Proof of Training (PoT): Harnessing Crypto Mining Power for Distributed AI Training
This addresses the problem of inefficient resource use in crypto mining for AI researchers and practitioners, but it appears incremental as it builds on existing blockchain and AI methods.
The paper tackles the challenge of integrating AI training with crypto mining by introducing the proof-of-training (PoT) protocol, which uses PBFT consensus to synchronize states and shows potential in task throughput, robustness, and security.
In the midst of the emerging trend of integrating artificial intelligence (AI) with crypto mining, we identify three major challenges that create a gap between these two fields. To bridge this gap, we introduce the proof-of-training (PoT) protocol, an approach that combines the strengths of both AI and blockchain technology. The PoT protocol utilizes the practical Byzantine fault tolerance (PBFT) consensus mechanism to synchronize global states. To evaluate the performance of the protocol design, we present an implementation of a decentralized training network (DTN) that adopts the PoT protocol. Our results indicate that the protocol exhibits considerable potential in terms of task throughput, system robustness, and network security.