AI Agent's Cross-Industry Exploration in the Web3 Field: From Manus to MCP

Recently, a product called Manus, the world's first universal AI Agent, has attracted widespread attention in the industry. Developed by the domestic startup Monica, this product saw a situation on its first day online where invitation codes were in high demand. As a universal AI Agent, Manus demonstrates strong abilities in independent thinking, planning, and executing complex tasks, capable of autonomously completing the entire process from planning to execution, such as writing reports and creating spreadsheets.

The explosive popularity of Manus has not only attracted attention within the industry but has also provided valuable product ideas and design inspiration for the development of various AI Agents. With the rapid advancement of AI technology, AI Agents, as an important branch of artificial intelligence, are gradually transitioning from concept to reality, demonstrating immense application potential across various industries, and the Web3 sector is no exception.

Core Components and Design Patterns of AI Agent

An AI Agent is a computer program that can autonomously make decisions and execute tasks based on its environment, inputs, and predefined goals. Its core components include:

• Large Language Model (LLM) as its "brain"
• Observation and Perception Mechanism
• Inferential thinking process
• Action Execution
• Memory and Retrieval

The design patterns of AI Agents mainly have two development paths:

1. Emphasis on planning ability: including REWOO, Plan & Execute, LLM Compiler
2. Emphasis on reflective ability: including Basic Reflection, Reflexion, Self Discover, LATS

Among them, the ReAct mode is currently the most widely used AI Agent design pattern. ReAct addresses diverse language reasoning and decision-making tasks by combining reasoning and acting within language models. Its typical process can be described as a cycle of "Thought → Action → Observation."

The Current Status of AI Agents in Web3

The popularity of AI Agents in the Web3 industry peaked in January this year and then significantly declined, with the overall market value shrinking by more than 90%. Currently, the major projects with substantial attention and market value mainly explore Web3 around the AI Agent framework, with three main models:

1. Launch Platform Mode: represented by Virtuals Protocol
2. DAO Model: represented by ElizaOS
3. Business company model: represented by Swarms

From the perspective of economic models, currently only the launch platform model can achieve a self-sufficient economic loop. However, this model also faces challenges, primarily that the issued AI Agent assets need to have enough "attraction" to form a positive flywheel. At present, most of the launched AI Agents are essentially memes, lacking intrinsic value support.

Exploration of MCP in Web3

Model Context Protocol (MCP) is an open-source protocol launched by Anthropic, aimed at addressing the connection and interaction issues between LLMs and external data sources. The emergence of MCP brings new exploration directions for AI Agents in Web3:

1. Deploy the MCP Server to the blockchain network to solve the single point issue and possess anti-censorship capabilities.
2. Empower the MCP Server with the ability to interact with the blockchain, reducing the technical barrier.

In addition, there is a solution based on Ethereum to build the OpenMCP.Network creator incentive network. This network aims to achieve automation, transparency, trustworthiness, and censorship resistance of incentives through smart contracts, utilizing technologies such as Ethereum wallets and ZK for signature, permission verification, and privacy protection during the operation process.

Summary and Outlook

The release of Manus marks an important milestone for the general AI Agent product. The Web3 world also needs a milestone product to break the skepticism that Web3 lacks practicality and is only hype. The emergence of MCP brings new exploration directions for AI Agents in Web3, including deployment to blockchain networks, interaction with blockchains, and building creator incentive networks.

Although theoretically, the combination of MCP and Web3 can inject decentralized trust mechanisms and economic incentive layers into AI Agent applications, the current zero-knowledge proof technology still struggles to verify the authenticity of Agent behaviors, and decentralized networks also face efficiency issues. This is not a solution that can succeed in the short term.

AI, as one of the grandest narratives in history, is inevitably merging with Web3. We need to maintain patience and confidence as we continue to explore the infinite possibilities in this field.