Intent-driven Best practices for Web3+AI?

Best practices for Web3+AI driven by intent?

Author: Kyle

Editor: Wendo

In the 8 years since the birth of the open-source blockchain network Ethereum, numerous competing chains have emerged, laying the foundation for the development of on-chain applications in finance, gaming, and other scenarios. The only regret is that after 8 years, the vast majority of internet users still cannot access the Web3 gateway based on blockchain.

According to data, out of a total of 5.16 billion internet users, less than 100 million users have blockchain wallets (the main entry point for Web3 networks and applications). Why is it so difficult for Web3 to break out?

• In-depth analysis of Pendle LSDfi integrates RWA DeFi Lego to stack new puzzle blocks
• Exclusive Interview with Connext Adopting a Points Mechanism to Determine the Number of User Airdrops, with Plans to Expand to 20 Chains by the End of the Year
• LianGuai Daily | Grayscale Wins Lawsuit Against the US SEC; Ethereum Foundation Launches Ethereum Execution Layer Specification

On the one hand, there hasn’t been a popular application familiar to Web2 users, such as in the fields of social media and e-commerce. More importantly, the usage and operations are complicated: for example, when interacting with the blockchain, users need to calculate “gas fees” (one of the network usage costs), and every interaction requires a wallet signature (similar to providing authorization and verifying transactions).

Recently, a new concept aiming to revolutionize the Web3 user experience has emerged – investment institution LianGuairadigm proposes “intent-centric” and expounds on the new approach of establishing a Web3 interactive experience centered around intentions: ideally, users only need to issue an instruction of intent, and all operations can be automatically executed behind the scenes, requiring only one signature from the user to achieve the intention.

The intent-centric approach of focusing only on the result and not the process sounds very user-centric. Does this simplified style resemble what AI chatbots do? In fact, artificial intelligence can indeed be applied in this context.

1. The New Focus of Web3

A recent research article by well-known cryptocurrency investment institution LianGuairadigm has made the concept of intent-centric the new focus of the Web3 track. The institution listed intent-centric as the first of the “Ten Web3 Directions Worth Paying Attention To,” spreading this unfamiliar term in the cryptocurrency community and generating excitement.

Intent-centric, as the name suggests, means “centered around intentions.” It focuses on the result rather than the process. Through optimization of protocols and infrastructure, intent-centric aims to make cumbersome on-chain operations “one-step completion.” More accurately, it hides the previously complex operation process, allowing users to achieve their goals seamlessly and directly.

For example, if a user wants to exchange the Ethereum blockchain version of the stablecoin USDT for ARB tokens on the Arbitrum blockchain, the process is somewhat similar to an international interbank transfer. It usually requires a series of operations, including opening a cross-chain bridge (a trust solution across systems), connecting the wallet (account), transferring USDT and the gas fee ETH (transaction fee) from the Ethereum network to Arbitrum, waiting for the cross-chain transfer of assets to complete, and then finding a liquid exchange on Arbitrum to execute the exchange.

The above is a familiar and habitual operation for experienced users on the blockchain. There are a large number of users every day who transfer assets, conduct transactions, and stake assets across different blockchains, and the complex operation flow is very unfriendly, especially for newcomers. Completing this series of basic operations is a challenge at every step.

intent-centric aims to solve these problems by turning long chains of operations into short chains, or more precisely, making these chains “invisible” to the user experience.

In an ideal scenario, users only need to express their intent, such as “Help me convert the USDT in my wallet to ARB on Arbitrum.” The intent-centric protocol will automatically perform the necessary operations behind the scenes, including cross-chain transactions, finding the optimal exchange path, paying for gas fees, until the conversion is completed. Throughout this process, users are unaware and only receive a clear result: the USDT in their wallet has been converted to ARB.

This is very close to the user experience of Web 2 applications. For example, when we use Alipay to make a purchase on Taobao, we simply wait for the arrival of the goods without needing to know how the money reaches the merchant.

The cross-chain transaction from USDT to ARB is just a simple example. In theory, as the concept of intent-centric matures and improves at the protocol level, any on-chain operation can be completed with just one click. For example, one-click purchase of a specific NFT, or one-click search for the highest annualized return investment product, and so on.

In short, the core idea of intent-centric is to optimize the user experience and enable even novice users to quickly navigate the world of blockchain.

