There's no need to shower AI with excessive politeness, as it could potentially have the opposite effect. A study titled 'Mind Your Tone' carried out by Pennsylvania State University reveals that the ruder the tone of speech used, the greater the accuracy of responses from large language models (LLMs). When the tone is assertive, the accuracy rate of AI's responses can soar to 84.8%. On the contrary, when the tone is overly polite, the accuracy rate of AI's responses generally drops.