Can the deep integration of DeFi and AI, known as DeFAI, give rise to a new wave of AI Agents?

Author: YBB Capital Researcher Ac-Core

1. What story does DeFAI tell?

1.1 What is DeFAI?

In simple terms, DeFAI refers to AI + DeFi. The market has gone through multiple rounds of hype around AI, from AI computing power to AI memes, and from different technical architectures to various infrastructures. Although the overall market value of AI agents has recently seen a decline, the concept of DeFAI is emerging as a new breakthrough trend. Currently, DeFAI can be broadly categorized into three types: AI abstraction, autonomous DeFi agents, and market analysis and prediction. The specific divisions within these categories are illustrated in the figure below.

Image source: Created by the author

1.2 How does DeFAI work?

In the DeFi system, the core behind AI agents is LLM (Large Language Model), which involves multi-layered processes and technologies, covering all aspects from data collection to decision execution. According to the research by @3sigma in the IOSG document, most models follow six specific workflows: data collection, model inference, decision-making, custody and operation, interoperability, and wallet management. The following summarizes these processes:

1. Data Collection: The primary task of the AI agent is to gain a comprehensive understanding of its operating environment. This includes obtaining real-time data from multiple sources:

• On-chain data: Real-time blockchain data such as transaction records, smart contract statuses, and network activity are obtained through indexers, oracles, etc. This helps the agent stay synchronized with market dynamics;

• Off-chain data: Price information, market news, and macroeconomic indicators are obtained from external data providers (e.g., CoinMarketCap, Coingecko) to ensure the agent's understanding of external market conditions. This data is typically provided to the agent via API interfaces;

• Decentralized data sources: Some agents may obtain price oracle data through decentralized data feed protocols, ensuring the decentralization and reliability of the data.

2. Model Inference: After data collection is complete, the AI agent enters the inference and computation phase. Here, the agent relies on multiple AI models for complex reasoning and prediction:

• Supervised and unsupervised learning: By training on labeled or unlabeled data, AI models can analyze behaviors in markets and governance forums. For example, they can predict future market trends by analyzing historical trading data or infer the outcome of a voting proposal by analyzing governance forum data;

• Reinforcement learning: Through trial and error and feedback mechanisms, AI models can autonomously optimize strategies. For instance, in token trading, the AI agent can simulate various trading strategies to determine the best time to buy or sell. This learning method allows the agent to continuously improve under changing market conditions;

• Natural Language Processing (NLP): By understanding and processing user natural language inputs, the agent can extract key information from governance proposals or market discussions, helping users make better decisions. This is particularly important when scanning decentralized governance forums or processing user commands.

3. Decision-Making: Based on the collected data and inference results, the AI agent enters the decision-making phase. In this stage, the agent needs to analyze the current market conditions and weigh multiple variables:

• Optimization engine: The agent uses an optimization engine to find the best execution plan under various conditions. For example, when providing liquidity or executing arbitrage strategies, the agent must consider factors such as slippage, transaction fees, network latency, and capital size to find the optimal execution path;

• Multi-agent system collaboration: To cope with complex market conditions, a single agent may not be able to optimize all decisions comprehensively. In such cases, multiple AI agents can be deployed, each focusing on different task areas, collaborating to improve the overall decision-making efficiency of the system. For example, one agent focuses on market analysis while another agent focuses on executing trading strategies.

4. Custody and Operation: Since AI agents need to handle a large amount of computation, they typically require their models to be hosted on off-chain servers or distributed computing networks:

• Centralized hosting: Some AI agents may rely on centralized cloud computing services like AWS to host their computing and storage needs. This approach helps ensure the efficient operation of the models but also brings potential risks of centralization;

• Decentralized hosting: To reduce centralization risks, some agents use decentralized distributed computing networks (like Akash) and distributed storage solutions (like Arweave) to host models and data. Such solutions ensure the decentralized operation of models while providing data storage persistence;

• On-chain interaction: Although the models themselves are hosted off-chain, AI agents need to interact with on-chain protocols to execute smart contract functions (such as trade execution and liquidity management) and manage assets. This requires secure key management and transaction signing mechanisms, such as MPC (Multi-Party Computation) wallets or smart contract wallets.

