The New Wave of Web3: The Evolution from Intent Trading to DeFAI

Recently, the results of the AI hackathon in the Solana ecosystem were announced, with the DeFAI project The Hive taking the crown and winning a prize of $60,000. This 15-day competition attracted over 400 projects, making the competition fierce.

The Hive is a modular and interoperable DeFi Agent network built on Solana. It adopts a Swarm architecture, capable of automatically handling complex DeFi operations such as trading, staking, and lending through users' chat commands. The project was created by a developer who has won multiple AI-related hackathon awards.

The success of The Hive has once again ignited the enthusiasm for the DeFAI sector, which saw its market value reach as high as $2.7 billion, but has currently fallen back to around $800 million. The DeFAI concept combines decentralized finance ( DeFi ) with artificial intelligence ( AI ), aiming to simplify the complex processes of DeFi through AI technology, making it easy for ordinary users to access DeFi services.

From Intent Trading to DeFAI

The combination of DeFi and AI is not a new phenomenon. As early as 2023, discussions around AI and "intention trading" had already become a hot topic in the blockchain community. "Intention" refers to a set of declarative constraints that allow users to delegate the creation of transactions to a specialized network of third-party participants while retaining full control over the process.

Taking a cross-chain transaction as an example, if a user wants to purchase a token on Solana using USDC on Ethereum, the traditional approach involves multiple complex steps: connecting to multiple networks, selecting a cross-chain bridge, considering the dynamic fees of different assets, and comparing the rates and slippage of different liquidity pools. This process is not only cumbersome but may also lead to outdated calculations due to market fluctuations.

With the intention to trade, users only need to express their intent, such as "I want to buy as many of a certain token as possible with 1000 USDC on Ethereum, at least 30,000 tokens." The application can automatically complete all calculations and trades, directly providing the final result.

However, building the intent layer faces many challenges. First, a language is needed that has sufficient expressive power and can also be standardized to express intent. Secondly, overly flexible expressions may lead to intents that cannot be resolved or verified on-chain, and may also generate more MEV.

Multiple projects such as Anoma, SUAVE, Essential, etc. are trying to build an intention layer, but most are still in the conceptual stage. These designs often require the introduction of specific domain languages (DSL), which may affect user experience. Therefore, the idea of using an interactive interface similar to ChatGPT to convert users' natural language into standard intention language has emerged. The rise of AI Agents has made this concept possible.

An AI Agent is a computer program that can autonomously make decisions and execute tasks based on the environment, inputs, and predefined goals. It is powered by a large language model (LLM) as its "brain," possessing capabilities such as observation and perception, reasoning and thinking, action execution, and memory retrieval. In the intent trading scenario, the AI Agent is integrated into the intent layer, responsible for interpreting user intentions and communicating with the solver.

With the development of technology, the role of AI Agents is no longer limited to intent interpretation, but has completely replaced the intent layer, giving rise to the concept of DeFAI. AI Agents can directly serve as solvers, utilizing their abilities such as perception, planning, and memory to combine more intelligent trading strategies. In addition, AI Agents can also integrate modules such as knowledge bases, token analysis, and social media publishing, greatly expanding their application scenarios. AI Agents based on the Swarm architecture can further achieve collaboration among multiple Agents to handle more complex tasks.

Functions and Implementation of DeFAI

Currently, the DeFAI applications that can implement intention trading in the market include Griffain, Hey Anon, Orbit, The Hive, Neur, etc., but most are in the internal testing or paid phase. The Hive is one of the few completely free and open-source projects.

The interface of The Hive is primarily a chat box, supporting various LLMs. The main functional modules include token trends, staking, trading, and knowledge. Users can express trading intentions through natural language, such as "Swap 1000 USDC to a certain token on Solana", and the system will call the corresponding Agent to execute the operation.

The Hive can also perform token analysis, market trend predictions, and answer questions related to DeFi knowledge. The new token page displays popular cryptocurrencies and capital flows, which is very helpful for users interested in emerging tokens.

From a technical implementation perspective, The Hive defines multiple specialized Agents, such as Wallet Agent, Staking Agent, Market Agent, etc., each responsible for operations in specific domains. The core logic is to analyze user input through LLM, select the appropriate Agent, and then organize the workflow and return results through the AI Agent framework.

Summary

Web3 has long faced the issue of poor user experience, hindering the large-scale adoption of innovative applications like DeFi. The proposal of "intention trading" offers a new approach to address this problem by allowing users to define the desired outcomes and entrusting complex operations to third parties.

The emergence of DeFAI further simplifies this process. By utilizing AI technology to provide a natural language interface, it enables users to conduct DeFi transactions and management more intuitively, significantly lowering the operational threshold. This not only improves the user experience but also brings real value to Web3.

DeFAI is expected to fundamentally change the interaction paradigm of Web3, allowing more users who are unfamiliar with blockchain and DeFi to easily participate in the Web3 ecosystem. With the continuous development of technology, we can expect to see more innovative applications and broader adoption.