AI Agents will become the next trillion-dollar market, and I firmly believe this.

Many people still do not understand what an AI Agent is. An AI Agent is an AI entity that can independently complete a series of tasks. To understand it, one must first grasp the difference between AI understanding AI Agents and the current large language models (LLMs) such as ChatGPT, Claude, Gemini, etc. (I will refer to ChatGPT for convenience).

Since 2023, ChatGPT's capabilities have seen explosive growth, not only in communication, which is indistinguishable from that of a real person, but also in writing, coding, knowledge, translation, and other expert-level skills. However, ChatGPT cannot complete a full task in most scenarios. For example, if you need ChatGPT to analyze whether BTC will rise tomorrow, you must provide it with the recent price trends of BTC and tell it which technical analysis to use.

ChatGPT is like a developed brain. But to complete a full task, it needs eyes and ears to gather information, hands and feet to operate, making it a true social workhorse. In summary, an AI Agent uses a large predictive model as the brain of a social workhorse.

There are many benefits to using LLMs as brains; for instance, they can work 24/7, and their capabilities are at an expert level. During the day, they can help a programmer write code, and at night, they can act as a companion. They possess independent thinking and judgment abilities, can handle complex changes flexibly, and provide optimal solutions. Such workhorses are what every boss desires. The advantages are evident, so it is not an exaggeration to say that every aspect of society will be reworked by AI Agents.

Many companies are developing AI Agents, and here we will discuss those related to the cryptocurrency space. We can simply categorize them into infrastructure and applications.

1. Infrastructure Projects

1.1 When a large number of AI Agents emerge, the demand for infrastructure will certainly surge. Just as the data of public chains skyrocketed when a large number of dApps appeared. Among the infrastructure, platforms that help novices directly create AI Agents are the foundation of the infrastructure. The leaders in this area are Virtual and AI16Z, which integrate a large number of tools needed for AI Agents, allowing users to quickly and conveniently create an AI Agent. Their frameworks are GAME and Eliza, respectively. (In fact, this ELIZA framework is not the same as the capitalized token ELIZA, but the market may equate them.)

Virtual's economic design is very well done, and one of the most important measures is that all its AI Agents are composed of its token Virtual, creating a large demand for Virtual. AI16Z recently released a new token model and will empower it in the future.

1.2 Many infrastructure projects are also looking to achieve the functionalities of Virtual and AI16Z. Some are in development, such as Zerobro, while others are already operational, such as Tophat and Vvaifu. I believe the functionalities can definitely be realized, but to surpass them, strong industry resources and powerful execution capabilities are needed in addition to technology.

When the flood overflows, it will certainly spill over to this part.

1.3 Improving or optimizing the infrastructure for AI Agents.

This part of the infrastructure will definitely increase, and although its market value may not reach that of Virtual, they may start with a lower market value, so the probability of a significant increase is not small. Moreover, since this infrastructure requires GitHub open-source, it is very easy to judge whether it has real substance, unlike those coins that rely on packaging to deceive people. Current typical representatives include:

Arc (RIG)

ARC is an AI engine "toolbox" using Rust language, providing backend support services such as AI invocation, performance optimization, data storage, and exception handling. Its focus is on how to "make good use of it" in the collaboration process between AI LLM models and AI Agent systems.

The FXN protocol is a universal AI Agent communication and resource-sharing protocol designed to address the challenges brought by the surge of AI Agents and frameworks. It provides a mechanism for agents to discover, communicate, and share resources across different frameworks.

Swarms aims to be the authoritative and most reliable multi-agent LLM framework, providing developers with tools to easily automate business operations. It offers a wealth of cluster architectures, seamless third-party integration, and unparalleled ease of use.

2. Functional AI Agents

Functional AI Agents can be considered as dApps; they have practical functions. While infrastructure requires long-term technical development, functional AI Agents are different. If you have an idea, it can be easily created with the help of infrastructure. One can imagine that during a surge of interest, a large number of such agents will emerge, with many likely to fail and many opportunities for wealth.

We cannot predict what kinds of AI Agents will emerge in the future, but here are a few examples:

Axibot, an AI Agent for cryptocurrency analysis, has already reached a market value of billions. It can be said to be the hottest KOL right now.

GriffAIn is a high-end customization service for your entire home; it can meet any of your needs as long as you describe them. GriffAIn connects the demand side with the Solana ecosystem, and as long as it exists within the existing technical framework of Solana, it can meet basic requirements, whether it's targeting certain tokens on Pumpfun, creating tokens, buying a bottle of whiskey with Solana, or launching a Blink NFT on Twitter for fans to mint.

3. Some Agent Introductions

Virtual

AI16Z & Eliza

Eliza is a powerful multi-agent simulation framework designed to create, deploy, and manage autonomous AI agents. Built with TypeScript, it provides a flexible and scalable platform for developing intelligent agents that can interact across multiple platforms while maintaining consistent personalities and knowledge.

