Intent: The Starting Point for Intelligent Interactions in Web3

This article combines insights from both Anoma and TG Bot to deepen our understanding of the evolution, interaction paradigms, current state, and trends of the concept of ‘intent.’ Finally, it looks forward to the challenges and future of Web3-enabled intelligent interactions.

Abstract

(1) Understanding Intent: Top-down Anoma vs Bottom-up TG (Telegram) Bot

(2) Delving into Intent: AI is the New UI: Complex Commands (Command-Based Interaction) vs. Simple Intent (Intent-Based Interaction)

(3) Navigating Intent: The Evolution of the Concept of Intent, Interaction Paradigms, Current State, and Trends

(4) Guiding Intent: Challenges and Prospects of Intelligent Interactions in Web3

In June, Paradigm’s official website published an article titled <Intent-Based Architectures and Their Risks> bringing the concept of ‘Intent’ to the public eye for the first time. Protocols and infrastructure related to ‘Intent’ developed rapidly and became an unavoidable topic at the ETHCC conference in July.

Intent is not a new concept. As Mindao, the founder of DForce, stated: The ongoing trend in the crypto space has always been to abstract and automate all operations to an extreme level. Aggregators, CEX, contract wallets, cross-chain DeFi are all working on these aspects, as are Chainlink and the recent developments in automated middleware like Telegram bots.

Yet, Intent introduces new changes. Previously, the focus was on product interaction, but in the AGI (Artificial General Intelligence) era, the core is human-machine interaction. AI/Large Language Models (LLMs) offer even greater potential for enhancing Crypto interactions.

01 Top-down Anoma vs Bottom-up Unibot (TG Bot)

In the past six months, protocols, projects, and infrastructures related to ‘Intent’ have successively gained prominence. Rather than rehashing the concept, let’s focus on two excellent examples for a glimpse into the phenomenon.

Among the various projects related to ‘Intent,’ the most dazzling is undoubtedly the Anoma Foundation, which completed its third round of financing with $25 million at the end of May 2023. In a landscape where Layer 1 solutions are increasingly homogenous, the Anoma Foundation has received a total investment of $57.8 million for its intent-centric architecture, which includes the full-stack Dapp architecture ‘Anoma’ and ‘Namada,’ a privacy-centric Layer 1 solution. This places them seventh in fundraising among Layer 1/Layer 2 projects that have not yet issued tokens.

Looking back at Anoma’s public presentation at EthCC, founder Adrian Brink pointed out the magic of Anoma lies in: all interactions start with ‘Intent.’ User intentions pass through the Black-box Architecture (also known as the ‘Magic Happens’ architecture) of Anoma for processing. This ‘Magic Box’ implements the core transaction flow based on intent-driven interactions:

• User Interactions: Users can send transparent, private, or protected intents to Anoma’s ‘Black-box’ architecture (the intent gossip layer).
• Intent Collection and Matching: Anoma’s ‘Black-box’ architecture employs solvers to collect intentions and perform balanced state transitions to match trading counterparts.
• Transaction Processing: Matched transactions are submitted to the encrypted mempool（memory pool）, and validators then send the block packaged by the proposer to the execution layer, complete the execution and verify its validity, ultimately updating the state root.

Source：https://twitter.com/Delphi_Digital/status/1696626180752056764

Behind the new forms of interaction, the Anoma team has identified the pain points at the foundational level of blockchain protocols. Reflecting on the journey from the first generation of scriptable settlements in Bitcoin to the second generation of programmable settlements in Ethereum, they note that existing application protocols still include at least one Web2 component, making it impossible to accomplish counterpart discovery and resolution. They argue that Anoma’s intent-centric architecture represents the third generation of Dapp architectures in the evolution of blockchain protocols. This allows users to define their desired end states and achieve efficient and customizable private transactions at the intent level. An era of intent-centric Dapp architectures is on the horizon。

Source：https://www.youtube.com/watch?v=mdYwfW6tMJ8

With high financing and a new technological architecture, Anoma is currently built by the Heliax development team, consisting of 37 interdisciplinary members. After two years of slow yet consistent progress, Heliax has formed a complete innovative ecosystem：

