Applications and Future Explorations of AI Agents in the Web3 Field

Recently, a global first universal AI Agent product called Manus has attracted widespread attention in the tech circle. This product, developed by a Chinese startup, possesses the capability to autonomously complete tasks from planning to execution, demonstrating unprecedented versatility and execution power. The meteoric rise of Manus has not only drawn industry attention but has also provided valuable product ideas and design inspiration for various AI Agent developments.

With the rapid development of AI technology, AI Agents, as an important branch of artificial intelligence, are gradually moving from theory to practice, demonstrating enormous application potential in various industries, and the Web3 industry is no exception.

Basic Concept of AI Agent

An AI Agent is a computer program that can make autonomous decisions and execute tasks based on the environment, input, and predetermined goals. Its core components include:

1. The large language model ( LLM ) as the "brain"
2. Observation and Perception Mechanism
3. Reasoning Process
4. Action Execution Ability
5. Memory and Retrieval Function

The design patterns of AI Agents mainly have two development paths: one focuses on planning capabilities, such as REWOO and Plan & Execute; the other emphasizes reflective capabilities, such as Basic Reflection and Reflexion.

The most widely used application currently is the ReAct model, whose typical process is: Think ( Thought ) → Act ( Action ) → Observe ( Observation ), referred to as the TAO loop.

According to the number of agents, AI Agents can be divided into two categories: Single Agent and Multi Agent. Single Agent focuses on the combination of LLM and tools, while Multi Agent assigns different roles to different agents to collaboratively complete complex tasks.

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 over 90%. Currently, the main projects that are relatively active revolve around exploring Web3 based on the AI Agent framework, with three primary models:

1. Launch Platform Mode: represented by Virtuals Protocol, allows users to create, deploy, and monetize AI Agents.

2. DAO Model: Represented by ElizaOS, it uses AI models to simulate investment decisions and combines suggestions from DAO members for investment.

3. Business Company Model: Represented by Swarms, providing an enterprise-level Multi-Agent framework.

From the perspective of economic models, currently only the launch platform model can achieve a self-sustaining economic closed loop. However, this model also faces challenges, as the assets to be issued must possess sufficient appeal to create a positive cycle.

Web3 Exploration of MCP Protocol

Model Context Protocol (MCP) is an open-source protocol launched by Anthropic, aimed at addressing the connectivity 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 achieve decentralization and anti-censorship.

2. Empower the MCP Server with the ability to interact with the blockchain, such as conducting DeFi transactions and management.

3. Build an Ethereum-based OpenMCP.Network creator incentive network.

Although these directions can theoretically inject decentralized trust mechanisms and economic incentives into AI agents, they still face challenges in technical implementation and efficiency.

Conclusion

The launch of Manus marks an important milestone for general AI Agent products. The Web3 world also needs a milestone product to break through external doubts. The emergence of MCP brings new exploration directions for AI Agents in Web3. The fusion of AI and Web3 is an inevitable trend, and we need to maintain patience and confidence while continuously exploring the infinite possibilities in this field.