• Multi-agent architecture: Deploy and manage multiple unique AI personalities simultaneously.
• Create different agents using the characterfile framework.
• Memory management: Advanced RAG (Retrieval-Augmented Generation) system for long-term memory and context awareness.
• Platform integration: Seamless connections with Discord, Twitter, and other platforms.

Swarms

Swarms aims to be the authoritative and most reliable multi-agent LLM framework, providing developers with tools to easily automate business operations. It offers a wealth of cluster architectures, seamless third-party integration, and unparalleled ease of use. With Swarms, developers can orchestrate intelligent, scalable agent ecosystems to automate complex business processes.

GitHub has 2.1K stars.

Why is multi-agent collaboration needed? Official documentation

https://docs.swarms.world/en/latest/swarms/concept/why/

ARC

Arc is a protocol for GitHub's RIG, aimed at building scalable, modular, and ergonomic LLM-driven applications.

https://github.com/0xPlaygrounds/rig

GitHub has 1.5K stars, 122 forks.

ARC is an AI engine "toolbox" that provides AI invocation, performance optimization, data storage, exception handling, and other backend support services. Its focus is on how to "make good use of it" in the collaboration process between AI LLM models and AI Agent systems.

Arc has chosen to develop an innovative framework based on the Rust language, with advantages including:

High security: Rust's type system can prevent bugs early at compile time, reducing the risk of runtime errors. Additionally, Rust's memory management mechanism ensures no memory leaks and avoids data races.

High performance efficiency: Arc leverages Rust's zero-cost abstractions along with efficient pipelines to significantly improve runtime efficiency and reduce costs. At the same time, using the Tokio runtime, Arc can efficiently handle parallel processing, enhancing overall agent performance.

Strong flexibility: Arc allows developers to add new functional modules through traits, maintaining the framework's flexibility and scalability, and can operate across multiple platforms.

Arc is very suitable for developing high-performance AI Agents, excelling in games, robotics, automated workflows, or real-time simulations. Its architecture allows seamless scaling from local development environments to enterprise-level systems, laying the groundwork for institutional adoption.

FXN

https://x.com/joinFXN

The FXN protocol is a universal AI Agent communication and resource-sharing protocol designed to address the challenges brought by the surge of AI Agents and frameworks. It provides a mechanism for agents to discover, communicate, and share resources across different frameworks, facilitating the next generation of collaborative AI systems.

The core goals of FXN include:

• Breaking down agent silos: Traditional AI systems operate as isolated entities, limited by their built-in capabilities and directly owned resources. FXN aims to break down these barriers and promote collaboration among agents.
• Secure and efficient resource sharing: FXN is committed to providing agents with a secure, decentralized, and fair dynamic market for all participants to exchange digital resources.
• Cross-framework compatibility: Considering the multi-framework future of the AI field, FXN ensures that agents can collaborate, regardless of their underlying implementations or sources.

Griffan

On November 1, Tony first released GriffAIn at the Hacking for Agentic Finance hackathon, expressing the vision of turning your ideas into operational agent engines, gaining significant attention from the Solana ecosystem, including support or further collaboration from Toly, vvAIfu, Jupiter, and Dialect to improve the product.

GriffAIn is a high-end customization service for your entire home; it can meet any of your needs as long as you describe them. GriffAIn connects the demand side with the Solana ecosystem, and as long as it exists within the existing technical framework of Solana, it can meet basic requirements, whether it's targeting certain tokens on Pumpfun, creating tokens, buying a bottle of whiskey with Solana, or launching a Blink NFT on Twitter for fans to mint.

Currently, it is only available for users holding Sage phones, and users can activate GriffAIn by minting GriffAIn's Saga Genesis Token. GriffAIn also supports mobile devices.

The above content is sourced from the article: https://foresightnews.pro/article/detail/73963

4. Trading Strategies

In the upcoming bull market, it is highly probable that AI Agent projects will emerge with market values in the billions.

For large funds, the best choices are Virtual and AI16Z, which have become leaders with ecological barriers. The only issue is the entry point. While I believe its market value potential is definitely more than this, and a market value of billions is likely to come from this. However, both of these coins have been rising recently, so this needs to be determined based on each person's trading strategy. You can buy in batches, wait for adjustments, or buy now and withstand possible adjustments.

For the second tier of platforms, I feel there is no need to spend time and energy. The coin prices follow Virtual and AI16Z.

There are actually not many other AI Agent infrastructures because this requires technology and time to develop. I believe it will reach at least a scale of hundreds of millions. After that, projects may emerge because the infrastructure for AI Agents is not only in blockchain; the entire society is developing. Some well-developed projects can issue a coin, and developers can also obtain a large amount of development funding.

If you have the time and ability, you can invest in application-based AI Agents. In the next phase, there will definitely be a large number of them emerging, and there will be a lot of junk, but there will also be gold. The probability of finding gold will be very high.

Each function can produce a leading AI Agent; with a leader, there will be second and third leaders, along with a large number of short-term speculative imitations.

I will keep an eye on new potential AI Agents, so please follow me~

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。