Will LLMs Kill Search? The Future of Information Retrieval

Overview

Explore the future of search in this insightful talk featuring Nils Reimers, director of ML at Cohere. Delve into topics such as the drawbacks of fine-tuning, the potential of retrieval augmented generation (RAG), and the limitations of lexical search. Learn about multilingual embed models, searching complex data, and Cohere-Rerank. Gain valuable insights through a comprehensive Q&A session, addressing the question: Will large language models (LLMs) ultimately replace traditional search methods?

Syllabus

Intro
Why is fine-tuning a bad idea
RAG - retrieval augmented generation
Issues with lexical search
Multilingual Embed Model
Search On More Complex Data
Cohere-Rerank
Conclusion
Q&A section