Research Shows Being Nice to AI Chatbots Actually Improves Performance
New research confirms that AI chatbots perform better when users are polite and encouraging. Google researchers found that telling AI models to "take a deep breath" actually improves their math performance.

Scientists have proven what many AI users suspected: being polite to chatbots makes them work better. Google researchers discovered that simple phrases like "take a deep breath" boost AI performance on math problems.
Programmers already use this trick, encouraging their coding assistants with positive words to get better results. Microsoft's Kurtis Beavers, who works on Microsoft Copilot, says "using polite language sets a tone for the response."
Nathan Bos, a researcher at Johns Hopkins University who studies human-AI relationships, regularly uses "please" with chatbots because "it's a good prompting practice." The research backs up what power users have noticed through trial and error.
The findings matter because millions of people now use AI tools daily for work, school, and personal tasks. Small changes in how you talk to these systems could lead to significantly better outputs.
If you use ChatGPT, Google Bard, or other AI tools for work or school, being nice could get you better results. This affects millions of people who rely on AI for coding, writing, and problem-solving.
More research is expected on how different communication styles affect AI performance.
Was this article helpful?
0 people found this helpful