logoalt Hacker News

bigstrat2003today at 1:47 AM0 repliesview on HN

You already shouldn't be using LLMs for either of those things. Doing so is tremendously foolish with how stupid and unreliable the models are.