From the course: Hands-On Generative AI with Multi-Agent LangChain: Building Real-World Applications

Unlock the full course today

Join today to access over 24,400 courses taught by industry experts.

Authoritarian vs. decentralized speaker selection

Authoritarian vs. decentralized speaker selection

From the course: Hands-On Generative AI with Multi-Agent LangChain: Building Real-World Applications

Start my 1-month free trial Buy for my team

Authoritarian vs. decentralized speaker selection

“

In multi-agent systems, we encounter problems that are quite similar to what we encountered in real life. There may be a situation where multiple agents want to speak at the same time. How do we decide who gets to speak first in this situation? Well, we have speaker selection algorithms that solve this problem. So let's take a look at two speaker selection methods, mainly decentralized and authoritarian, and see how we can use them in LangChain to decide who gets to speak when. First, let's take a look at the authoritarian method. In this method, we have the conversation history, which can act as the context which is taken in by a director or chairperson. And the director then directs the flow, deciding who speaks and when. This approach ensures that every agent contributes in a manner that aligns with the predetermined objectives of the conversation. Keep in mind, we have three agents here. One of them is the director agent who directs the flow of the conversation. Next, let's shift…

Contents