Comment

Comment by Isolated_Routes

This is interesting. I'm curious how you calibrate it. I've found that sometimes AI is very confident about being able to do something (or not do something), and I've had to pick its response apart a bit to see what the real answer is. Are you getting accurate results with one pass?

No comments yet

Be the first to share your thoughts.

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton