Search

This proposal involves a possible path forward towards a decentralized system offering a degree-of-belief consensus on a non-trivial subset of all natural language statements, including those about reality.

Its possible uses include:

• Decentralized, formally verified scientific research aggregation.
• Automated literature review and meta-analysis.
• Proof- and incentive-assisted debate.
• Trustless fact-checking.
• Decentralized governance in arbitrary contexts.

The system would consist of three layers that would be designed and built sequentially, each extending the system's capabilities upon implementation:

• Layer/phase 1: optimistic decentralized formal proof verification with reverse dependency tracking. Introduces all statements verifiable in the proof system.
• Layer/phase 2: formal definition and implementation of "extractors" and "injectors" mapping in and out of proofs via non-L1 "soft" axioms. Increases the scope to real-world statements not depending on real-worl