A user’s seemingly casual question on X about the electricity costs of saying ‘please’ and ‘thank you’ to AI models caught OpenAI CEO Sam Altman’s attention. Altman responded with a mix of humor and insight, stating that these courtesies have amounted to ‘tens of millions of dollars well spent.’ While Altman’s tongue-in-cheek response suggested he hadn’t done a precise calculation, it sparked a discussion about the implications of being polite to AI chatbots like ChatGPT.
Kurt Beavers, a director on Microsoft Copilot’s design team, shed some light on the matter. According to Beavers, using polite language when interacting with AI ‘sets a tone for the response.’ He explained that when an AI model detects politeness, it’s more likely to respond in kind. This insight suggests that being courteous to AI isn’t merely a redundant habit or misplaced anthropomorphism; it actually influences the AI’s response.
The conversation highlights an interesting dynamic in human-AI interaction. While being polite to machines might feel unnecessary, it appears to have a tangible effect on how AI responds. As we continue to interact with AI systems, understanding these nuances becomes increasingly important. The discussion also raises questions about the broader implications of our interactions with AI and how they might shape our digital experiences.
