Cryptographic Reciprocity Theory (CRT) is a framework for constructing systems in which cooperation emerges from rational self-interest over extended time horizons.
This document describes the theory without philosophical interpretation.
CRT does not assume altruism or moral alignment.
It assumes that agents respond to:
- Incentives (what benefits them)
- Observability (what others can see)
- Delayed consequences (what happens later)
The theory is based on three premises:
Single encounters favor defection (betrayal).
Repeated encounters favor cooperation (tit-for-tat).
When agents know they will interact again, cooperation becomes more valuable than immediate gain.
You don't need to trust someone before cooperating with them.
Trust emerges from repeated, predictable behavior over time.
CRT systems make behavior observable so trust can develop naturally.
Promises fade.
Contracts can be breached.
Institutions can be captured.
Cryptographic commitments persist regardless of: - Who created them - What authority enforces them - Whether the creator is still alive
CRT builds on the Iterated Prisoner's Dilemma.
In a single-round Prisoner's Dilemma: - Defection (betrayal) is always optimal - Cooperation is irrational
In repeated games: - Defection in round N hurts you in round N+1 - Cooperation becomes rational over time - Tit-for-tat strategies emerge as optimal
CRT extends this by: - Making behavior observable (blockchain records) - Making history persistent (immutable ledger) - Extending time horizons (21-year cycles)
CRT uses these Bitcoin primitives:
Participant generates a Bitcoin key pair.
Public key becomes their identity.
Private key remains secret.
No registration required.
No email or phone number.
Just cryptography.
Participant creates a time-locked transaction: - Locks funds until block height X - Signs transaction with private key - Broadcasts to Bitcoin network
The commitment is now: - Publicly verifiable - Immutable - Enforceable by consensus rules
Other participants can: - See the commitment on blockchain - Verify the signature - Track the participant's history - Update their assessment of reliability
This is reputation without a centralized database.
If participant defects: - Their history shows the defection - Future partners see this - Cost of cooperation with them increases
If participant cooperates: - Their history shows reliability - Future partners prefer them - Access to valuable relationships improves
Time-locks solve the commitment problem.
Without time-locks: - Promises can be broken - Commitments are cheap - Signals are unreliable
With time-locks: - Breaking commitment means losing funds - Making commitment is costly - Signals are expensive to fake
This creates credible commitment.
CRT reputation is not a score.
It is an observable history.
Traditional reputation systems: - Collapse identity to a number (5 stars, 1000 points) - Hide context and nuance - Can be gamed or manipulated
CRT reputation: - Preserves full history of actions - Maintains context (when, with whom, under what conditions) - Cannot be erased or edited
Participants form their own judgments based on observable facts.
CRT aligns incentives through:
Short-term defection creates long-term costs.
Reputation damage compounds over time.
Actions cannot be hidden or denied.
Blockchain provides permanent record.
No one has privileged access.
Everyone faces the same incentives.
Time-locked commitments are expensive to fake.
Only serious participants will make them.
Let: - A = set of agents with persistent identities - L = public ledger of all actions - T = time horizon (e.g., 21 years)
For each agent i at time t: - Hi,t = history of actions up to time t - Ri,t = reputation derived from Hi,t
Agent i chooses strategy si to maximize:
V_i = Σ (payoff_t × discount_factor^t)
Where: - payofft depends on actions and reputation - discount_factor < 1 (future matters less than present)
In CRT systems: - Defection at time t reduces Ri,t+1 - Lower reputation reduces future payoffs - Long time horizon T makes future losses significant
Thus, cooperation becomes rational.
| Traditional | CRT |
|---|---|
| Trust required upfront | Trust emerges from observation |
| Reputation controlled by platform | Reputation derived from blockchain |
| Commitments enforced by law | Commitments enforced by cryptography |
| Exit may be restricted | Exit always possible |
| Rules can be changed arbitrarily | Rules are transparent and symmetric |
CRT cannot: - Force cooperation (agents can always defect) - Eliminate all uncertainty (unknown unknowns remain) - Work with fully anonymous participants (identity must be persistent) - Prevent Sybil attacks without cost-of-entry (proof-of-funds or proof-of-work) - Encode subjective judgments (only objective facts)
CRT is suited for: - Long-term business relationships - Inheritance and intergenerational transfer - Decentralized marketplaces - Reputation-based access control - Coordination without institutions
CRT is NOT suited for: - Anonymous transactions (requires persistent identity) - Instant finality (Bitcoin confirmations take time) - Off-chain enforcement (only on-chain actions are observable) - Complex subjective disputes (requires human judgment)
See CRT in practice: Architecture
Understand time-locks: Time-Locked Covenants
Learn about reputation: Reputation & Incentives
CRT is a framework, not a solution.
It solves some problems by creating new constraints.