Models · Hacker News ·
Humanising LLM Outputs Is Dumb
An essay argues that deliberately humanizing large language model outputs can make them less clear or useful. It examines how stylistic choices affect the quality and interpretation of AI-generated text.