Vitalik redefines on-chain governance logic: prediction markets + anti-capture preference layer, DAO may usher in a new paradigm

February 2 News, Ethereum co-founder Vitalik posted on Farcaster that the future of on-chain mechanism design is not complicated. The core will follow a “two-layer structure” model: one layer is an open execution mechanism similar to prediction markets, and the other layer is a preference setting tool that is resistant to capture and non-financialized. He believes this is the key path to solving the current DAO governance problem of being controlled by capital.

Vitalik stated that the first layer should be a market system that maximizes accountability, where anyone can participate in buying and selling, and bear economic responsibility for decision outcomes through “profit from correct judgment, loss from incorrect judgment.” This mechanism, in an permissionless environment, naturally possesses auditability and accountability, making it very suitable as the foundation of a “decentralized execution agency.”

The second layer is used to express the community’s true value orientation. Vitalik emphasized that this layer cannot rely on token voting because tokens lack diversity, and anyone can gain control through acquisition. Preference expression must be decentralized, anonymous, and should adopt anti-collusion schemes like MACI to reduce manipulation and collusion risks, thereby stimulating intrinsic motivation rather than profit-seeking behavior.

He also pointed out that in some scenarios, a centralized execution team can be used to improve efficiency, but these executors must also be subject to constraints and evaluations from the non-financialized preference layer to ensure their actions align with collective values rather than serving a few interests.

Finally, Vitalik provided a clear design idea: any on-chain system should be broken down into two questions—who executes and how to execute; how preferences are expressed, and how to evaluate the executors. As concepts like “decentralized governance mechanisms,” “prediction market decision systems,” and “MACI anonymous voting models” gradually mature, this approach may push Web3 governance models into a new stage.