Why should AI framework standards move towards “chainization”?

Author: Haotian

Many people still don’t understand why I have been calling on AI framework standard project parties to move towards “chainization”? Perhaps during the first two bull-bear periods, Chain infra carried too many expectations. Finally, it has reached the era of AI Agent "application", and everyone feels intimidated by the "chain". However, in order for AI Agents to make more reliable independent decisions and collaborate with each other, they must move toward "chaining."

The currently popular frameworks such as ELIZA, ARC, and Swarms are basically still in the "conceptual stage". This stage cannot be falsified and returned to zero, nor can it be confirmed and exploded. They are basically in the infancy stage where valuation cannot be quantified. This is the first hurdle for issuing assets on Github. Only by finding the possibility of implementing the outlined framework and vision can it be unanimously recognized by the market.

Let's take a closer look at frameworks such as ELIZA, ARC, and Swarms. Whether they are for ultimate single AI Agent performance optimization or multi-AI Agent interaction and collaboration frameworks, they essentially have to sort out a set of traceable logic and rules for AGI large model API calls.

After all, the data is off-chain, the reasoning process is difficult to verify, the execution process is opaque, and the execution results are uncertain.

From a short-term perspective, TEE provides a set of off-chain Trustless low-cost, highly feasible implementation solutions that can accelerate the application of AGI to the autonomous decision-making process of AI Agents. From a longer-term perspective, a set of "on-chain consensus" is also needed to assist. to become more reliable.

For example, ELIZA wanted to build an AI Agent autonomous private key hosting solution based on its framework. It used @PhalaNetwork's TEE secure remote authentication capability to ensure that the execution code of the AI-Pool is not tampered with before calling the private key signature, but this is only TEE The first small step in the direction of AI Agent.

If we can put the complex preset execution logic into the Agent Contract and let the Validators of the Phala chain participate in the verification, a set of chains based on chain consensus constraining TEE execution details will be opened. By then, AI Agent will drive demand for TEE, and TEE will drive the forward flywheel of chain empowerment.

The logic makes sense. TEE can ensure that the private key is invisible, but how to call the private key, based on what preset rules, how to trigger the risk control emergency response, etc. In the short term, transparency can be achieved by leaving it to an open source code base, but in the longer term, doesn’t it all rely on a decentralized verification consensus to verify and confirm in real time?

Therefore, "chainization" can accelerate the AI ​​Agent framework to the practical application stage, and can also bring new incremental opportunities with Crypto infra.

The direction has been very clear. For most people, looking for and bullish, the earliest chained AI Agent framework and the earliest old chain to support AI Agent, this is the Alpha under the new trend of AI Agent.