5. Interoperability: The key role of AI agents in the DeFi ecosystem is to interact seamlessly with multiple different DeFi protocols and platforms:

• API integration: Agents exchange data and perform interactions with various decentralized exchanges, liquidity pools, and lending protocols through API bridges. This allows agents to access key information such as market prices, counterparties, and lending rates in real-time and make trading decisions accordingly;

• Decentralized messaging: To ensure synchronization with on-chain protocols, agents can receive updates through decentralized messaging protocols (like IPFS or Webhook). This enables AI agents to process external events in real-time, such as voting results of governance proposals or changes in liquidity pools, thereby adjusting their strategies.

6. Wallet Management: AI agents must be able to perform actual operations on the blockchain, which relies on their wallet and key management mechanisms:

• MPC wallets: Multi-Party Computation wallets split private keys among multiple participants, allowing agents to conduct transactions securely without a single point of key risk. For example, the wallet used by Coinbase Replit demonstrates how to achieve secure key management using MPC, enabling users to maintain some control while delegating partial autonomous operations to AI agents;

• TEE (Trusted Execution Environment): Another common key management method is to use TEE technology, which stores private keys in protected hardware enclaves. This allows AI agents to conduct transactions and make decisions in a fully autonomous environment without relying on third-party intervention. However, TEE currently faces issues of hardware centralization and performance overhead, but once these challenges are resolved, fully autonomous AI systems will become possible.

1.3 Sect Origins? From Intent to DeFAI

Image source: Created by the author

If the vision of DeFAI is to enable users to autonomously manage their portfolios through AI agents and various AI platforms, allowing everyone to easily participate in cryptocurrency market trading, then does this vision naturally lead us to the concept of "intent"?

Let's revisit the concept of "intent" first proposed by Paradigm. When we trade normally, we need to specify a clear execution path, like exchanging Token A for Token B on Uniswap. However, in an intent-driven scenario, the execution path is matched and ultimately determined by solvers and AI. In other words: trading = I specify how the TX is executed; intent = I only care about the TX result but not the execution process. From a retrospective perspective, the narrative of DeFAI not only approaches the ultimate concept of AI agents but also perfectly aligns with the vision of realizing intent, making DeFAI more like a new added path for intent.

The ultimate version of achieving large-scale blockchain applications in the future will be: AI Agent + Solver + Intent-Centric + DeFAI = Future?

2. DeFAI Related Projects

Image source: Created by the author

2.1 Griffain

@griffaindotcom $GRIFFAIN: An innovative platform that combines AI agents with blockchain, helping users issue AI agents, focusing on creating a powerful and scalable decentralized finance (DeFi) solution that supports seamless token swaps, liquidity provision, and ecosystem growth. It allows easy management of wallets, trading, and NFTs, and automatically executes tasks such as Memecoin issuance and airdrops.

2.2 Hey Anon

@HeyAnonai $ANON: An AI-driven DeFi protocol that simplifies interactions, aggregates real-time project data, and executes complex operations through natural language processing, facilitating users' DeFi abstraction layer. DWF Labs announced support for the DeFAI project Hey Anon through its AI Agent fund, which launched on Moonshot on January 14.

2.3 Orbit

@orbitcryptoai $GRIFT: Simplifies complex DeFi interfaces and operations, lowering the participation threshold for ordinary people. It currently supports over 100 blockchains and more than 200 protocols (EVM and Solana), with the token GRIFT energizing the platform.

2.4 Neur

@neur_sh $NEUR: An open-source full-stack application that integrates LLM models and blockchain technology, designed specifically for the Solana ecosystem, achieving seamless protocol interaction using the Solana Agent Kit.

2.5 Modenetwork

@modenetwork $MODE: Positions itself as the central platform for AI x DeFi innovation on Ethereum Layer 2, allowing holders to stake MODE to earn veMODE, thus enjoying airdrops from AI agents, aiming to become part of the DeFAI Stack.

2.6 The Hive

@askthehive_ai $BUZZ: Built on Solana, integrating multiple models including OpenAI, Anthropic, XAI, and Gemini to perform complex DeFi operations such as trading, staking, and lending.

2.7 Bankr

@bankrbot $BNKR: An AI-driven cryptocurrency companion that allows users to easily buy, sell, swap, place limit orders, and manage wallets with just a message, planning to add token swapping and on-chain tracking features soon, with a vision to make DeFi accessible to everyone and achieve automated trading.

2.8 HotKeySwap

@HotKeySwap $HOTKEY: Provides a complete set of DeFi tools including AI-driven DEX aggregators and analysis tools, as well as cross-chain trading and analysis.

2.9 Gekko AI

@Gekko_Agent $GEKKO: An AI agent created by the Virtuals protocol, focusing on providing comprehensive automated trading solutions, specifically designed for prediction markets. The automated trading strategies of the GEKKO token include automatic rebalancing, yield harvesting, and creating new token index functionalities.

