In essence, the vast majority of AI agent projects that issued tokens were rushing ahead. These tokens can be seen as a quantitative representation of market and RWA (real world attention) expectations for the project's future under those conditions. However, due to the market's emotional fervor and the extreme concentration of RWA, the delivery capabilities of these projects have been greatly overlooked.
Looking back to January, the market automatically formed a valuation path: Oh, this is a framework project, so it benchmarks AI16Z; Oh, this is a defai project, so it benchmarks xxx, and so on. As long as there is a somewhat decent official website, and if there are real-name devs and a GitHub repo with good star/fork performance, it basically starts with a market cap of several million. Plus, at the beginning of the year, a bunch of star projects graduated from the Solana hackathon. I remember a teacher made a table of all the hackathon projects, opened a scraper, and monitored to directly buy CA to fully capture the profits.
But in fact, to say without any reservation, almost all AI agent tokens at that time were AI agent memes. No one cared whether their token design was reasonable, or even whether these tokens had utility. The market only knew: 'Oh, this project is good, so the coin price will rise, so I want to buy.' However, the relationship between the vast majority of AI agent tokens and the projects is unclear. Having a consumption scenario where a token is used to pay for product interface usage is already in the top 50%. Some project tokens are not even issued by the project team themselves but are launched by third-party retail investors and sent to the dev, who accepts them.
The current market is truly a major wave washing away the sand. Projects with capable teams must have accumulated enough funds in December and January. Therefore, if they are still building, they must want to create something. For these projects, I am willing to give some patience. Since February, I have been tracking some projects that are still continuously updating in my daily reports. I noticed that the vast majority of AI agent tokens during the market bubble period have already collapsed into AI agent memes. However, the surviving projects are all delivering continuous content products and are striving to resist collapse, working hard to 'fake it till they make it':
@virtuals_io is building an agent commerce protocol, continuously improving the @GAME_Virtuals framework, and is committed to building a Virtuals nation;
@griffaindotcom has already been able to make the agent into a blink, smoothly using USDC to exchange for energy calling the agent;
@arcdotfun developed the launchpad Forge and the agentic app store Ryzome from the framework;
@HeyAnonai is constantly improving token utility. Recently, the founder @danielesesta has become a master of Chinese abstraction;
@elizaOS has undergone a rebrand, building AIDAO based on its most widely adopted framework and is also developing its own launchpad;
We can see that AI is definitely a mainstream narrative for a long time in the future. The speed of iteration of technology and methods is extremely fast (you can see that content related to MCP has already blossomed in recent days). If the Web2 world has mindshare, then AI certainly occupies a high position. From a metaphysical perspective, the next ten years will transition from an official-killing era to a food-injury era. People will gradually develop from being disassembled by various standardized systems and forced to progress along prescribed routes, to individual micro-motivations having greater influence. This process undoubtedly requires the help of AI, or to put it bluntly, AI has to some extent dominated this process.
I digress, let’s come back. The AI agent meme is indeed almost dead, and along with it, some truly aspiring AI agent tokens are severely weakened. However, we can indeed see that these surviving projects are continuously building. I believe AI agents will definitely come back because, in fact, apart from RWA, the overall background hasn't changed. What we need to do is to survive until the second stage of this hype curve.
