A better way to give feedback on LLM output. And why nobody built it until now.Continue reading on Medium »