จาก Ethereum สู่ CROPS แห่ง AI: ชุด "ตัวแปรช้า" ที่ Vitalik ย้ำอยู่เสมอนี้ คืออะไรกันแน่?

In the past period, Vitalik has repeatedly mentioned a somewhat unfamiliar word: CROPS.

The systematic appearance of this concept can be traced back to March 13th. The Ethereum Foundation board released the "EF Mandate" document, clearly stating that its primary focus will be Ethereum's censorship resistance, open-source nature, privacy, and security—that is, CROPS—and to serve user sovereignty through these, while maintaining extractability resistance and a more seamless user experience.

This statement is actually very important, especially as AI begins to enter wallets and automated execution scenarios. CROPS is no longer limited to Ethereum's value proposition; it potentially becomes a question of whether users can continue to control their digital lives in the age of AI.

I. What Exactly is CROPS?

To understand CROPS, one must first break free from a common misconception: Ethereum certainly needs to improve performance and reduce costs, but its goal isn't just to compete with other public chains on who is faster or has lower fees.

Although speed and cost are the most intuitive metrics for short-term user experience, looking at the longer term, Ethereum's stance has become increasingly clear over the past two years. What it truly aims to provide is a set of more fundamental capabilities: allowing users to hold assets, express identity, sign transactions, and participate in coordination without relying on a single platform, surrendering ultimate control, or being arbitrarily blocked by a centralized service.

This is the significance of CROPS.

In the context of the EF Mandate, CROPS mainly points to five directions, also the acronym of their keywords: Censorship Resistance, Capture Resistance (actually added later by Vitalik), Open Source, Privacy, Security:

• C - Censorship Resistance: Ensuring the immutability of transactions and smart contracts, preventing termination due to external political pressure or centralized entities;
• R - Capture Resistance: Preventing the governance, development roadmap, and key access points of Ethereum from being long-term controlled by a small group of stakeholders;
• O - Open Source / Openness: Maintaining fully open-source code and ensuring absolute permissionless access to the ecosystem;
• P - Privacy: Using cryptographic techniques to protect users' right to be unobserved on top of a transparent ledger;
• S - Security: Upholding fundamental principles to provide unbreachable ultimate settlement security;

Viewed together, these constitute a set of selection and guiding principles with a clear orientation, very much in line with Ethereum's consistent value trajectory.

For example, at the protocol layer, it means Ethereum needs to continuously improve censorship resistance, client diversity, validator decentralization, formal verification, etc.; at the application layer, wallets, RPCs, browsers, signing interfaces, and account systems need to reduce reliance on centralized entry points; at the user experience layer, security shouldn't rely solely on users understanding complex transactions themselves, but should be advanced through clearer signing displays, more verifiable interactions, and better risk warnings before operations occur.

This is also why the EF has recently promoted more specific directions around security, privacy, protocol resilience, and ecosystem public goods. For instance, the Ethereum Audit Subsidy program aims to lower the barrier for developers in the Ethereum ecosystem to obtain high-quality security audits. By extension, this isn't just about subsidizing costs; it's about pushing "security" from an expensive service affordable only by large projects, further towards more small and medium-sized developers.

In late May, Vitalik again discussed his view on the EF's future direction, emphasizing that the EF should become a smaller, more principled, and more focused organization on long-term sustainability, rather than trying to cover all needs within the ecosystem. The reason is pragmatic: the EF doesn't have unlimited resources and lacks a continuous income stream from staking or transaction fees. Therefore, it should allocate its limited resources to tasks critical for achieving CROPS values in Ethereum, tasks that other entities cannot reliably undertake.

In other words, in the current transitional historical phase for Ethereum, CROPS is not an abstract slogan where "ideology comes before reality," but rather an external definition guiding what the EF should and should not do.

II. When CROPS Meets AI: The Intersection of Two Parallel Universes

Vitalik Buterin's most recent push to bring CROPS into a broader discussion was within the context of AI.

On May 28th, Vitalik Buterin posted an update on his progress with localized AI, stating that DeepSeek V4 has released a 2-bit quantized version capable of running in approximately 90 GB of VRAM, achieving speeds of about 35 tokens/s on Apple hardware and 7 tokens/s on AMD hardware. He argued that true "CROPS AI" should support multiple hardware platforms, not just be synonymous with "decentralized AI."

He also pointed out a significant overlap between the CROPS Ethereum access layer and CROPS AI, for example, achieving paid remote LLM calls via zero-knowledge proofs, and private Ethereum RPC reads. He suggested more AI models fine-tuned for Ethereum scenarios should emerge to improve smart contract, protocol code, and ecosystem security.

This effectively places Ethereum and AI within the same problem framework.

In the past, our discussions of AI often focused on model capabilities—whether it can write code, especially if it can handle complex tasks instead of humans. But from a user security perspective, the real change brought by AI isn't just being "more capable." It's that it is changing the entry point for digital operations.

