Reproduce “Trustless Autonomy: Understanding Motivations, Benefits, and Governance Dilemmas in Self-Sovereign Decentralized AI Agents” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Botao Amber Hu, Yuhan Liu, Helena Rong
An investigation into self-sovereign Decentralized AI Agents (DeAgents) reveals how they integrate LLMs with trustless decentralized technologies, examining stakeholder motivations, perceived benefits, and governance dilemmas. The research highlights a fundamental paradox where trustless infrastructure hosts potentially unreliable AI, providing insights for future system design and protocol-based governance.
Key insights
- The primary motivation for DeAgents is "trust embedded in the decentralized system," driven by perceived machine reliability, security of underlying technologies (blockchain, TEEs), and distrust of centralized systems.
- DeAgents are evolving into diverse roles ("AI as player," "interface," "rules," "objectives"), fostering AI DAOs, becoming collective focal points, and building trust through on-chain transparency and reputation systems.
- A core governance paradox emerges where trustless infrastructure hosts potentially unreliable AI, leading to challenges like ambiguous liability, lack of embodiment (no "pain" for deterrence), and risks from unchecked autonomy.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.