Automatic Generation of Distributed Algorithms with Generative AI

Fault-tolerant distributed algorithms such as Reliable Broadcast, Causal Broadcast, Total Order Broadcast, and Consensus, are at the core of many modern distributed systems. However, the development of distributed algorithms by humans is a laborious and complex process. This work presents a novel approach to generating distributed algorithms using Generative Artificial Intelligence that allows for automating the process of generating such algorithms. The paper also summarizes our initial results on using the approach to generate Reliable Broadcast algorithms.