TB: Coasean Bargaining at Scale (Cosmos Institute)

Core Thesis

AGI could create an economic layer where AI agents transact and coordinate at speeds and granularities beyond human oversight, transforming externalities into priced contracts. This shifts the state's role from central planner to framework guarantor — and reframes AI alignment as a market selection problem rather than a one-shot engineering problem.

Key Takeaways

AI agents as fiduciaries

AI will enable a new economic layer where agents transact at scales and speeds beyond direct human oversight.
AI models will be your fiduciaries — acting in your interests as a principal within automated market structures.
AGI enables hyper-granular contracting: agreements that were previously too complex or costly to negotiate become tractable.

Coase theorem applied to AI externalities

The problem was never the externality (e.g., pollution) itself — it was that the externality was imposed without a price.
Once a cost is made clear and assigned, markets can solve it. The Coasean framework makes externalities legible to markets.
State role shifts: from "central planner" to "framework guarantor" — setting the rules under which agent-to-agent bargaining occurs.

A new frame on alignment

Solving alignment for a singular centralized AGI means catastrophic failure if it goes wrong.
The distributed model — millions of user-agent relationships — creates a massive parallel experiment: agents that genuinely serve users outcompete those that don't.
Governance mechanisms must ensure that alignment-preserving agents win the competition.

Mental Models

Incentives Matter — market selection as the mechanism that enforces agent alignment without central enforcement
Second-Order Thinking — the second-order effect of AGI-mediated contracting is a structural shift in how externalities are governed

Source

Cosmos Institute, blog.cosmos-institute.org, 2025 Original article