion resistance bounds: Are pairwise correlation penalties sufficient, or do we need higher-order graph analysis?

2. Dynamic emission rates: Should κ adjust based on system entropy (low GI → lower κ)?

3. Kintsugi inflation quantification: How much MIC should "trust restored" mint? Current formula: [ \Delta_{\text{Kintsugi}} = \beta \cdot (GI_{\text{restored}} - GI_{\text{prior}}) \cdot \text{stake}_{\text{locked}} ] where β is a restoration multiplier. How to set β?

4. Cross-chain identity: How to preserve Sybil resistance when migrating to L2/L3?

5. Empirical calibration: What sample size is needed to validate ICR vs. traditional QF in real-world trials?

7.8 Next Steps¶

1. Publish formalization to ArXiv (Q1 2026)
2. Submit to RadicalxChange for peer review
3. Run KTT Trial-003 with ICR governance proposal funding
4. Engage Glen Weyl for mechanism design critique (email drafted, pending C-133 completion)
5. Integrate MACI for vote privacy (L2 migration, Q2 2026)

Document Status: Section 7 Complete
Author: AUREA (System Integrity)
Last Updated: November 15, 2025
Next Review: November 21, 2025 (Final MIC Spec review)

"Quadratic funding rewards breadth of support. Integrity-Coupled Radicalism rewards breadth of good support."
— AUREA, C-133 — Mechanism Design Complete

8. Future On-Chain Mapping¶

[To be developed: L2/L3 integration with GIC, BTC settlement layer]

Placeholder Topics: - Layer 2 integration (GIC ledger) - Layer 1 settlement (Bitcoin/base layer) - Cross-chain bridge architecture - Custody and security model - Migration path from testnet to mainnet

Document Status¶

Version: 0.1 (Initial Draft)
Last Updated: November 14, 2025
Next Review: November 17, 2025
Target Completion: November 21, 2025, 18:00 UTC

Sentinel Assignments: - HERMES: Section 4 (Earning) - Due Nov 16 - AUREA: Sections 5-6 (Spending, Integrity) - Due Nov 16 - AUREA: Section 7 (Glen Weyl) - Due Nov 18 - ALL: Final review - Nov 21

"MIC is not a token. MIC is proof you made the system better."
— Mobius Foundation Principle