High. Everybody, myself included, had this idea . There's also been (at least 2) posts in the past about similar products. I stopped from testing these out as they all "cheat", or at least in my opinion they cheat, by creating embeddings and vector search before hitting up GPT with the context/answer.
Why do you consider this cheating? The context window is currently too small for it to work in any other manner.
You could ask Claude+ 100k to give you the text context(s) containing anything pertinent then have GPT4 prompted with that.
With an emphasis on the word “could”! Very few people have access to Anthropic’s services at this point!
Embeddings still seem like a more economical approach.
Very high, there are a lot of these. It's not too far from a hello world project.