This paper critically examines both the promises and pitfalls associated with Web3 decentralization within the evolving landscape of global governance, increasingly influenced by Generative AI (GenAI). It questions whether Web3 technologies meaningfully redistribute power or primarily strengthen a new cyber-libertarian elite composed of technologically skilled individuals with significant digital capital. Drawing from extensive fieldwork action research conducted in Silicon Valley since August 2022, as well as direct interactions with scholars and practitioners—including detailed dialogues held in Washington, DC, in April 2025—the analysis explores the underlying techno-political beliefs driving the global Web3 movement. Advocates commonly depict Web3 as an effective means to break centralized data monopolies and challenge traditional state institutions. However, this paper argues that decentralization does not inherently lead to democratization. Instead, Web3 potentially reproduces or even deepens existing hierarchies rooted in algorithmic control and digital privilege, especially when integrated with GenAI systems. To illustrate this point, the paper delineates three emerging paradigms shaped by Web3 logic: (i) Network States, promoting crypto-libertarian models of governance; (ii) Network Sovereignties, enabling collective self-determination for marginalized groups; and (iii) Algorithmic Nations, constructing decentralized, data-driven infrastructures aimed at global collaboration. While innovative and provocative, these paradigms simultaneously raise significant concerns about inclusivity, equity, and the concentration of technocratic influence. The paper concludes by proposing hybrid policy frameworks that acknowledge global power imbalances and local contextual realities, highlighting the urgency of pursuing digital justice and fostering international cooperation within a GenAI-powered Web3 environment.