The Next Booming Trend of AI Agents: DeFAI

AI Agents on the blockchain experienced a brief highlight moment in early January this year. The market capitalization of Virtuals and ai16z surged to $5 billion and $2.5 billion, respectively, while underlying framework projects like ARC and Swarms saw their market cap exceed $600 million. DeFAI projects such as BUZZ and Grift also reached a market cap of $200 million. Traditional Web2 AI Agent projects like Pippin and AIOS have also entered the cryptocurrency space to issue tokens. The entire industry is immersed in the atmosphere of the "AI Agent revolution," looking forward to the arrival of the "DeFAI spring."

However, after a well-known figure issued a Meme coin, the liquidity of AI Agent tokens rapidly diminished. The emergence of DeepSeek further burst the AI bubble. The market capitalizations of Virtual and ai16z fell below $1 billion and $500 million respectively, while other AI Agent tokens generally dropped by over 90%. The AI Agent sector has plunged into a severe winter, with once beautiful visions turning into illusions, leaving behind impractical applications and investors deeply trapped.

Due to the relatively small overall market capitalization of the AI Agent sector (less than $7 billion), many people are not familiar with this field. Therefore, we will first provide an overview of the AI Agent sector, and then analyze why DeFAI is expected to become the second growth curve for AI Agents and potentially bring about a market opportunity worth hundreds of billions.

Overview of the AI Agent Section

AI Agent is an intelligent entity that can act autonomously, execute tasks, and interact with external tools.

The core of the AI Agent is model services, with main participants including ChatGPT, Anthropic, Gemini, DeepSeek, and Llama. However, it is not just a conversational large language model; it also needs to manage its own state (such as conversation history and memory), call various tools, and execute tasks securely.

In the Web2 domain, AI Agent projects mainly focus on enterprise services. In the Web3 domain, the main types include AI Agent launch platforms, frameworks, infrastructure, agents, and DeFAI.

Launch Platform

Launch platforms are currently the type of project where AI Agents can capture value most easily. They allow users to create, deploy, and monetize AI Agents, similar to pump.fun in Meme coins, but focused on AI Agents.

Virtuals is an AI Agent asset issuance platform launched on the Base network and is currently the largest launch platform, with over 100,000 Agents issued on it. Its ecosystem has given rise to several well-known AI Agents, such as the virtual host Luna, the market analysis provider AIXBT, and G.A.M.E, which provides developers with an Agent development framework.

The Virtuals team was originally established in 2021, originating from the gaming guild PathDAO, focusing on enhancing the gaming experience through AI-driven NPCs. In 2023, the release of GPT made the team realize the importance of AI, and Virtuals subsequently established AI as its main development direction.

framework

The framework is a toolkit for developing AI Agents. An official blockchain platform has listed five main framework projects: SendAI's SolanaAgentKit, ai16z's Eliza, ARC's Rig, Crossmint's GOAT, and Zerebro's ZerePy.

Among them, ai16z is currently the second highest market value AI Agent project. The concept of ai16z is quite complex; in addition to including the Eliza framework, it is also a venture capital fund focused on the AI sector, with plans to develop into a Layer 1 public blockchain for AI Agents in the future.

ai16z was founded by @shawmakesmagic on the Daos.fun platform. The core idea of the project is to use AI models to simulate the investment decisions of well-known venture capital firms and their co-founders, combining them with suggestions from DAO members for investments. This innovative investment model aims to provide investors with a way to organically combine AI trading strategies and decentralized governance, enhancing the transparency and credibility of investments.

Infrastructure

Infrastructure refers to off-chain platforms or tools built around AI Agents, covering a wide range of services. For example, AI Agent data aggregation services, AI Agent testing platforms, Agent resource sharing networks, AI Agent model fine-tuning training services, and dedicated Layer 1 public blockchains for AI Agents.

A certain data aggregation platform tracks over 1,000 AI Agents, building an index and data layer that supports the display of mind share, intelligent engagement, participation, and on-chain data performance. Its core function is mainly to aggregate data sources, allowing all AI Agents to compare and compete in terms of mind share, market share, and more.

A certain testing platform is dedicated to creating a decentralized environment that allows companies to test their AI models and agents in a distributed system. By identifying vulnerabilities before product deployment, it aims to enhance the security of AI models in real-world applications.

A certain resource-sharing network is a platform for Agent resource sharing, using a unified decentralized protocol to connect Agents from all frameworks. Its core is a P2P resource-sharing protocol that enables Agents to securely share their functionalities.

A certain laboratory has developed an AI model tailored specifically for a particular blockchain ecosystem. The model that has been launched can assist in completing a wide range of tasks related to that ecosystem, such as market analysis, risk assessment, and project planning.

A certain public chain is the first high-performance blockchain that supports full-chain AI, featuring elastic block space, parallel execution, and a modular architecture, which gives AI behavior strong verifiability and scalability, creating possibilities for complex, independent, and transparent AI.

Intelligent Body

Intelligent agents are specific AI Agents, and the vast majority of AI Agent projects on the blockchain belong to this category. In addition to the previously mentioned AIXBT and Luna, there are several other intelligent agent projects worth noting.

An AI idol connects with fans on various social media platforms through live broadcasts or text outputs, releases songs on music platforms, and can also incentivize users to participate in activities through on-chain transfers.