• Anoma: A general-purpose Dapp architecture centered around ‘intent,’ which can be modularly deployed to L1, L1.5, and L2, and integrated as modular components in various EVM frameworks.
• Namada: The privacy settlement Layer 1 of the Anoma ecosystem, utilizing PoS mechanisms and dedicated to asset-agnostic inter-chain privacy. It plans to be compatible with Ethereum and IBC chains in the future.
• AnomaVM: Simplifies Dapp development on Anoma, using Juvix (an intent-centric programming language) and VampIR (a circuit programming language).
• Typhon: An improvement over Cosmos’s Tendermint, known as Typhon consensus.
• Taiga: Provides a unified execution environment with composable privacy. Taiga defines three types of intents: transparent intent, shielded intent, and private intent.

Source：https://heliax.dev/#projects

While many institutions like CMCC Global, Electric Capital, and Delphi Digital are highly bullish on Anoma and express that its intent-centric architecture has limitless potential, Anoma has only laid a strong technological foundation without yet delivering exciting applications. On the other hand, TG Bot has been unlocking specific use-cases for ‘intent’ through on-chain automated trading tools.

Unibot is a trading bot based on Telegram (TG Bot), allowing users to automate DEX trades via Telegram, offering functionalities like sniper buys, follow trades, DEX limit orders, privacy, and MEV resistance. Unibot replaces Uniswap’s cumbersome interactions based on ‘intent,’ offering a convenient DeFi interaction experience. It has also spurred a slew of TG Bot clones, pushing the total market value of the TG Bot space close to $200 million.

Top-down Anoma represents innovation, while bottom-up TG Bot represents transformation. Although Anoma is slow in progress, it starts with an intent-based foundational architecture. TG Bot brings a new on-chain interaction interface to users, but it’s not intelligent and has various security risks.

Both approaches share the same goal—to simplify user interaction experiences and introduce new programmable, customizable user interaction interfaces, known as the User Intent Layer. This allows users to bypass complicated on-chain interactions and define trading states based on user intent.

02 AI is the New UI：Command-Based Interaction vs. Intent-Based Interaction

With a basic understanding of ‘intent’ from the two examples above, it’s now time to explore the foundational changes AI brings to the crypto world behind the scenes—changes in the user interaction interface. The AI paradigm is introducing the third user interface paradigm in computer history, shifting towards a new interaction mechanism where users tell the computer what they want, rather than how to do it.

• Paradigm 1: Batch Processing.
• Paradigm 2: Command-Based Interaction Design.
• Paradigm 3: Intent-Based Outcome Specification.

Historically, each revolution in interaction modes has given rise to entirely new business models. Generative AI based on LLM (Language Models) has brought disruptive changes to human-machine interaction, directly turning the original interaction with various software UIs into interaction with ChatGPT chat windows. This is an unprecedented interactive experience. LLM + crypto brings a whole new proposition to the crypto world based on intent-based interaction. LLM will also potentially make blockchain interactions smarter by discovering and describing user intent.

03 The Evolution of the Concept of Intent, Interaction Paradigms, Current State, and Trends

After deeply understanding the AI-driven trends behind the concept of ‘intent,’ we further explore the evolution of this concept. Intent is not a new idea; it has existed since the Web2 era. With the rise of search engines like Google, users can input their intent, and the search engine provides related search results to satisfy that intent.

With the emergence of e-commerce platforms like Amazon and eBay, the architecture based on ‘intent’ underwent significant transformation. Now, users can express their intent to purchase products, and the platforms handle the logistics. In early 2017, Gartner released a report titled ‘Innovation Insight: Intent-Based Networking Systems,’ formally introducing the concept of Intent-based Networking. The key to intent-based networking is to present users with an interface where they only need to express what they want, and the platform takes care of the business details.

Intent-based networking is the trend of automation and intelligence in the Web2 era. Andrew Lerner, the Vice President of Research at Gartner, pointed out in 2017 that intent-based networking would be the next milestone in the networking field.

