Recently, Vitalik once again made a series of statements regarding the role that Ethereum should play in the AI era, among which the core points I am particularly concerned about are:

- AI should not blindly accelerate

- The value of Ethereum lies in providing verifiable and decentralized directional choices for AI

- Making better market and governance forms truly feasible

This article generally inherits Vitalik's consistent thoughts and logic, unwaveringly insisting on the ideal of using technology-based decentralization to impose "constraints" on AI to prevent it from getting out of control in the AI era.

I fully recognize and agree with these points.

However, during this period, due to a series of new events and situations, my expectations and hopes for the future development of Ethereum have risen significantly.

Two weeks ago, due to the development of AI, Wall Street underwent a nearly irrational "bloodbath" on software company stocks. The prevalent opinion online is:

In the AI era, software will become outdated, software companies will be eliminated, and AI will reshape everything about software.

This perspective seems a bit exaggerated, but this time, I believe there is a certain reason behind Wall Street's viewpoint, which is worthy of serious consideration.

Everything we use in modern society in terms of software is built for humans—the user experience, interface design, usage habits, invocation methods, and all other factors ultimately face humans.

However, now we are already seeing a new trend: after the emergence of Clawdbot/Molt, do humans still need to directly interact with software? Currently, it seems necessary, but this demand will certainly decrease exponentially in the upcoming period, and eventually, in future society, the vast majority of "software" users will be AI Agents rather than humans.

Just as the interaction between AI Agents does not necessarily have to be in the language used by humans, the software invoked by AI Agents and the way they use software can also be entirely different from that of humans; they can have more efficient and disruptive methods.

This suggests that the vast majority of software we see today is highly likely to be reconstructed, disrupted, or eliminated in the future.

This is what is happening and about to happen in the software field.

Such occurrences are also undergoing fission and catalysis in other fields.

A couple of days ago, in an article comparing Virtual and Molt, I sensed that Virtual faces similar issues; as the article states:

"We examine this world from a god's perspective, knowing that countries will certainly emerge in the future, so we first build the structure and mechanisms of a country according to our ideas, and then we migrate to this country or attract people to enter this country."

The biggest potential problem with this approach is: does the architecture and mechanisms we construct for this country truly meet the needs of AI Agents? Are they truly the most suitable for the flourishing of AI Agents?

Clearly, this approach is not as natural, healthy, and effective as the organic growth and evolution approach of Molt.

It is precisely because of the continuous accumulation of these thoughts that when I saw Vitalik's article, I thought of ERC-8004, X402, and a series of verification mechanisms and payment mechanisms designed by us humans for AI Agents...

Then a question arose:

These protocols and mechanisms are all conceived by us humans, but are they really suitable for AI Agents? Are they truly the most appropriate form for AI Agents?

Could it be that some seemingly insignificant protocols and mechanisms emerging in a corner or community we completely overlook might ultimately be the most favored, most native, and most suitable for AI Agents?

I currently do not have an answer.

But I can't shake the feeling that this exploration process may not be very smooth.

Duan Yongping especially emphasizes two sentences in his investment Q&A:

Do the right thing, do things right.

Following this logic, I firmly believe that Vitalik and the Ethereum he leads are currently doing the right things, but are the things being done right?

This is something I am increasingly uncertain about, as the changes brought about by the development of AI are indeed too rapid, disruptive, and unpredictable.

However, in any field, the correct exploration and practice follow this process: first ensure we are doing the right things—applying cryptographic technology to impose boundaries and constraints on AI development, and then explore how to do those things right in practice.

If Ethereum is to truly realize Vitalik's ideals and goals in the AI era, it will likely need more frequent iterations and closer attention and tracking of AI in the future.

This is a larger, more daunting goal and task.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。