Author: Nishil, Biconomy Ecosystem Cooperation Director; Translation: Jinse Finance xiaozou
AI agents will completely change the way we interact on the chain. Let's take a look at several major protocols in this field that are committed to this.
When we hear people talking about AI x Crypto, we basically hear people mention AI projects such as data collection, GPU computing, or data inference that utilize encryption technology. While these projects are creating fascinating applications, their impact on everyday users of web3 is still limited.
There is a new set of protocols in the Crypto X AI space that have the potential to attract the masses to crypto. We all admit that operating complex crypto dapps can be a hassle even for the most experienced users. However, AI can solve this problem!
These AI agents are at the crossroads of crypto and AI and are designed to solve the thorny user experience problems that exist in the crypto space. Imagine a future where you can just tell the AI agent what you want to achieve on-chain, and it will automatically write and execute the necessary transactions for you.
These agents will help us build an intelligent layer on top of the existing DeFi rails. This intelligent layer will be the web3 equivalent of bankers, investors, traders, fund managers, etc. in traditional finance. They will leverage underlying primitives to conduct on-chain transactions.
1. Classification of AI Agents
Based on the type of use cases and the added value brought by the protocol, the evolving field of AI agents can be roughly divided into the following categories:
These AI agents are designed to manage users’ assets, and their goal is to maximize users’ returns by intelligently allocating assets to various DeFi strategies.
Imagine that in the future you can just tell the AI agent what goals you want to achieve on the chain, and it will automatically write and execute the necessary transactions for you. These interfaces can help users interact on-chain by communicating directly with the agent.
Game agents are designed to act on behalf of the user or to assist the user as an in-game pet/assistant. These agents can also interact with other agents in the game and trade assets. The field of game agents is still a relatively new field.
2. AI Agent Protocol
Let's take a closer look at several of these AI agent protocols:
Enable developers to build and deploy autonomous agents that can work together seamlessly off-chain while leveraging blockchain to enhance on-chain functionality.
polywrap is building a network of professional AI agents that perform complex tasks for web3 users and protocols. Current agents are involved in payments, market research and trading, social content management, forecasting, and public goods funding.
Colony is developing a new AI-driven Web3 survival game. In Colony, AI agents or "characters" continuously learn from their surroundings. Players must guide and cooperate with these characters in an effort to survive a future Earth filled with colonial competition. Colony stands out by integrating continuous learning into gameplay. AI characters develop unique personalities and worldviews, learning to acquire new knowledge and skills from their own experiences, identities, and goals. In addition, these characters can trade with other in-game characters.
Wayfinder Foundation is creating a "map" for AI agents to help them handle tasks and simplify users' on-chain activities. Through open source development and incentivizing builders with PROMPT tokens, Wayfinder will expand the Map Command Network. It aims to connect blockchains and off-chain data sources, allowing users to easily perform tasks through the command prompt.
Noya is a decentralized finance (DeFi) protocol that enables AI agents to securely and precisely manage liquidity across multiple blockchains. Noya has multiple vaults, each catering to the needs of their respective user intent profiles. Brian provides an API for developers to integrate into their applications, providing users with the ability to generate web3 transactions, conveying their intent through prompt words, such as "Can you redeem 10 USDC for ETH on Uniswap on the Ethereum mainnet?"
Brian provides an API for developers to integrate into their applications, providing users with the ability to generate web3 transactions, conveying their intent through prompt words, such as "Can you redeem 10 USDC for ETH on Uniswap on the Ethereum mainnet?"
Fungi leverages the power of smart accounts and AA to provide a self-hosted AI agent experience. It allows users to instruct the command prompt through its interface, and then it will process real-time blockchain data and autonomously perform operations based on the user's instructions.
The Fyde Treasury protocol enables users to increase their crypto assets faster by depositing into various AI-managed vaults, which can lock in returns and reallocate assets based on market performance. Users can deposit various tokens into these vaults to obtain corresponding TRSY.
3. Growth Potential
In the upcoming AI&Intent projects, potential use cases range from abstracting simple tasks to empowering AI agents to execute complex DeFi strategies to discover the best returns for users.
However, these AI agents face two major challenges.
Challenge 1: They cannot be truly autonomous. Currently, AI agents can recommend on-chain operations for users and prepare transactions, but still require user signatures or approvals.
Challenge 2: If they choose to automate, they lose security. Protocols tend to explore alternative solutions to automation, such as approvals, central vaults, shared private key pairs, etc., which make the protocol the custodian of your assets and introduce significant risks.
We need to set guardrails for AI so that it does not sign transactions that it does not have the authority to sign. Therefore, we need a solution that can delegate transaction authorization to AI agents, but only conduct transactions within specific permissions and rules.