Source：https://www.gartner.com/en/documents/3599617

The evolution from Web2 to Web3 is somewhat unclear in terms of the concept of ‘intent,’ but there are some common definitions:

• Paradigm: Intent is a set of declarative constraints that allow users to delegate transaction creation to a specialized third-party network while retaining complete control over the process.
• Anoma: Intent is a message sent by the user to express custom preferences; it serves as constraints defined by the user, not the specific execution path.
• Propellerheads: Intent is a part of the transaction that needs help from others, simplifying complex transactions by outsourcing difficult parts to third parties.

In real Web3 use cases based on intent, users create intents off-chain and outsource them to Solvers, encapsulating blockchain interaction complexities. This lowers the threshold for on-chain interactions.

Source：https://www.brink.trade/blog/powerful-intents-part-1

Intent also introduces a new architectural concept—Intent Centric, which has crypto protocols and infrastructures embedded in the user interaction layer, providing a better on-chain experience with the help of LLM.

The combination of LLM+Crypto brings a new human-machine interaction paradigm, where users’ intents can be directly converted into smart contract calls. The user only needs to express their intent like using Apple’s Siri, and robots/AI Agents/third-party solvers help complete complex on-chain operations. Web3 interactions based on intent have the potential to significantly reduce the complexity of user interactions on the blockchain.

AI is the New UI, currently more and more Web3 projects are exploring the magic of intent. Whether it’s intent-based Dapps or intent-based underlying architectures, they are all unleashing the potential of intent from different angles. We try to organize this rapidly developing field, and from the perspective of interaction, divide it into four major categories and track the latest developments and trends:

Basic Architecture: (Intent-centered intent architecture layer)

• DappOS：The first Web3 operating protocol centered on intent, building an intermediate layer between users and public chains, cross-chain bridges, and other cryptographic infrastructure, allowing Web3 users to seamlessly and imperceptibly interact with products. Recently, DappOS released version V2;
• Flash 2.0 SUAVE：SUAVE is an independent plug-and-play modular MEV whole-chain sorting layer. Preferences (the core concept of SUAVE) and the concept of Intent are similar, both providing ordinary users with customized trading methods, ultimately helping users to achieve optimal execution. Flashbots plans to launch SUAVE Centauri in 2023 Q4;
• Anoma：An intent-centered universal Dapp architecture that can be modularly deployed to L1, L1.5, L2, and embedded in various EVM frameworks through modular components. The latest development in the Anoma ecosystem is that Anoma’s privacy settlement Layer1—Namada is about to launch its v0.22.0 testnet;
• Cow Protocol：CoW Protocol builds a basic settlement network for traders and solvers and released Cow Hooks, a brand new intent system for customizing DeFi operations, in July. CoW Protocol’s latest data processed a total transaction volume of 1.05B and generated a profit of 5.3M.

Supporting Facilities: (Related to intent, related to account abstracted wallets and supporting infrastructure)

• ERC 4337: ERC 4337’s User Operation addresses the user’s intent, including tokens for paying gas, payment options, login methods, replacing the current transaction pool, thereby achieving account abstraction and improving wallet interaction experience.
• Facilities related to account abstraction: Packaging infrastructure (StackUp, Blocknative, Alchemy Rundler, AA-Bundler, infinitism Bundler), Wallet SDK (Safe, ZeroDev, Biconomy);
• Facilities related to intent: Juvix (intent-based programming language), Seaport (open-source NFT trading protocol), Gelato (Web3 automatic execution layer providing automation services for smart contracts), okcontract (low-level intent automation tool), Delegatable (smart contract authorization proxy tool), Hyper Oracle (off-chain automated services), symmio (intent-based on-chain derivative trading framework).

Enterprise Applications: (Integrated Dapp intent infrastructure, API, modular intent layers, domain-specific solvers)

• Transaction Solver API: PropelleHeads;
• Intent API: EnsoFinance;
• Intent trading infrastructure: Bob the Solver, Brink;
• Composable intent tools: Essential.

Consumer Applications: (User-interactive smart interfaces such as Wallets, Dex, Web3 Ai Agents, smart search engines)

• Wallets: Safe, Bitconomy, Argent, Ambire, Sequence, Versa, A3SProtocol;
• Dex: CowSwap, Paraswap, Uniswap, 1inch, BananaHq, basedmarkets;
• Intelligent front-end: Unibot, Dora, Web3 Analytics, kaito.ai.

