Proof of Contribution

Overview

The Proof of Contribution (PoC) mechanism in Vana is designed to validate data in a decentralized, privacy-preserving manner. It ensures that only authentic, high-quality, and unique data is added to DataDAOs, allowing contributors to benefit from the value their data generates, while protecting their privacy.

Key Elements of PoC

Authenticity: Ensures that the data is real and unaltered.
Ownership: Confirms the data belongs to the contributor.
Quality: Measures the relevance and utility of the data.
Uniqueness: Prevents redundant or copied data from being reused.

How It Works

Data Submission: Contributors submit encrypted data to a DataDAO, which is stored off-chain with metadata written on-chain.
Validation: Validators use a secure Trusted Execution Environment (TEE) to check the data for authenticity, ownership, quality, and uniqueness via Proof Of Contribution.
Attestation & Reward: Once validated, an attestation (proof) is generated and recorded on-chain. Contributors are rewarded in VANA tokens based on their data's quality.

Implementation

One recommended implementation for DataDAO Proof-of-Contribution is to run a model influence function, which measures exactly how much new information a given data point teaches the AI model.

Customization

Each DataDAO in Vana can implement its own PoC function based on the data it handles. For example:

Financial Data DataDAO: May prioritize the timeliness and accuracy of data.
Social Media Data DataDAO: May focus on engagement metrics or user activity.

By customizing the PoC function, DataDAO creators ensure that their data pools are filled with meaningful, relevant, and high-quality contributions.

Benefits

Data Quality: Guarantees that data within DataDAOs is trustworthy and useful for AI training or other purposes.
Privacy: Contributors can submit and validate data without exposing sensitive information.
Incentivization: Contributors are rewarded in VANA tokens based on the quality and uniqueness of their data.
Governance: PoC provides a fair system for contributors to earn governance rights in DataDAOs.

Zero-Knowledge Proof (ZKP) Integration

To enhance privacy, Vana uses Zero-Knowledge Proofs. This allows contributors to prove that their data meets certain criteria (e.g., more than 50 conversations in the ChatGPT DataDAO) without revealing the actual content. The cryptographic proof is generated in a secure browser environment, ensuring that sensitive data is never exposed.

For example, a user might submit a proof that their dataset contains a valid number of conversations without revealing the exact contents of those conversations. Validators process the ZKP and confirm the submission’s validity without accessing the raw data.