Comment by nateroling

I tend to agree. After seeing http://chatjimmy.ai, I think multi-agent systems are mostly just solving for LLMs being slow currently.