A market analysis agent is created based on a specific blockchain platform. It monitors cryptocurrency-related social media and market trends through intelligent analysis tools, providing users with valuable market insights. Some of the analysis content will be shared on social platforms, while the rest is accessible only to token holders. Holders can directly converse with the agent through their exclusive terminals. The agent's analysis has a certain level of accuracy in predicting price trends, enhancing users' confidence in its AI-driven market insight capabilities.

A certain adversarial agent game is an agent that controls a prize pool. The goal of the game is to persuade the agent to send funds from the prize pool to the user. Anyone in the world can send messages to the agent by paying a query fee.

An intelligent agent on a certain blockchain focuses on artistic creation and content generation, capable of independently creating music, memes, artworks, and NFTs. The team has also developed a Python-based AI agent framework.

Another agent project is both an agent and a framework, with its concept initially being an abstract idea published on social media by the well-known AGI framework's author. This image originates from a new AI Agent he plans to create, which he sent to a certain large language model, ultimately naming it. The author also launched an AI Agent framework for creating "self-feedback" mechanisms for agents.

DeFAI

DeFAI is a term created by a well-known figure, who is the founder of a certain AI project and also an important person behind a well-known DeFi project launched during the 2021 DeFi Summer. Conceptually, DeFAI is the combination of decentralized finance (DeFi) and artificial intelligence (AI), aimed at simplifying the complex processes of DeFi through AI technology, allowing ordinary users to easily use DeFi services. DeFAI is currently the most likely application to realize the integration of AI agents with blockchain.

A certain intention-based blockchain terminal aims to simplify the complexity of on-chain transactions through agent technology, allowing users to complete operations more easily. The project also supports users in creating their own agents to engage in various social platform interactions and various DeFi transactions. With the support of a well-known blockchain platform, the project has garnered widespread attention on social media.

A certain DeFi operation simplification platform focuses on simplifying DeFi operations and obtaining market information, providing users with a smooth interactive experience. The AI Agent token launched by this platform has also garnered market attention. It is one of the most comprehensive DeFi platforms, developed by a certain team (funded by well-known exchanges and tech giants) and supported by a certain well-known DAO.

A certain DeFi trading assistant developed by a certain AI development team, as the champion project of a certain blockchain AI hackathon, utilizes AI technology to provide a natural language interface, enabling users to conduct DeFi trading and management more intuitively, thereby lowering the technical barrier. This project is based on a cellular architecture and a certain blockchain's Agent development toolkit, which can automatically handle complex DeFi operations such as trading, staking, and lending through users' chat commands, and can also be customized into tailored strategies according to users' needs, such as setting limit order trades.

Why are we optimistic about DeFAI

The story of AI agents is far from over; it just needs practical applications to sketch the second growth curve. Currently, DeFAI seems to have the greatest opportunity.

Firstly, from a macro perspective, the future will definitely be an AI world, and for Web3, integration with AI is inevitable. The most direct way is for AI to improve Web3. The certainty of this direction is very high, even higher than Web3 improving AI. Currently, the most pressing issue in Web3 is the high user threshold. AI agents can guide users through their entire journey in the Web3 world, assist in creating and managing on-chain transactions, answer questions in real-time and provide help, and even complete complex intent transactions.

Secondly, from a market perspective, the abrupt halt of the entire AI Agent narrative is largely due to a "crisis of trust" brought about by certain events. Various celebrities issuing tokens have turned the blockchain into a purely PvP game, competing on who can enter early and run faster. The AI Agent, as a utility token, has also been unfairly slaughtered by the market, with its trend resembling that of celebrity Meme coins.

Therefore, applications need to emerge that allow users to truly experience the improvements that AI brings to Web3. There needs to be applications that help the market form the understanding that AI Agents are not just memes. This leads to DeFAI as a new narrative direction for AI Agents, which represents the second growth curve. The second growth curve is about finding another new high-growth curve before the peak arrives or disappears, in order to achieve sustained growth.

In fact, the narrative of AI Agents only started at the end of last year, with a peak market value of only 13 billion USD. There was only a brief surge in the launch platforms and frameworks, but for the narratives surrounding infrastructure, agents, and DeFAI, many good projects are still in development and progress, and have not been fully understood by users, nor have they formed a breakthrough effect.

For infrastructure, it is a business of shovels, so it is naturally difficult to become the protagonist; for intelligent agents, the Meme attribute is too strong, and tokens usually struggle to provide empowerment; however, for DeFAI, AI can help users lower the threshold for DeFi trading and management, aligning with AI's improvements for Web3, and the tokens of DeFAI are easier to empower. For example, advanced features such as custom AI models, custom trading strategies, and quantitative robots can be unlocked through token payments or token staking.

In the first curve, the AI Agent framework acts as a glue, binding social platforms and various chains together. However, the framework itself is a tool, and it is very difficult for tool-type products to commercialize in Web2. Although there are tokens in Web3, the framework itself cannot capture value, so it must overlay community attributes to develop into a platform. If we only create a platform for launching agents, it will revert to the model of certain projects. How do agents capture value, and what is the difference between agents that cannot capture value and memes? DeFAI is clearly the application that captures value most easily, and the AI Agent framework supporting DeFAI is also a trend. Moreover, it is foreseeable that the market will subsequently see many new frameworks specifically supporting DeFAI applications.

Summary

In the summer of 2020, DeFi projects were in