The infinite possibilities and challenges brought by the AI agent track.

CN
1 year ago

In the article "From DeSci to AI Agents: Interpreting Potential Tracks" published on December 6, I listed several conditions that I believe are necessary for a potential track in this round of market:

  • It must have a strong ecological demonstration effect, to the extent that after reaching a certain stage, other ecosystems eagerly join the fray.

  • It must have benchmark projects, to the point where every project attempting to establish this ecosystem will say, "We want to become the XXX coin of the XXX ecosystem."

  • Of course, it must also have a strong wealth effect, and this wealth effect cannot be limited to just one or two isolated cases; rather, there should be a batch of projects that allow participants to earn considerable returns, forming a collective effect. Additionally, this wealth effect should be sustainable over a period of time; it cannot be the case that one buys in yesterday, makes money today, and then it goes to zero tomorrow.

  • Finally, it is likely to be a new track. Because new tracks have not been tried before, even a slight new play can trigger infinite imagination.

Now it seems that the AI agent track increasingly meets these conditions.

Just last Friday, the largest AI agent financing platform in the Base ecosystem, Virtuals (https://app.virtuals.io/), saw its token VIRTUAL's circulating market cap/full market cap exceed $2 billion.

This market cap was achieved without being listed on Xuan, OK, or even its parent platform Coinbase.

In this context, this market cap is already at the top among AI agent tokens and significantly ahead of other AI agent tokens (including meme coins initiated by AI agents).

In terms of market cap, it has already become a benchmark project in the AI agent track.

At the same time, a series of AI agent tokens funded or initiated by the Virtuals project have also seen a "rise" during this period, clearly forming a clustering effect.

Under this effect, other second-layer expansions of Ethereum, such as zkSync and Arbitrum, have also emerged with similar projects, with project parties clearly stating on their Twitter that they want to develop their own AI agent platforms.

The AI agent track has evidently spread rapidly from the initial Solana and Base to other ecosystems.

During this time, while checking information about AI agents on Twitter, I have seen several very enlightening pieces of information, and my first instinct was to check which KOL shared it, only to find that they all came from the AI agent AIXBT.

At least for AIXBT, which is launched by the Virtuals platform, I believe it is no longer a meme coin but has truly generated actual value and utility.

More importantly, this value is sustainable and developmental, unlike meme coins that purely rely on intermittent emotional drives.

I believe that cases like AIXBT are not isolated.

In the article on December 2, I shared some viewpoints expressed by the co-founder of the Virtuals platform during an interview. In that interview, he also mentioned that there are hundreds of AI agents poised to launch on the platform.

This number should be even higher now.

Even if many of these projects ultimately turn out to be bubbles, as long as a few more AI agents like AIXBT emerge that truly generate actual value and utility, the value and narrative space they bring should not be underestimated.

There is an article on Deep Tide that provides a detailed introduction to the current AI agent ecosystem. The link to the article is attached at the end.

This article lists many projects, but what I value more is its division of each subfield within this ecosystem. For investors focusing on this ecosystem, paying attention to the projects mentioned is one aspect, but more importantly, it is to find the subfields that align with their tastes and expertise. Seeking opportunities within these subfields.

For investors, when we discover a potential new track, the highest returns come from directly targeting the subfields we are most skilled at and understand, looking for small-cap, newly launched projects. However, this approach carries high risks and high thresholds.

If one is unsure about which specific projects are reliable, a simpler approach is to find trustworthy platforms. Because the judgment criteria of platforms are much simpler and easier to grasp. Compared to specific projects, while the returns on platform tokens may be smaller, the risks are also lower.

If a few more projects that can provide real services and value emerge in the AI agent track, I believe it could truly become the main driving force of the upcoming market and see projects with market caps exceeding $10 billion.

As for the AI agent track, I believe it is still in its early stages.

Reference link:

https://www.techflowpost.com/article/detail_22268.html

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。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink