Should you use RAG, or just leverage the long context LLMs have?
This is a longer video, but there was a bit too much "context" to include! Should you still use RAG if LLMs can handle really long contexts? I cover two papers here, one from DeepMind and one from Nvidia. My TLDR is, you still need RAG, but longer context LLMs will eat away at the use cases where you might need RAG. #technews #ainews #ai #llm #genai #generativeai #productmanager