🌏The Web3 Use Case

The Flexibility Conundrum

One very interesting challenge is how to define Intents, allowing them to be adaptable and solve a wide array of problems. Intents as a concept can serve multiple needs and solve many issues, but how do we set a definition to avoid semantic hindrances?

Overly rigid definitions can hinder their Web3 potential, becoming a niche for specific domains and leading to fragmentation in the space, e.g., Intent solution providers focusing on token swaps, others on yield, others yet on gaming, etc. Therefore, having defined guidelines is crucial. We've distilled it down to three key directives to follow:

Motivation

What fuels Intent solving is critical. User needs are central to the Intents solution, but motivation also encompasses incentives for all participants in a multi-party transaction. This aligns well with game theory and reduces friction for everyone involved.

Understanding the motivation behind an Intent is crucial so the system can tailor effective execution strategies. It's essentially the "why" behind the intent. The motivation can provide a deeper understanding of the user's goal. The users may only sometimes be aware of their motivation, but a well-designed system can deduce it from the context and the user's choice of words. This helps provide an accurate, contextually relevant, and personalized response.

Example

"I want to trade my NFT for tokens I own." is the need. This example lacks specificity, but the users' preferences and motivations can improve it, i.e., a bias toward spending stablecoin for the trade might include liquidity preferences (I prefer a stablecoin vs. Ether, which is illiquid due to price appreciation). The user would like to trade with reputable buyers. A system that grasps these objectives provides an Intent closer to the user's preferences and better execution paths. Thus, an improved Intent could be "I want to trade my NFT for USDC with a reputable buyer over the next 24 hours".

The User's Profile

In the Web3 ecosystem, users often have public profiles that include information about their transaction history, token holdings, and interactions with various decentralized applications (dApps). By analyzing this data with user privacy and consent in consideration, Intent-based systems can infer specific preferences and tailor the intent resolution process accordingly. For example, suppose a user's transaction history shows a higher frequency of interactions with DAI than USDC. In that case, the Intent resolution system can assume that the user prefers DAI as their stablecoin. This information can be used to prioritize DAI-based solutions when resolving the user's intents, providing a more personalized experience.

Intent definition language domain

Natural Language Processing (NLP) enables us to move beyond rigid keywords to genuine understanding. A well-defined Intent language domain isn't about restricting the user from expressing their objectives but providing language the system can interpret. Thus, the derived Intent execution plan clearly and concisely accomplishes the user's goal. This involves solving the user's explicit request, expressed through their words. By engaging in semantic understanding, we can "read between the lines" and unlock new levels of insight akin to the value we already derive from advanced LLMs. The Intent language domain should cover various possible utterances for the same goal and be flexible enough to accommodate variations.

Intent Structure and Semantic Flexibility: The Blueprint for Execution

The Intent structure includes the linguistic expressive freedom the user can use and the circumstances under which the Intent is valid. The structure also includes any semantic annotations that might be needed based on the user's request. It's important for the format to be Web3 flexible and robust to handle a variety of user inputs and scenarios. This ensures the system has sufficient cognitive understanding capacity to respond effectively, regardless of how users phrase their requests.

Last updated