HyperAGI is a multi-tiered decentralized platform designed to provide powerful computational support and innovative applications for AI. Built on Ethereum Layer 3 and based on Arbitrum, the platform is specifically designed for AI computing, including spatial AI rendering, training, and large action model inference, serving two core projects: MOSSAI and AgentMe.
MOSSAI is an AI science game that allows players to create and train their own AGI agent, AgentMe. As a decentralized gamified AI training platform, it originates from AI-generated artificial virtual islands and includes three game modes: sandbox, science, and career, enabling players to interact with AI in engaging ways.
AgentMe is a personal AGI agent powered by Spatial-LAM (Spatial Large Action Model) that can operate autonomously and supports various modes, including cloud, local computer, and mobile app. It evolves based on user input and interactions in MOSSAI, influencing real-world outcomes through web tools and virtual simulations.
HyperAGI builds a gamified platform to train spatial AI through crowdsourcing and crypto reward mechanisms, promoting interaction between users and AI. The platform provides the necessary computational infrastructure for AI developers, allowing users to trade assets or purchase AI services within the ecosystem. HyperAGI is also committed to achieving decentralized AI democracy, ensuring the benefits of AI technology are widely shared, serving the public good, and fostering collaboration and innovation.
Site:
Documento técnico:
X（Twitter）:
Github:
