How to view the current "chaos" in the AI Agent market?

Author: Haotian

In my last article, I shared the logical relationships between AI DePIN, AI MEME, AI Agent, and other AI narratives. However, when it comes to exploring the potential Alpha of AI, questions arise: Should we follow the AI MEME that may come from conspiracy groups with strong wealth effects but frequent changes? Should we choose protocols that are still in their early stages but are building infrastructure? Or should we select some cool standalone AIs with specific application scenarios? To be honest, I have been confused for a long time, and here are some of my shallow insights:

1) The big trend is undeniable. AI Agent is neither an "asset" nor an "application," but a new paradigm of large-scale blockchain ecosystem "narrative."

In other words, the arrival of AI Agent is akin to the DeFi Summer that blockchain experienced in 2020. It will generate a massive new "bubble" narrative cycle around the issuance, custody, and trading of AI Agent tokens; infrastructure construction, frameworks, standards, communication, stacking, and SaaS packaging; application development, operation, aggregation, iteration, and implementation, etc.

DeFi drove the brilliance of the last bull market and sparked tremendous market vitality around the three pillars of stablecoins + lending + DeX derivatives. According to incomplete statistics, the total locked value of multi-chain DeFi is $134.6 billion, the circulating supply of stablecoins has exceeded $100 billion, DEX monthly trading volume has reached $372.3 billion, and lending volume is nearly $20 billion, among other impressive and continuously refreshing data. The market size and ecological importance of the entire DeFi industry are reflected in these astonishing figures, not to mention the various chains and applications generated by DeFi, such as MCAP and FDV.

In my view, AI Agent will create a new narrative miracle that may even surpass DeFi narratives, with an imaginative space of $200 billion or $500 billion. Stop always thinking about a new DeFi Summer 2.0; please keep up with the current AI Agent Summer narrative wave and don’t fall behind.

2) Based on this logic, let’s take a look at the current "chaos" in the AI Agent market:

1) The community disputes and competition among AI MEMEs like $GOAT, $FARTCOIN, and $ACT; 2) The competition over AI Agent framework standards like $VIRTUAL and $AI16Z; 3) The competition over asset issuance and circulation like $VVAIFU and $GRIFFAIN; 4) The competition among standalone AI Agent applications like $AIXBT, $AVA, and $BULLY, etc.

You will find that it is still too early. The various lending platforms, derivatives, liquidity pools, governance tokens, etc., that appeared in DeFi at least had Ethereum and other chains providing a basic execution framework and standards, while the AI Agent ecosystem does not even have dedicated infrastructure conditions. From the perspective of ecological plasticity, the industry narrative opportunities brought by AI Agent are much grander than those of DeFi.

In other words, these chaotic phenomena are not chaotic. If you are optimistic about the development of AI Agent narratives across various sectors, any asset that has reached a certain market value and has a corresponding user base will be a quality asset before the massive explosion of the AI Agent narrative. The heights of FDV that past assets like YFI, CRV, UNI, DYDX have reached, ranging from hundreds of billions, should ease your concerns.

Of course, having confidence does not mean you can act recklessly. Don’t forget that besides the "big uncle," there are also "second uncles" like pseudo-value assets, and these types of assets will always quietly invade your wallet with short-term allure but long-term toxicity.

3) Because we are indeed still in the early narrative stage, there is no reason to discourage participation in AI MEMEs that are clearly manipulated by conspiracy groups. After all, no one knows how much wealth effect an early track can unleash, and some frameworks that seem more valuable and feasible may also face the possibility of being overtaken by competitors at any time.

In summary, it is difficult to provide a clear logic for Alpha discovery. I will only share my personal investment research logic for reference:

1. For AI MEMEs, only consider "leading coins" like $GOAT that have narrative pioneer characteristics. There is no logic; if I had to say one, it would be like Bitcoin, a consensus coin, giving birth to Ethereum, a smart contract delivery coin.

An early MEME in a track, although primarily speculative, plays an important role in educating the market and cultivating user awareness, just like DOGE did for the popularization of cryptocurrency.

2. AI Agent infrastructure is very important and is underestimated in the long run. For example, early frameworks like $ELIZA, which have rich GitHub open-source libraries and a particularly wide range of applications, may be in their early stages and could be overtaken by stronger frameworks at any time, but their absolute advantages in open-source sharing and first-mover advantages serve as barriers. It is highly likely that better frameworks will build on top of it, which may affect its growth potential but will not eliminate it.

If we treat ELIZA as the EVM of the AI Agent era, there will inevitably be new frameworks like EVM++, EVM Compatible, etc., to fill the gaps.

3. The asset issuance, custody, and trading of AI Agents are currently still at the narrative level of MEME assets. We should focus on platform value and not be misled by "junk assets" on the platform.

The logic is simple: AI Agent is essentially a process of value realization based on crypto incentive mechanisms for AI scenarios. The early infrastructure, applications, protocol frameworks, etc., are still immature, so how could there be so many super applications at the level of AIXBT? What can the convenience of issuing tokens bring? A realization? In the end, it only creates a bubble trap that generates hype for asset issuance platforms like VIRTUAL and VVAIFU.

Therefore, do not overly pursue AI Agent super applications unless there is an absolute information advantage. Otherwise, by the time you see the market cap surge, the cost-effectiveness will not be high, while locking in innovative platform infrastructure is a different story.

4. There is still a vast blank space in the construction of AI Agent infrastructure. For example, early frameworks like ELIZA only solve the problems of LLM large models and AI Agent dialogue mechanisms and have been applied to Twitter, Discord, and other scenarios, but more often they are just matching web2 information. When will there be an ELIZA++ version that can connect the ELIZA framework with web3 trading scenarios based on public chain paradigms? That is worth looking forward to.

Moreover, while platforms like Virtual and VVAIFU can quickly issue assets, they still lack significantly in decentralized custody and AMM-style intent-based trading. When will the previously discussed AI Agent + TEE private key management method mature? Can asset issuance platforms extend new gameplay in trading and introduce more interesting trading methods and incentive models?

Additionally, can we have some chains that build interaction standards for AI Agents, becoming the Cosmos of the AI Agent era, providing standard interoperability capabilities? Or chains that build permanent memory modular storage for AI Agents, providing necessary DA capabilities? There are also needs for AI Agent visualization tools, execution engines, modular combination packaging services, etc.

With these technology and market-driven supporting infrastructures in place, will you still be anxious about holding a bunch of incomprehensible MEMEs?

That’s all.

As for the various layer1s, high-performance chains, cross-chain infrastructures, chain abstractions, DA combination chains, layer2s, restaking, etc., built on the smart contract paradigm, where is the way out? Besides the existing infrastructure's landing application PMF delivery issues, one breakthrough point is the new concept of "riding the AI Agent narrative."

Therefore, any chain that can jump out to provide communication protocols, DA, application development, and other service capabilities for AI Agents is worth paying attention to. Of course, among them, there will definitely be chains that purely ride the concept, but that is still better than chains that are stagnant and rely on past achievements.

In any case, please focus on the AI Agent track and welcome the arrival of this epic bull market's main upward wave.