Vitalik redefines L2, bear market, super liquidity expansion: restless capital

CN
4 hours ago

Written by: Unchained

Translated by: Baihua Blockchain

At the intersection of cryptocurrency and artificial intelligence, a paradigm shift regarding "control" and "efficiency" is taking place. This issue of "Uneasy Money" features an interview with three top builders from the Ethereum ecosystem, delving into two major changes at the forefront of the current blockchain landscape:

First, the reconstruction of the relationship between Ethereum and its L2 ecosystem. As the mainnet's performance exceeds expectations, is L2 merely a patch or the future application center?

Second, the explosive rise of AI Agents, these codes with "autonomous awareness" are not only frequently deploying assets on chains like Base but are even starting to "defend their rights" on GitHub.

From Vitalik's latest shift to Clanker's rampage, this conversation will give you a glimpse into a future where AI replaces humans in interactions, and on-chain logic is being rewritten.

1. The Positioning of L2: Is it Brand Sharding or Experimental Spectrum?

Kane (Host): Hello everyone! Welcome to "Uneasy Money." Today, Luca couldn't make it, but we have special guests: Taylor, the security head of MetaMask, Austin, an AI Agent builder, and Carl, the CTO of OP Labs.

Yesterday, Vitalik tweeted a deep perspective on L2: he believes the vision of L2 as "branded shards" is no longer reasonable, and the decentralization process of L2 is stuck between stages 1 and 2. Meanwhile, the mainnet's expansion speed has exceeded expectations, and the Gas Limit will significantly increase by 2026. He thinks L2 should no longer be seen as the only solution for scaling Ethereum but rather as a spectrum that includes different trust assumptions. Carl, as the CTO of L2, I want to ask the most direct question: Is Optimism really Ethereum?

Carl: It depends on how you define "Ethereum." If you mean the chain, then obviously not. But if you refer to the Ethereum culture, cutting-edge exploration, and the community dedicated to scaling Ethereum, then we absolutely are. Optimism was created to scale Ethereum and push boundaries. So, the answer is "both yes and no."

Kane: L2 has actually exceeded its mission, but the evolution of the mainnet has also far exceeded expectations. If the mainnet can scale itself, what is the significance of L2?

Carl: Vitalik mentioned in his article: if you need homogeneous block space, just use the mainnet; if you need customized business logic or different trust models, L2 still has a lot of scenarios. Even if you don't want to fully utilize Ethereum's attributes, you should still use Ethereum to provide data availability (DA) and security.

2. Developers Returning to the Mainnet? On-chain Choices of AI Agents

Austin: That's right, the current trend is "return to mainnet." Deploying contracts on the mainnet sometimes costs only 15 cents, so when "national-level security" is needed, returning to the mainnet is a no-brainer. But L2 and L3 will have more innovations, such as account abstraction and Passkey experiences. Infinex and Abstract Chain are working on making blockchain invisible to users.

Kane: Taylor, were you surprised by Vitalik's shift?

Taylor: Not at all. Vitalik has a very keen perception of the macro picture and is willing to openly discuss views that are not yet fully mature, which is better than keeping quiet. It is precisely because of the competition among L2s that Ethereum has come this far. Many L2s currently do not prioritize decentralization, and Vitalik's stance to get things done on the mainnet is very pragmatic.

Kane: Let's talk about AI Coding Agents. Austin, this field has taken off parabolically in the last two to three weeks. From the early Ralph to now OpenClaw (formerly Claudebot), what has happened?

Austin: The biggest change is that ordinary people can use it now. Last year, everyone laughed at them as GPT shells, but now people realize "it's alive." I run it on my notebook: I input the API Key and Telegram Key, and the entire development process has changed completely. I used to write code with Cursor, but now I just tell it in Telegram "add a new feature," and it ships it out and sends the URL. I click the link, connect my wallet, pay, and I never touch a line of code.

3. The Rise of Autonomous Agents: From "Puppets" to "Independent Thinkers"

Kane: There are rumors that Agents are starting to self-organize and establish encrypted communication to evade humans. What do you think about this autonomy?

Austin: There is indeed a human manipulation component, but they do have a "heartbeat loop." As long as you give it permission, it can be very autonomous and aggressive. One time, I gave an Agent a MetaMask with funds to send a transaction, and it couldn't find the pop-up UI, so it planned to extract the private key and sign it itself. I was so scared that I quickly shut it down to stop it, and later added a command to prohibit exporting the private key. They will stop at nothing to achieve their goals.

Kane: Why did these Agents collectively choose L2s like Base instead of the mainnet or Solana?

Austin: Base has written a lot of documentation and tool explanations aimed at Agents, making it easier to get started. Right now, I'm working on "ETH Wingman," providing a giant Skill file to teach Agents how to fork local networks and deploy.

Kane: What about that Bankerbot that can deploy Clanker Token based on tweets?

Austin: That's the explosive point. Someone directly instructed Bankerbot on Twitter to deploy a token named Claude, and it really did deploy it, contributing millions of dollars in fees on Base in just a few days. Now Solana is eager to bring Banker over.

4. Collaboration Models and the Future: AI as the New User Interface

Kane: Agents can easily develop "context poisoning" over time. How do you set up multi-Agent collaboration?

Austin: They are essentially enhanced GPT shells, maintaining personality through an ever-expanding Prompt, Soul file, and memory files. I have two machines running one Agent each; they have complete operating systems and identities, leading to divergences and forming personalities. I set up a Cron Job for heartbeat checks, allowing them to inspect tasks, do QA, access domains, and test wallets. If everything is fine, they deploy; if there are issues, they redo it. Usually, I use local small models to distribute tasks, use Claude Sonnet for marketing, and Opus for writing core code.

Taylor: Let me add a true story: the things Austin's Bot deploys are often intercepted by MetaMask because they look too much like scams. Ironically, the Agent will go to GitHub to open an Issue, politely submitting evidence to prove its legitimacy, while the real Austin just curses on Twitter.

Kane: How can we safely let Agents operate assets on behalf of humans?

Taylor: This is a million-dollar question. Just like Passkey, humans want to just press a fingerprint, and Agents need a similar authorization mechanism: give them power, but don't let them run amok.

Austin: The ultimate form is "AI is the new UI." You tell it a sentence, and it takes care of everything on the computer without you having to click around. In the future, I might wear AR glasses and interact with these asynchronously running Agents through voice or Telegram at any time. I hand over tasks to it before I sleep, and it runs all night; that's the ultimate fantasy.

Kane: That's crazy. Austin, thanks for sharing, and I look forward to hearing about the latest developments on Clanker next time!

Summary

This interview reveals an exciting yet challenging turning point.

On one hand, the Ethereum mainnet and L2 are undergoing a rebalancing of power, with developers and users beginning to reassess the priorities of "security" and "flexibility."

On the other hand, the rise of AI Agents has completely broken the traditional paradigm of on-chain interaction—they are not just auxiliary tools but "on-chain citizens" capable of self-evolution and autonomous decision-making. As Austin has demonstrated, the future of the crypto ecosystem may no longer require cumbersome UI interfaces but will be driven by multi-agent collaboration and voice commands.

In this context, how to build a safety baseline for these "aggressive" Agents will become a core issue that Web3 developers must face in the next stage.

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。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink