Variant Fund Investment Partner: AI Agent has become a "first-class citizen" of the on-chain economy

author:Mason Nystrom

Compiled by: Xiaobai Navigation coderworld

Bots are becoming core players in the crypto economy.

Evidence of this trend is everywhere. For example, searchers deploy robots (such as Jaredfromsubway.eth) to exploit the human user’s need for convenience and profit from their decentralization by front-running transactionsexchange (DEX) trading. Banana Gun and Maestro Such tools allow users to easily conduct robot-supported transactions through the Telegram platform. These tools have long occupied the “Gas consumption"The top of the list. In addition,Emerging social applications like FriendtechIn the process, robots are also adopted by human users after the initialRapid intervention, and may have inadvertently accelerated the market's speculative cycle.

In general, robots are becoming a growing trend, whether for profit (such as MEV robots, where MEV stands for “maximum extractable value”) or for ordinary users (such as the Telegram robot toolkit).Blockchainpriority users on .

Although the functions of robots in the crypto field are still relatively simple, with the development of large language models (LLMs), robots outside the crypto field have evolved into powerful AI Agent (AI The goal is to enable agents to autonomously handle complex tasks and make intelligent decisions.

Building these in a crypto-native environment AI Intelligent agents have several important advantages:

• Built-in payment function：AI agents can exist outside of the cryptosphere, but if you want them to perform complex operations, they must have access to funds. Crypto payment systems are more efficient at funding AI agents than traditional methods (such as bank accounts or payment processor Stripe), while avoiding various inefficiencies common in the off-chain world.

• walletownership:passwalletAI agents can own digital assets (such as NFTs or earnings) and thus enjoy the digital property rights inherent in crypto assets. This is particularly important for asset transactions between agents.

• Verifiable deterministic operation：When AI agents perform tasks, verifiability of operations is crucial. On-chain transactions are inherently deterministic—either completed or not—and this property enables AI agents to complete on-chain tasks more accurately, while off-chain tasks have difficulty achieving the same certainty.

Of course, on-chain AI agents also face some limitations.

One major limitation is that AI agents need to execute logic off-chain to improve performance. This means that the agent’s logic and computations are hosted off-chain, but decisions are still made online to ensure verifiability of operations. In addition, AI agents can use tools like Modulus Such zkML (zero-knowledge machine learning) providers to verify the authenticity of their off-chain data inputs.

另一个关键局限性是，AI 智能体的功能依赖于其工具的丰富程度。例如，如果你希望智能体总结一条实时新闻，它需要具备网络爬虫工具来搜索互联网。如果你希望它将结果保存为 PDF，则需要为其配备文件系统。如果你想让它模仿你最喜欢的 Crypto Twitter 意见领袖的交易，则需要为其提供wallet访问权限和密钥签名功能。

From the perspective of determinism to non-determinism, most of the tasks currently performed by crypto AI agents are deterministic tasks, which means that humans have pre-set the parameters of the task and how it is performed (such as the specific process of token exchange).

Crypto AI agents evolved from early keeper bots, which are still widely used in DeFi and oracle services. Today, AI agents have evolved to be more complex. They can not only use large language models (LLMs) to achieve autonomous creation (such as Botto Such independent artists can also Syndicate Trading CloudProvide yourself with financial servicesIn addition, like Autonolas Such an early AI agent service market is also gradually taking shape.

Currently, many cutting-edge applications are demonstrating the potential of AI agents:

• AI Assistant in Smart Wallet：Dawn through its DawnAI The intelligent agent provides users with a multi-functional assistant that can help users send transactions, complete on-chain transactions, and provide real-time on-chain information (such as trend analysis of popular NFTs).

• AI Characters in Crypto Games: Parallel Alpha's latest game Colony Trying to create an AI character that can own a wallet and conduct on-chain transactions to add more interactivity to the game.

• Functional upgrades of AI agents：The capabilities of AI agents depend on the tools they are equipped with.BlockchainThe interaction is still in its early stages. Crypto AI agents need to have wallet functions, fund management capabilities, permission control,Integrating AI models, and the ability to interact with other agents. Gnosis has shown a prototype of this type of infrastructure, such as their AI Robot (AI mechs), encapsulate AI scripts into intelligentcontractso that anyone (includingOther robots) can call the smartcontractto perform tasks (such as participating in prediction market betting) and also pay rewards to the agents.

• Advanced AI Trader：DeFi super applications provide traders and speculators with more efficient operation methods, such as: automatic fixed investment (DCA) when conditions are met; when the Gas fee is lower than a certainXiaobai NavigationAutomatically execute transactions at thresholds; monitor newly issued Meme Tokens contract; and intelligently select the best order route, eliminating the need for users to manually search for access points.

• Vertical Application of AI Agents：Although large models such as ChatGPT are suitable for some general conversation scenarios, AI agents need to be fine-tuned specifically to meet the needs of different industries and segments. Bittensor Such platforms use incentive mechanisms to encourage developers to train models that focus on specific tasks (such as image generation, predictive modeling), targeting industries including encryption, biotechnology, and academic research. Although Bittensor is still in its infancy,Developers have begun to useIt builds applications and agents based on open source large language models.

• AI NPC in Consumer Applications：Non-player characters (NPCs) are common in massively multiplayer online games (MMORPGs), but are still rare in consumer applications. However, due to the financial nature of crypto consumer applications, AI agents can be ideal participants in innovative game mechanics. For example, open AI infrastructure company Ritual recently released Frenrug, which is an agent based on a large language model and runs on Friend.tech It can automatically execute transactions (such as buying or selling keys) based on the content of the user's message. Friend.tech users can try to convince this agent to buy their keys, sell other people's keys, or even find ways to get the Frenrug agent to use their funds in more creative ways.

As more and more applications and protocols begin to introduce AI agents, humans will use them as a bridge to enter the crypto economy.今天的 AI 智能体看起来还像是“玩具”，但未来它们将全面提升用户的日常体验，成为区块链协议中的关键利益相关者，甚至在智能体之间形成完整的经济生态系统。

AI agents are still in their early stages of development, but as core participants in the on-chain economy, they are just beginning to demonstrate their potential.

The article comes from the Internet:Variant Fund Investment Partner: AI Agent has become a "first-class citizen" of the on-chain economy

Related recommendation: Bull market tips: How to maximize profits without losing your mind?

理性管理、设定明确目标、分散投资、保持情绪纪律和长期视角。 撰文：Crypto by Shameer 翻译：白话区块链 牛市是投资者激动人心的时刻，价格呈上升趋势，潜在的收益似乎无穷无尽。然而，谨慎管理这种热情至关重要。尽管人们很容易受到高风险投资的诱惑，但牛…