AI users often employ polite phrases such as “please” and “thank you” when interacting with chatbots like ChatGPT. According to OpenAI CEO Sam Altman, these words have already cost the company tens of millions of dollars in electricity, but he emphasized that it was “money well spent.”
tens of millions of dollars well spent–you never know
— Sam Altman (@sama) April 16, 2025
Although there are no exact calculations, this user behavior has sparked a discussion about the appropriateness of politeness when addressing generative AI. Microsoft Copilot representative Kurt Beavers notes that using polite language affects how the model forms its response—if the model “notices” politeness, it is more likely to reply in the same tone.
Thanks to this feature, users can adjust the style of communication with AI, receiving friendlier or more neutral responses depending on their preferences. This can be useful for those who value a certain level of etiquette even in digital interactions.
At the same time, it is noted that the model recognizes not only politeness but also other emotional nuances in requests, including harsh or expressive statements. This allows users to experiment with the style of their questions to elicit the desired reaction from AI.
