Hacker News new | ask | show | jobs
by Gormo 558 days ago
Maybe just having a confidence weight assigned to each sentence the LLM generates, reflected in tooltips or text coloring, would be a big improvement.