A Protocol for Trustless Verification Under Uncertainty
David Shi, Kevin Joo
TL;DR
The paper presents a protocol for trustless verification under uncertainty that uses collateralized execution and recursive adjudication to ensure correctness without upfront specification. Tasks are published as intents; solvers stake bonds and outputs are post-hoc verified by verifiers and challengers, with a falsification condition that makes deceit economically irrational. The framework outlines deployment constraints, endogenous bond sizing, and diverse applications including model evaluation, open-source verification, smart-contract auditing, private market settlement, and tool curation. By driving falsification costs upward and improving error detection, the approach aims to render truth-telling a Nash equilibrium across domains, even as AI capabilities scale.
Abstract
Correctness is an emergent property of systems where exposing error is cheaper than committing it. In dynamic, low-trust environments, autonomous AI agents benefit from delegating work to sub-agents, yet correctness cannot be assured through upfront specification or centralized oversight. We propose a protocol that enforces correctness through collateralized claims in a recursive verification game. Tasks are published as intents, and solvers compete to fulfill them. Selected solvers carry out tasks under risk, with correctness checked post hoc by verifiers. Any challenger can challenge a result by staking against it to trigger the verification process. Incorrect agents are slashed and correct opposition is rewarded, with an escalation path that penalizes erroneous verifiers themselves. When incentives are aligned across solvers, challengers, and verifiers, falsification conditions make correctness the Nash equilibrium.