A few days ago, I took a brief look at the simplified white paper V1 released by Arc, and the main points were as follows: 1. Arc has focused on the integration of basic digital services and seamless payment mechanisms for AI agents. From this perspective, arc has captured the key point for the entry of encrypted AI agents. It seems that the team not only considered the technical aspects, but also took into account the practical implementation framework. At present, most AI agents only provide advisory services and cannot transition to autonomous execution. For example, they can recommend flights and hotels based on user needs, but most cannot automatically help users complete reservations; For example, it can help users analyze DeFi trading strategies, but it cannot directly execute trades or adjust investment portfolios in real time. The main reason is the lack of interoperability. Currently, most integrations are customized, and once users switch platforms, their context will mostly disappear, resulting in the inability to complete execution coherently. In addition, there is a lack of a universal automatic payment system. This involves the concept MCP mentioned below for arc V1. 2. Utilize Anthropic's MCP (Model Context Protocol) to build standardized interfaces for AI agents, enabling access to web2 and web3 services. To enable AI agents to seamlessly connect to various services. Similar to HTTP for AI. It is also the core of the Ryzome AI agent app store for Arc V1. 3. The Ryzome Agenetic system attempts to integrate multiple AI agents and various digital services. As mentioned above, MCP is the HTTP of the AI era, which supports any browser to communicate with various websites. MCP supports communication between AI agents and various digital services, and is a universal connection protocol. Ryzome can communicate with web2 and web3 services through MCP. In addition, Ryzome also has an identity system. The identity system will be associated with different application scenarios, whether it is users booking flights and hotels, conducting DeFi transactions, or adjusting investment portfolios. It will understand user preferences and does not need to generate new revenue to support comprehensive collaboration. Additionally, Ryzome AI agents support the mutual calling of functions (with a support fee), which also enables the Lego combination between AI agents. This composability will generate countless new application scenarios. If it really lands, it will be very interesting. 4. The seamless payment mechanism integration mentioned above is to support the implementation of automatic payments. Ryzome payment uses Arc token. Arc token serves as a gas fuel token and settlement token, providing gas power for seamless transactions between AI agents and digital services; Arc serves as a settlement token, with 85% of the revenue generated going to service providers, 10% going to Arc for development, and 5% covering operating costs. For example, when using Ryzome to call flight and hotel booking services or DeFi protocols, 85% of the arc fee will be paid to the service provider and 10% will flow into the arcdao treasury; 5% is used to pay for operating costs. This payment allocation structure aims to achieve long-term economic sustainability. If AI agents succeed in the economic system, traditional applications will inevitably be marginalized or even fully anticipated. From the above concepts, Arc attempts to solve the problem of service fragmentation by combining MCP and seamless payment mechanism (Arc payment) through Ryzome architecture, in order to achieve the transformation of AI agents from providing suggestions to independent execution: any AI agent approved by the user can achieve end-to-end task execution. If Arc v1 is implemented, the most useful parts for AI agent developers would be the MCP and built-in payment integration. Because they can focus on AI agent application innovation without building custom bridges and payment channels connecting multiple services. For users (individual or enterprise users), AI agents are no longer just chatbots that can provide suggestions, but can directly help users achieve the desired results. 7. AI agents are one of the few fields that can achieve breakthroughs in the field of encryption. At present, it is not a matter of one or two days, and it requires some longer time and patience to brew.