Eugene Yan 4/30/2023

Interacting with LLMs with Minimal Chat

Read Original

The article critiques the prevalent chat-based interface for Large Language Models (LLMs) and proposes a more intuitive, click-driven UX. It details a prototype book recommendation system that uses embeddings, approximate nearest neighbors, and cached keywords to provide context, allowing users to interact with an LLM 'librarian' through minimal typing.

Interacting with LLMs with Minimal Chat

Comments

No comments yet

Be the first to share your thoughts!

Browser Extension

Get instant access to AllDevBlogs from your browser