So, how is the vision of intent-centric realized technically? In fact, its principle is not difficult to understand. It involves decomposing intent and assigning each step of the process to specialized protocols to complete.

Take Bob the Solver, which has performed well at the recent ETHGlobal LianGuairis hackathon, as an example. It is an infrastructure based on intent-based transactions, consisting of two parts: the Solver and the Account Abstraction Wallet (AA Wallet).

Bob the Solver brings an intent-centric template

The Solver acts as the coordinator, responsible for identifying user intent, categorizing intent, and planning the optimal path to achieve it. Once the path is determined, the Solver will construct the necessary transactions to implement the user’s intent, similar to creating a “process outline,” and then forward them to the programmable Account Abstraction Wallet.

The Account Abstraction Wallet is responsible for executing the work. It consists of a bundler and a LianGuaiymaster. The bundler plans the transactions sent by the Solver, while the LianGuaiymaster manages and pays for the related gas fees.

Bob the Solver provides a simplified template for the intent-centric track. Following this approach, as long as there are professional Solver solutions and wallet programming solutions, it is possible to achieve one-click access to intent in various scenarios.

2. Potential integration with AI

The on-chain application ecosystem has undergone 8 years of development, and there are already a large number of transaction-based, lending-based, financial-based, game-based applications, as well as new assets like NFTs. In the context of the growing on-chain ecosystem, the emergence of intent-centric comes at the right time. Only by achieving a transformative upgrade in user experience can the on-chain world achieve exponential user growth.

Currently, there are relatively few new protocols under the narrative of intent-centric. However, similar applications aligned with its vision have already emerged before.

Decentralized encrypted asset trading application 1inch is a typical case. Compared to the famous Uniswap, the characteristic of 1inch is that it allows users to trade encrypted assets across multiple DEXs in one transaction. By scanning multiple DEXs at the same time, 1inch can find the best price for a specific trading pair and execute the transaction for users at the most favorable price.

Before the emergence of 1inch, users who wanted to complete a transaction in a way that maximizes value usually had to visit multiple DEXs for comparison, while also considering factors such as trading slippage and gas fees. However, 1inch efficiently discovers the optimal exchange path by using a specific algorithm and aggregating dozens of DEXs, allowing users to complete the optimal transaction in real time at the best price.

1inch trading page

Aggregation is a simple and effective way to simplify user operations. However, the blockchain world is not just about simple transactions of A to B. There are hundreds of independent and open public chains, which also support a variety of on-chain application-like digital assets. Aggregation alone is not enough to solve the problem of most cumbersome operations.

Therefore, the realization of intent-centric vision still has a long way to go, and in this process, highly intelligent AI may become a powerful assistant.

Whether it is natural language input of intentions or decomposing goals, calculating the optimal path, and executing operations, AI can play its advantages.

In many cases, user intentions are complex, and the expression of intentions may not be accurate, which may cause the solver to have difficulty in accurately understanding the user’s intentions and therefore unable to plan the optimal solution. However, AI trained by specific models can more accurately identify user intentions and infer potential goals and needs based on the user’s transaction request sources, transaction data, etc.

In terms of decomposing goals and executing operations, AutoGPT based on GPT-4, a large language model launched by OpenAI, has proven its power. Just by issuing a task, AutoGPT can plan and execute automatically. Its characteristics are in line with the requirements of intent-centric.

In the eyes of many professionals, the realization of intent-centric cannot be achieved without the help of AI. After all, the efficiency of AI retrieval and execution is much higher than that of humans, and the involvement of AI will accelerate the arrival of the user-friendly era of blockchain.

Of course, when developers hand over the “intermediate layer” operations from intention to result to AI and other third-party execution layers, it means that there are multiple parties involved behind the “one-click implementation of intentions”, and security issues must be taken seriously. On the one hand, intent-centric protocol providers need to establish punishment mechanisms for malicious behavior and provide a secure and stable third-party execution layer. On the other hand, they also need to improve their technical security capabilities to prevent algorithms from being exposed or AI from being “deceived”. Once there is a problem in any link, the rights and interests of users will not be guaranteed.

Intent-centric presents an exciting future for the Web3 industry, and it is expected that more secure and user-friendly “intent” applications will appear, which will revolutionize the user experience of blockchain from the user’s perspective.

We will continue to update Blocking; if you have any questions or suggestions, please contact us!