Following the main line of AI Agent, how to select potential tokens in CEX?

I believe that many people feel like picking up treasures everywhere when they open DEX. After all, there are popular projects with 10-300M technology and good quality everywhere. However, when they open CEX, there are old narrative projects with 300M-3B everywhere, which makes them feel mixed. Chen Plus is beyond words. But in fact, CEX also has a set of logic for picking up treasures according to the trend. Listen to my simple chat:

1. This wave of AI Agent storytelling is evolving rapidly, from AI MEME -> Single AI application -> AI Agent Launchpad -> AI Agent framework standard -> AI Agent chaining.

Needless to say, pure MEME has always been a fast-moving momentum-building logic; single-body AI needs to strive for experience, and AIXBT can dominate in the short term; Launchpad platform will definitely be deprived of the first-mover advantage by Virtual; frameworks and standards are backed by complex technical logic, imagination The space is large, but it is still in the stage of chaos; only the direction of AI Agent chaining is relatively "traceable".

2. The logic of "chainization" must not be that some public chain jumps out and says, I can support all AI Agents to run on the chain. There must be such a public chain, for example,

@NEARProtocol, ICP, BNBChain, all have such potential. But the problem is that at this stage, AI Agent is still in the initial stage of MEME-based "issuing assets". It can only run AI Agent, but it cannot mobilize the enthusiasm of developers, which is of no help.

Therefore, the first wave of "chained" narratives must come from the ai16z ecological series. Like @focEliza.

3. Moreover, the core point of "chaining" is to build infrastructure for the large language model LLM to be better compatible and serve AI Agent, such as:

1. If AI Agent wants to generate private keys and manage assets autonomously, it must rely on TEE infrastructure. Then, projects that can provide TEE mature solutions are potential Alphas in the narrative of CEX. Because TEE has always been inconspicuous under the competitive pressure of traditional ZK, MPC, FHE and other cryptographic algorithms, it has suddenly become the core infrastructure of the new wave of AI narrative.

Just imagine some projects with small market capitalization, low presence, solid technology, and new trends. They will definitely be potential treasures among many CEX targets. ($PHA’s show logic is exactly like this. If you agree with TEE’s big infrastructure for AI The value and significance of Agent, the stamina of the TEE track must still be there).

2. In order for AI Agent to achieve hierarchical optimization and matching of Memory, the goal must be to build Data Avaliability (DA) capabilities suitable for AI Agents. The traditional EVM public chain has extended Blob plug-in space, third-party DA War and many other things in the DA sector. Wonderful narrative.

How to create DA capabilities exclusive to AI Agent will definitely be the focus of the topic: including how the DA block space accurately records the effective semantics of the LLM context, how the Character role setting plug-in interacts with the DA block in real time, and the data in the multi-block space. How to effectively match multi-modal interactions, the cost of hierarchical storage of data in multi-block space (short-term, long-term, working memory, etc.), etc., are all problems that the new DA Builder needs to overcome urgently. This is also the reason why focEliza uses DA ability as the core breaking point.

3. To achieve single-modal verifiability and multi-modal trusted interaction, AI Agent must rely on public chain-level trusted verification processing. In the aspect of TEE infrastructure that solves the problem of generating large models and applying private keys, only TEE obviously cannot solve the problem of physical explosion of single-point TEE hardware, as well as the problem of consensus verification of TEE execution programs.

In the long run, these will all be achieved by relying on the naturally decentralized node verification consensus and smart contract collaborative calling environment of the blockchain.

In other words, a blockchain distributed system with strong consensus will make the AI ​​Agent’s intelligence “reliable” and guaranteed. Therefore, following this logic, we are currently working on projects such as the zkVM underlying framework, ZK Oracle oracle solution, ZK Bridge cross-chain solution, and "Chain Abstraction" public chain-level application solutions for AI Agents, all of which are also in this round. Among the narrative trends of "chaining".

above

When TEE+DA+Oracle+zkVM+ chain abstraction and a series of solutions that provide "chaining" capabilities for AI Agents mature, at this time, AI Agents will need decentralized distributed computing power and a decentralized fine-tuned reasoning environment. , decentralized data sources, decentralized IP communication and incentives, and a series of things done by AI Platform and other platforms will truly become "rigid needs."

To a certain extent, based on this step in the direction of AI Agent chaining, large AI platform projects such as io, Aethir, Vana, and SaharaAI will also have a place to play.

Clarify this logic, how to obtain the true scriptures in the era of chaos on DEX, and how to pick up treasures in silence on CEX, without further explanation, you will naturally have a clear judgment.