Author: Haotian

I believe many people feel like they are picking up treasures everywhere when they open DEX, as there are hot projects with technical quality ranging from 10M to 300M everywhere. However, when opening CEX, it is filled with old narrative projects ranging from 300M to 3B, mixed feelings abound. But in fact, CEX also has a logic for treasure hunting along the trend; let me briefly talk about it:

1) This round of AI Agent narrative is rapidly evolving, from AI MEME -> standalone AI applications -> AI Agent Launchpad -> AI Agent framework standards -> AI Agent blockchainization.

Pure MEME, no more to say, has always been a logic of rapid hype; standalone AIs need to compete on experience, and in the short term, AIXBT can dominate one side; the Launchpad platform will definitely be devoured by Virtual, taking full advantage of first-mover benefits; frameworks and standards, due to the complexity of technical logic, have vast imaginative potential, but are currently still in a chaotic fighting stage; only the direction of AI Agent blockchainization is relatively 'traceable'.

2) The logic of 'blockchainization' is definitely not that a public chain jumps out and says, I can support all AI Agents running on the chain. Such public chains definitely exist, for example,

@NEARProtocol, ICP, BNBChain, all have such potential. But the problem is that at this stage, AI Agents are still in the early stage of MEME 'asset issuance', only able to run AI Agents but unable to motivate developers, which is futile.

Therefore, the first wave of 'blockchainization' narratives must come from the ai16z ecological series. For example, @focEliza.

3) Moreover, the core point of 'blockchainization' lies in building infrastructure that better accommodates and serves AI Agents for large language models (LLM), such as:

1. For AI Agents to achieve autonomous generation of private keys and management of assets, they must rely on TEE infrastructure. Therefore, projects that can provide mature TEE solutions will be potential Alphas among the old narratives in CEX. Because TEE has always been inconspicuous under the competitive pressure of traditional ZK, MPC, FHE, and other cryptographic algorithms, but suddenly becomes the core infrastructure of the new wave of AI grand narratives.

Imagine some low-market-cap, low-profile, technically solid projects that have tapped into a new trend will definitely be potential treasures among many CEX targets (the performance show logic of $PHA is just like this; if you agree with the value of TEE infrastructure to AI Agents, the potential of the TEE track must still be present).

2. For AI Agents to achieve layered optimization and matching of Memory, the goal must be to build Data Availability (DA) capabilities suitable for AI Agents. Traditional EVM public chains have extended the DA section with many exciting narratives such as Blob plug-in space and third-party DA War.

How to build DA capabilities exclusively for AI Agents will definitely be a focus of discussion: including how DA block space accurately records effective semantics of LLM context, how Character role-setting plugins interact in real-time with DA blocks, how data in multiple block spaces effectively match multimodal interactions, and the cost issues of hierarchical data storage in multiple block spaces (short-term, long-term, working memory, etc.) are all challenges that the new type of DA Builder urgently needs to tackle. This is also the reason why focEliza takes DA capabilities as the core breakthrough point.

3. For AI Agents to achieve verifiability in unimodal and trustworthy interactions in multimodal, it must rely on trustworthy verification processing at the public chain level. The TEE infrastructure addresses the generation and application of private keys for large models, but TEE alone cannot solve the problem of single-point TEE hardware being physically compromised, nor can it resolve the consensus verification issue of TEE executing programs.

In the long run, this must rely on the naturally decentralized node verification consensus and the cooperative calling environment of smart contracts provided by blockchain.

In other words, a blockchain distributed system with strong consensus will make the intelligence of AI Agents 'reliable' and guaranteed. Therefore, following this logic, efforts are currently being made to provide AI Agents with zkVM underlying frameworks, ZK Oracle solutions, ZK Bridge cross-chain solutions, 'chain abstraction' public chain application solutions, and other projects, all of which are part of the narrative trend of this round of 'blockchainization'.

Above

When solutions providing 'blockchainization' capabilities for AI Agents, such as TEE + DA + Oracle + zkVM + chain abstraction, etc., mature, at that point, the decentralized distributed computing power, decentralized fine-tuning inference environment, decentralized data sources, decentralized IP communication and incentives, and other necessary functions provided by AI platforms will truly become 'essential'.

To some extent, based on this effort towards the blockchainization of AI Agents, major AI platform projects such as io, Aethir, Vana, SaharaAI, etc., will have a usable place.

Clarifying this logic, understanding how to extract the true essence during the chaotic epochs above DEX and how to pick treasures in silence within CEX will naturally lead to clear judgments.