A user on X, formerly Twitter, raised a curiosity-provoking question, “I wonder how much money OpenAI has lost in electricity costs from people saying ‘please’ and ‘thank you’ to their models.”

This query, which seemed random at first, prompted a response from OpenAI CEO Sam Altman, who replied to the user’s post on X. Altman stated that the cost of processing these polite phrases has added up to “tens of millions of dollars well spent — you never know,” likely speaking in a humorous tone, as it’s unlikely he conducted an exact calculation.

Altman’s response inspired Futurism to investigate whether being polite to AI chatbots like ChatGPT is a waste of time and electricity. The outcome of this inquiry revealed that politeness is not merely a redundant habit or an attempt to avoid offending our potential future AI overlords.

According to Kurt Beavers, a director on the design team for Microsoft Copilot, using polite language sets the tone for the AI’s response. When an AI model detects politeness, it is more likely to respond politely as well. This suggests that being courteous to AI is not futile.

However, it’s worth noting that profanity can also be useful in certain contexts, demonstrating that the relationship between language and AI is complex and multifaceted.