Im not full read up on RAG pipelines, but has anyone ever tried to make the database a neural net itself? I.e get rid of any sort of traditional databases, and then you basically just have some sort of autoencoder?
In some sense there's probably a database compression scheme that does something similar. Usually people care too much about fidelity
ActorNightly · · focus · HN ↗
tyromaniac · · focus · HN ↗