Is Decentralized Artificial Intelligence Governable?

Abstract: 

The rise of decentralized artificial intelligence (DeAI) represents a transformative shift in AI development, leveraging blockchain and distributed ledger technologies to decentralize data, computation, coordination, and economic models. Advocates of DeAI highlight its potential to address many of the limitations of centralized AI systems, including single points of failure, trust deficits, and the monopolization of value. However, DeAI also introduces profound challenges, particularly in governance. Unlike centralized systems, DeAI operates across global, borderless networks, making it resistant to traditional regulatory mechanisms. This paper investigates two central questions: (1) How do decentralized technologies enable self-sovereignty and potential unstoppability of decentralized AI?  (2) What are the regulatory and governance challenges posed by decentralized AI and how can they be addressed? Through an analysis of DeAI’s key technical enablers, we explore how these systems achieve operational autonomy and resilience akin to virus spreading over biological networks, a characteristic we conceptualize as “unstoppable in the wild.” We identify the inherent characteristics of DeAI that challenge traditional oversight, including its global reach, immutability, and adaptive survival strategies, rendering the technology “ungovernable” in the conventional sense. We contribute to raising awareness of DeAI governance research in this rapidly emerging field, emphasizing its potentially striking impact. 

Bio: 

Botao ’Amber’ Hu is a social computing researcher and experiential futures designer. 

He directs Reality Design Lab, an independent interdisciplinary research and design lab exploring the intersection of soma design, speculative design, spatial computing, and social computing. As a researcher, his interest lies in the governance of Decentralized AI (DeAI), protocol science, and complexity science. 

As a designer, his primary focus is creating socially experiential futures using mixed reality as the main medium.  He also serves as a visiting scholar at University of Nottingham, and visiting lecturer at the China Academy of Art. 

His work has been featured at top conferences such as SIGGRAPH, CSCW, CHI, UbiComp, WWW, TEI, ISEA, IEEEVR, IEEEVIS, ISMAR, Ars Electronica, SXSW, and TEDx and has received accolades including the SIGGRAPH Best in Show, CHI Best Interactivity, Webby, Red Dot, iF Design, Good Design, A’ Design, Core77 Design Award, and grants from the Ethereum Foundation. 

He holds a bachelor’s degree in computer science from Tsinghua University and a master’s degree in computer science with AI concentration from Stanford University.

Links

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5110089 

https://reality.design