Comment by skissane

I sometimes go in the opposite direction - generate LLM output and then rewrite it in my own words

The LLM helps me gather/scaffold my thoughts, but then I express them in my own voice