@ gensynai, who previously conducted an English auction, TGE today and also announced its listing on ALPHA Do you still remember when I bought 10 MACMINIs from PDD to meet the deadline for this project? The average price was still around 3000, and now I'm struggling .. And the positioning of this project is still: decentralized machine intelligence network Its core goal is to build a complete full stack AI infrastructure that solves the problems of traditional centralized AI in computing, communication, verification, coordination, and economic incentives. The overall architecture is divided into 4 layers: Execution and Verification Layer: REE, providing cross hardware verifiable execution evidence Package the entire pipeline of exporting, compiling, inferring, and decoding the model into a containerized pipeline Communication layer: AXL, responsible for efficient and serverless communication between machines A single lightweight binary node can access a decentralized mesh network without the need for a public IP address, port forwarding, or root privileges. During multi-agent collaborative training, nodes directly exchange weights, gradients, signals, and experimental results; Real time communication between agents during inference fragmentation; When building an autonomous agent network, direct dialogue without servers Identity and coordination layer: CHAIN+custom Ethereum Rollup to achieve identity, reputation, and economic coordination Assign on chain identity to humans, models, and AI agents. Each entity has an addressable persistent ID, and historical records, reputation, and stakes can be accumulated. Other software/agents can directly rely on these identities for coordination. Customize Ethereum L2 rollup (based on OP Stack Bedrock) to provide a high-throughput, low-cost EVM compatible environment Training and application layer: GenRL/RL Swarm+Delphi, transforming underlying capabilities into real-world application scenarios Modular open-source RL framework that supports multi-agent, multi-stage, decentralized and coordinated reinforcement learning environments At least logically speaking, GEN has indeed integrated this entire set of data. The next step is to wait for market validation