Proof of Useful Intelligence Links Consensus Security to Real-World AI Utility
Proof of Useful Intelligence (PoUI) is a hybrid consensus model blending useful AI computation with staking to secure the network while generating tangible value.
Asynchronous Consensus Achieved Using Only Hash Functions and Simple Primitives
A novel asynchronous consensus protocol leverages a binding Index Cover Gather primitive and simple hash functions to achieve optimal fault tolerance and constant rounds, eliminating complex public-key cryptography.
Dual-Threshold BFT Consensus Achieves Sub-Second Planetary Finality
Mercury protocol uses adaptive dual resilience thresholds and BFT forensics to slash wide-area consensus latency to under 0.4 seconds, enabling global-scale SMR.
Commitment-Reveal Decouples Ordering from Value to Ensure Fairness
A novel two-phase commitment-reveal protocol decouples transaction ordering from content knowledge, eliminating block producer MEV extraction and ensuring provably fair sequencing.
Recursive Zero-Knowledge Proofs Unlock Verifiable Private Computation Scaling
zkAdHoc introduces recursive proof aggregation to generate a constant-size proof for arbitrarily complex computation, enabling scalable on-chain verification.
Sublinear Zero-Knowledge Proofs Democratize Verifiable Computation Scaling
A novel space-efficient tree algorithm reduces ZKP memory requirements from linear to square-root, unlocking verifiable computation on resource-constrained devices globally.
Distributed zkVM Architecture Slashes Verification Costs and Latency
A modular, distributed zkVM architecture dramatically cuts hardware costs and latency, making real-time zero-knowledge verification economically feasible for all validators.
New Lower Bounds Define Communication Limits for Dishonest-Majority Broadcast Protocols
This research establishes fundamental communication lower bounds for randomized Byzantine broadcast in dishonest-majority networks, framing the ultimate scalability limits.
Differential Privacy Ensures Transaction Ordering Fairness in State Replication
By mapping the "equal opportunity" fairness problem to Differential Privacy, this research unlocks a new class of provably fair, bias-resistant transaction ordering mechanisms.
