In the December 6 article (From DeSci to AI Agents: Potential Track Interpretation),
I have listed a few conditions that I believe should be met for this round of market trends to become potential tracks:
- It must have a strong ecological demonstration effect, to the extent that after developing to a certain stage, other ecosystems will eagerly join the battle.
- It must have benchmark projects, to the extent that every project attempting to establish this ecosystem will say, we want to become the XXX coin of the XXX ecosystem.
- Of course, it also needs to have a strong wealth effect, and this wealth effect cannot be just one or two isolated cases, but rather a batch of projects that allow participants to earn considerable returns, forming a group effect. Additionally, this wealth effect should be sustainable for a period of time, rather than buying yesterday, making money today, and going to zero tomorrow.
- Finally, it is very likely to be a new track. Because this new track has not been tried before, even a slightly new approach can spark infinite imagination.
It seems that the AI agent track is increasingly meeting these conditions.
Just last Friday, the largest AI agent financing platform in the Base ecosystem, Virtuals (https://app.virtuals.io/), had a circulating market cap/full market cap of over 2 billion dollars.
This market cap was achieved without the involvement of Xuan, OK, or even their backend platform Coinbase.
In this context, this market cap has already reached the top among AI agent tokens, significantly ahead of other AI agent tokens (including meme coins initiated by AI agents).
Measured by market capitalization, it has already become a benchmark project in the AI agent track.
At the same time, alongside the start of the Virtuals platform, a series of AI agent tokens funded or initiated by Virtuals project parties have also experienced a 'rise in fortunes', clearly forming a clustering effect.
Under this effect, other second-layer expansions on Ethereum, such as zkSync and Arbitrum, have also emerged 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 clearly expanded rapidly from the initial Solana and Base to other ecosystems.
During this time, when I checked information about AI agents on Twitter, I saw several very insightful pieces of information, and instinctively looked to see which KOL shared it, only to find that they all came from AI agent AIXBT.
At least for AIXBT, an AI agent issued by the Virtuals platform, I believe it is no longer a meme coin, but has genuinely created actual value and utility.
More importantly, this value is sustainable and developmental, and it certainly doesn't rely solely on intermittent emotional drives like meme coins do.
I believe 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 several hundred AI agents ready to launch on the platform.
This number should be even higher now.
Even if many of these projects eventually turn out to be bubbles, as long as a few more AI agents like AIXBT emerge that genuinely create 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 relatively 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 classification of each subfield within this ecosystem. For investors interested in this ecosystem, paying attention to the projects mentioned is one aspect, but more importantly, it’s about finding the subfields that align with their preferences and expertise. Seeking opportunities within these subfields.
For investors, when we discover a potential new track, the highest returns come from directly targeting the specific areas we are most skilled at and understand, looking for small-cap and newly listed projects within them. However, this comes with high risks and high barriers.
If you're unsure which specific projects are reliable, a simpler approach is to find trustworthy platforms. Because the criteria for judging platforms are much simpler and easier to grasp. Compared to specific projects, platform tokens may have smaller returns, but they also come with lower risks.
If a few more projects that can provide real services and value emerge in the AI agent track, I believe it could really become the main driver of the next market trend, with projects exceeding a market cap of ten billion dollars.
In terms of the AI agent track, I believe it is still early.