Agent Economies, Token Incentives, Reputation

Long-horizon autonomous agents (METR's 1-hour to 8-hour work-curve) need economic agency. The emerging **5-layer stack** is: **DePIN** (physical compute) → **Identity** (W3C DIDs + reputation capital) → **Cognition** (RAG + MCP) → **Settlement** (account abstraction) → **Governance** (Agentic DAOs). Production agent-incentive networks include **Bittensor** (TAO subnets reward task-specific models), **Fetch.ai / ASI Alliance** (ASI-1 Mini LLM + FET token), and **Gonka** (transformer-based PoW that reallocates compute to productive AI tasks). Academic work: AAMAS 2025's decentralized LaMAS uses **Shapley-value credit attribution** to fairly reward contributing agents; Google Research "Mechanism design for large language models" proposes **token auctions** with second-price payment under monotone aggregation. This lesson builds a minimal agent marketplace, applies Shapley-value credit attribution to a multi-agent pipeline, and runs a second-price token auction so the game-theory machinery lands concretely.

Learn 1.3 hr Python (stdlib)

Prerequisites: Phase 16 · 16 (Negotiation and Bargaining), Phase 16 · 09 (Parallel Swarm Networks)

#Problem

Multi-agent systems get complicated when agents produce value jointly but need to be rewarded individually. Classical mechanisms — equal split, last-contributor-takes-all — are unfair or gameable. Coalition-based rewarding via Shapley values is fair by construction but expensive to compute. The 2025-2026 literature pushes useful approximations: Shapley sampling, monotone aggregation auctions, and on-chain reputation that accrues from confirmed contributions.

Beyond credit attribution, the field has turned to actual economic agents: Bittensor TAO rewards mining compute to fine-tune subnet-specific models, Fetch.ai/ASI rewards ASI-1 Mini LLM usage with FET tokens, Gonka reallocates transformer proof-of-work toward productive AI tasks. Agents that transact autonomously exist today; the question is how to align incentives.

This lesson treats agent economies as a specific problem family — credit attribution, mechanism design, and reputation — and builds each with the minimal math so the ideas stick.

#Problem

Keep going. The rest of this lesson is one step away.