I'm sure the output of this LLM could be improved with some hacking but why bother? Are we really at a point where we're nitpicking about the best way to automa

I'm sure the output of this LLM could be improved with some hacking but why bother? Are we really at a point where we're nitpicking about the best way to automate away any kind of interpersonal communication?