Retrieval augmented generation
is way easier to search then:
a script that runs a search and then the LLM takes the output of that and reformats it into an answer.
So if people want to look into it further and research what it is, instead of taking some persons 1 sentence explanation, they can.
Ironically trying to search for that phrase would work better in a RAG then a standard key word search.



Asked it the following to test it:
Took a while, actively showed the source articles it was looking into while it was processing which were clickable. Here’s a pdf of the response which is long, and well referenced, pretty interesting IMO, but here’s the initial overview:
Which my partner, whose taken some climate classes in college, said sounds right. If anyone thinks this is wrong please feel free to call it out.