Skip to content

Novelty & Feasibility Assessment

What's Genuinely Novel

  1. Domain Transfer: Network Backpressure → Monetary Flow Control - Nobody has applied Tassiulas-Ephremides backpressure routing to monetary flows in token economies. Backpressure routing (1992+), congestion pricing (Kelly 1998), and streaming payments (Superfluid 2020+) exist independently. No work unifies all three.

  2. Demurrage-Backpressure Distinction - Demurrage: stock-based, time-decay, drives uniform velocity. Backpressure: flow-based, capacity-constraint, drives allocation efficiency. Orthogonal axes, complementary mechanisms.

  3. AI Agent Pipeline with Upstream Propagation - Multi-agent pipelines naturally exhibit backpressure dynamics (slow downstream → upstream throttles). No existing agent payment protocol addresses flow control.

  4. Timing - Agent payment ecosystem (March 2026): Google AP2, Coinbase x402, OpenAI-Stripe ACP, Visa TAP all focus on authorization/trust. None address flow control.

What's NOT Novel (Acknowledge)

  • Backpressure routing with throughput optimality (Tassiulas 1992)
  • Network congestion pricing with proportional fairness (Kelly 1998)
  • Streaming payments (Superfluid, Sablier 2018+)
  • Token economy modeling via dynamical systems (Zhang-Zargham 2020)

Prior Art Map

Stream 1: Network Flow         Stream 2: Network Economics
  Tassiulas (1992)               Kelly (1998)
  Neely (2010)                   Srikant (2004)
         ↘                        ↙
    ┌─────────────────────────────────────┐
    │   BACKPRESSURE ECONOMICS (gap)      │
    │   Capacity-based monetary routing   │
    └─────────────────────────────────────┘
         ↗                        ↖
Stream 3: Token Engineering    Stream 4: Monetary Velocity
  Zhang-Zargham (2020)           Gesell (1916)
  Cha et al. (2025)              Fisher (1933)
                  ↘              ↙
           Stream 5: AI Agent Economies
             Google AP2, x402, Stripe ACP

Feasibility

Domain Feasibility Capacity Signal Verification
AI Agent Pipelines Strong Queue depth, latency, throughput Task completion proofs
Decentralized Compute Strong GPU util, VRAM, queue depth TEE attestation
Public Goods Funding Partial Team size, PRs, deployments Soft signals
Subjective Goods Not feasible No verifiable signal N/A