Skip to content

Comment on GPT-4o mini: advancing cost-efficient intelligenceparent

Comments

Wouldn't a lawyer wanting to run queries against the entire tax code have a model that was fine-tuned on all of that data though? I mean, vs. doing RAG by sending the entire tax code on each request.

Unclear, but fine-tuning has many problems not faced by RAG:

- More prone to hallucinations

- Worse at citing sources for people to double check outputs

- Can't be updated without retraining

- Can't impose knowledge access controls for different users

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.