To use a common analogy, applications used to be relatively clear interfaces. We open a wallet to transfer, a DApp to trade, a browser to search, a social product to post. Each application had relatively defined boundaries. But with the emergence of AI Agents, these boundaries become increasingly blurred. Users no longer click on functions one by one; they express their intentions using natural language:

"Help me find the best cross-chain route," "Help me execute a swap," "Help me organize my assets," "Help me call a specific DeFi strategy," "Help me generate and send a transaction..."

This sounds convenient, but it raises a crucial question: When AI becomes your digital agent, what transactions exactly is it signing on your behalf, and what privacy is it potentially exposing?

If the AI runs purely on centralized cloud servers, user's asset information, transaction intents, address relationships, identity preferences, and operational habits could be concentrated in the hands of a few service providers. Especially when on-chain operations rely on opaque APIs, centralized RPCs, black-box plugins, and unverifiable inference processes, users might become more convenient but also find it harder to know what exactly they are giving up.

This is the question CROPS AI aims to answer.

An AI more aligned with CROPS isn't just powerful; it should also be censorship-resistant, open, privacy-preserving, and secure as much as possible. Ideally, it should run locally, at least minimizing reliance on centralized cloud services in sensitive scenarios, minimizing information leakage, and allowing users to understand, confirm, and retain ultimate control.

In other words, AI cannot just be a smarter black box. Especially in Web3 scenarios, AI in the future might not just help you summarize articles, write code, or do customer service, but directly participate in asset management and automated execution.

The closer it gets to user assets, the more important CROPS becomes.

This is also the reason why the CROPS Ethereum access layer and CROPS AI intersect.

III. What Web3 Increments Can Be Mined from This Intersection?

From this perspective, it's very natural that Vitalik recently pointed out the intersection between the CROPS Ethereum Access Layer and CROPS AI.

Whether it's Ethereum or AI, the core problem users face is becoming the same: How can I use AI assistance without completely handing over my privacy, identity, assets, and choice to centralized intermediaries?

• On the Ethereum side, this manifests as: How do users access on-chain data? How do they connect to RPC? How do they sign transactions? How do they confirm if a DApp interaction is safe? How do they avoid all wallet queries, balance reads, and transaction broadcasts going through a few centralized services?
• On the AI side, this manifests as: How do users invoke models? How are prompts and personal data protected from misuse? How can local models handle sensitive tasks? When remote large model capabilities are needed, how can users minimize exposing their identity and intent?

These two sets of problems seem different but are very similar at their core.

For example, when Ethereum users check balances, read transaction histories, or simulate transaction results, they often need to go through RPC services. RPC seems like just a technical interface, but it can know your IP, addresses, query habits, asset structure, and interaction paths. If this data is collected centrally, users' on-chain privacy can be gradually pieced together.

Similarly, when AI users invoke remote models, they might also expose their preferences, financial information, or even identity clues. If, in the future, users use AI to handle wallet operations, the risk amplifies further.

Therefore, the ZK-paid remote LLM calls and private Ethereum RPC reads that Vitalik mentioned are essentially attempts to solve the same problem: How can one gain service while invoking remote capabilities without exposing all their information?

This is the intersection of CROPS Ethereum and CROPS AI. On one side, a more private, more verifiable, less trust-dependent on-chain access layer. On the other side, a more open, more localized, and more secure AI execution environment. Together, they might form the new entry point for users entering the digital world in the future.

Extending this logic outwards from CROPS' core framework, the entire Web3 ecosystem (especially the wallet layer as a traffic gateway) will undoubtedly assume more roles:

When users start expressing on-chain needs in natural language, the wallet transforms from just a signing tool into a console for users' digital actions. It needs to help users judge: Can this DApp be connected? What exactly will happen with this transaction? Is this AI Agent calling unnecessary data?

From this perspective, CROPS isn't just an abstract value; it will directly influence the design direction of wallet products and drive the evolution of integrating Web3 interaction experiences and the wallet track over the next decade.

Final Thoughts

Under current market conditions, many people's attention to purely conceptual discussions might have waned.

But the colder the market, the easier it is to overlook those technological variables that aren't immediately exciting in the short term but truly determine the long-term direction.

The reason CROPS deserves attention isn't because it creates a new hype cycle, but because it reframes the long-term problems of Ethereum and AI within a single framework: When digital systems become increasingly powerful, can users still retain their control?

After all, security and privacy cannot just be afterthoughts or patches.

From this perspective, in an era where AI is accelerating its takeover of the digital world, this may be the truly positive variable that makes Ethereum worth continuing to build upon and use.

In an era where AI is accelerating its takeover of the digital world, being more understandable, more verifiable, more privacy-preserving, and more secure might be the genuine reason why Ethereum remains worth building upon and using.