interesting read, I think a valid point for accelerating code review of LLMs is funnily enough to make sure you have multiple LLMs reviewing each other in a sor

interesting read, I think a valid point for accelerating code review of LLMs is funnily enough to make sure you have multiple LLMs reviewing each other in a sort of pipeline. Some research already shows this increasing accuracy of AI generated code, maybe could evolve into a real review process that works?