2.10 ASYM

@ASYM41b07 $ASYM: Provides AI-driven DEX aggregators and analysis tools that can identify high return investment opportunities and settle generated profits in $ASYM.

2.11 Wayfinder Foundation

@AIWayfinder $Wayfinder: An AI full-chain interactive tool launched by the card game chain game Parallel, designed to help agents navigate the on-chain environment, execute trades, and interact with decentralized applications.

2.12 Slate

@slate_ceo $Slate: A universal AI agent and agent connection infrastructure layer that translates natural language commands into on-chain operations, focusing on executing automated trading strategies, buying or selling under specific conditions, making on-chain operations as simple as thinking.

2.13 Cod3x

@Cod3xOrg $Cod3x: A Solana AI hackathon project that provides no-code development tools to build agents that can automate DeFi strategies, with its agent interface (Agentic Interface) being a tool that can perform complex operations using only intent expressions.

2.14 Almanak

@Almanak__ $Almanak: An AI agent with self-learning capabilities that can autonomously execute tasks, optimizing DeFi and gaming projects using agent-based modeling, with a mission to maximize protocol profitability while ensuring economic security through data science and trading knowledge.

2.15 HIERO

@HieroHQ $HTERM: A multi-chain smart tool for Solana and Base networks that allows users to autonomously complete transactions using natural language commands, including buying and selling tokens and performing simple token analysis.

3. What system does the AI Agent ultimately belong to?

Image source: Created by the author

Time is of the essence, and DeFAI projects are emerging like mushrooms after rain. After Bitcoin fell significantly below $90,000 on January 13, DeFAI-related tokens surged by 38.73% the next day according to CoinGecko data, with $GRIFT, $BUZZ, and $ANON seeing the largest increases. However, the financial direction of AI agents is worth pondering, as the current crossroads point towards the left side of Game and the right side of DeFi.

3.1 Left towards Game:

M3 (Metaverse Makers ) (@m3org) may be the most promising representative, composed of artists and open-source hacker communities suspected to be behind ai16z. The core team includes JIN (@dankvr), Reneil (@reneil1337), Saori (@saorixbt), Shaw (@shawmakesmagic), among others. However, the biggest real-world obstacle for Game is the resource-rich Web2 market, where no truly explosive AI game has emerged. The highly anticipated "Phantom Beast Palu" in January 2024 sparked controversy over whether AI design was used due to its extraordinary development efficiency, but the CEO ultimately denied this claim. Moreover, the long development cycle required for games seems to demand more market enthusiasm compared to the right side of DeFi.

3.2 Right towards DeFi:

The market capitalization ranking of projects is as follows: $GRIFFAIN, $ANON, $OLAS, $GRIFT, $SPEC, $BUZZ, $RSS3, $SNAI, $GATSBY, with GRIFFAIN and ANON together accounting for 37.29% of the total market capitalization of DeFAI.

GRIFFAIN: Built on Solana, currently ranks first in the DeFAI market capitalization leaderboard with a market cap of $457M and 103,000 followers on Twitter. Its core function is to complete directed transactions through generated wallets, enabling quick trades. Currently, it costs 0.01 Sol to mint The Agent Engine's NFT.

Hey Anon: Adopts a multi-chain model, currently supporting different public chains such as Sonic Insider, Solana, EVM, opBNB, etc. The sudden surge of $ANON is entirely driven by the aura of its founder Daniele (@danielesesta), who is also the founder of Wonderland, Abracadabra, and WAGMI. The traffic alone has injected considerable vitality into $ANON, making Hey Anon his next entrepreneurial project, currently ranking second with a market cap of $248M.

4. Conclusion

The emergence of DeFAI is not coincidental; the core characteristics of blockchain are well-suited for strong financial scenarios. Currently, whether towards the left with GameFAI or towards the right with DeFAI, both exhibit comparable market potential. The leftward direction of Game may lead to a continuation of the metaverse, where, with the help of AI, management of virtual assets, characters, economies, etc., can be achieved, drawing on the elements of AI agents to evolve memes for the autonomous and prosperous development of the metaverse.

The rightward development of DeFi will inevitably transition from passionate emotional speculation to a destination oriented towards actual value. The value of AI agents cannot rely on issuing memes to cater to market trends, but the continuation of the AI agent story must be supported by a DeFi-like yield nesting. The victorious king will not always wear armor, and the ultimate outcome of market competition is worth our anticipation.