In the past few weeks, the ecosystem of AI Agents has rapidly expanded, with more and more tokens and projects entering this space.

Projects are gradually differentiating into application-type and platform-type. Application-type projects focus on solving specific problems or providing personalized experiences, while platform-type projects provide tools for building and launching AI Agents.

The combination of "platform + marketplace" has become an increasingly popular model.

Of course, the representative projects at this stage are mainly AIXBT Virtuals, ai16z (Eliza framework), and the recently launched OK's GOAT, all of which are receiving significant attention.

Especially today, the launch of the first fully autonomous Eliza deploying the token $METAVERSE in TEE seems to have opened another big door for AI Agents.

However, AI Agents are definitely not independently issuing tokens. The high interest in AI Agents, whether it be the application-type project AIXBT or the platform-type project ai16z, is limited to the ability to independently issue tokens as a development focus. Through various low-threshold, no-code, or natural language methods, everyone can quickly create AI Agents.

Currently, it seems that projects focusing on platform/framework types in the AI Agent space have a more resilient lifespan compared to pure AI Agent applications. A good platform can spawn many AI Agent applications, and this is the simplest reason.

However, for the emerging market of AI Agents, this is quite common in the blockchain market. Whether it is ICOs, IEOs, mining machines, blockchain games, DeFi, or inscriptions [sounds familiar?], they all tend to bring significant market speculation when they first emerge, but the outcomes are often quite similar [although the aim is simply to create a profit, seizing the right timing for entry and exit can still help us achieve our profit goals].

As for what AI Agents will develop into in the future, I don't care that much.