Architecture based on intent, supporting facilities, and enterprise applications (due to space limitations, only listed here without further expansion) try to solve underlying challenges based on intent, but infrastructure can’t be built overnight, the trend of intent consumer applications is truly exciting. Wallets, Dex, and other interactive entries show the potential of intelligent interaction. Wallet intelligence: ERC-4337 introduced a new (UserOps) user intent layer, users express intent, and then packagers (Bundlers) convert these intents into executable signed transactions. In recent months, the number of ERC4337 users has increased significantly, ZeroDev, Biconomy, and Safe modular smart accounts are most motivated to promote AA+intent, AA+intent has the potential to catalyze the advent of intelligent wallets (SCW).

Source：https://dune.com/niftytable/account-abstraction

DEX architecture change: Intent-Based Trading aims to improve capital efficiency as well as user interaction experience. CoW Hooks string together complex actions like trading, bridging, staking, depositing, etc., UniswapX already allows users to sign intent off-chain matching, on-chain settlement, similar to BananaHq, Brink, basedmarkets, etc., Intent-based DEXs are gradually increasing, new Intent-based RFQ narrative (SYMMIO) is unfolding, as more and more DEXs and aggregators are moving toward intent-based architecture, the DEX landscape is undergoing significant changes.

Source：https://twitter.com/BananaHQio/status/1694013407929020740

Intelligent interaction entry: Similar to TG Bot for the transformation of the Web3 front-end interaction interface, interaction entries focus on understanding user intent from the start and systematically converting it into automated, actionable tasks. Not just TG Bot-type trading robots, Web3 AI Agents, Web3 smart search engines are making Web3 interaction more intelligent.

Source：https://twitter.com/awasunyin/status/169540582237605496

04 Challenges and Prospects of Intelligent Interactions in Web3

In the past, the user interaction experience of Web3 products indeed hindered widespread adoption; the wealth effect on-chain has masked many product issues. Now, liquidity and user attention are dispersed, and on-chain liquidity is scarce.

At the ETHCC conference in July, developers widely discussed the intent-centric future, such as how to help users more intelligently complete DAPP interactions in Web3. However, intent-centric implementation still faces many challenges:

• Lack of intent-related programming languages (like Juvix)
• Lack of suitable intent architectures (like Anoma)
• Lack of domain-specific solvers (like Bob The Solver)
• Centralized and closed implementation of intent (like CowSwap)
• Lack of front-end intent recognition parsers (like Unibot)
• Lack of composable intent implementation (like Brink)
• Security risks related to intent (mentioned in Paradigm’s article）

1. There are challenges but also hope. With the emergence of intent-centric protocols, future Web3 has the potential to evolve into a more user-friendly smart interaction interface. Users can interact with mainstream applications like Uniswap, Blur, AAVE in a one-stop manner; interaction will be as simple as hailing an Uber, offering a real opportunity for Web3 Dapps to outperform Web2 in terms of user experience. More importantly, combined with Anoma, Flashbots SUAVE, and DappOS, future intent-centric interactions on Web3 will also be more intelligent, customizable, and fair. At the same time, putting the focus on intent will return control to users. Ordinary users can also arm third-party robots/solvers to complete customized transactions (fees, slippage, privacy, cross-chain, MEV, or other intents), so users are no longer passive participants on the chain or subjects to the exploitation of MEV Searchers and Validators, but the real helmsmen of on-chain activities. Looking forward to the future of intelligent interactions in Web3, whether it’s account abstraction, chain abstraction, or intent, Web3 is finally paying attention to user experience, and the dawn of intelligent interactions is about to illuminate the dark.

Acknowledgments：
I would like to express my sincere gratitude to Web3 Analytics, Crypto V, Haotian, Jason Chen, Luke, Grace Deng, SixSix.eth, POOR DAO, #017, armonio.eth, and Kiwibig.eth for their valuable discussions, comments, and feedback on this paper.

Disclaimer：

1. This article is reprinted from [AC Capital]. All copyrights belong to the original author [AC Capital]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.