Three: tireless trial and error. Cannot agree more. I figured this probably be the biggest advantage of LLM considering for other variables humans hold the same-level competency.