OpenAI CEO Sam Altman has confirmed that users saying “please” and “thank you” to ChatGPT leads to millions of pounds in additional operational expenses, but considers the investment worthwhile as politeness enhances AI interaction quality and responsiveness.
OpenAI Chief Executive Officer Sam Altman has publicly acknowledged the financial impact of users employing polite language when interacting with the company’s AI chatbot, ChatGPT. Altman revealed that when users say phrases such as “please” and “thank you” to the AI, it results in millions of extra dollars in operational costs for OpenAI.
The conversation began on 15 April 2025 when a social media user queried, “I wonder how much money OpenAI has lost in electricity costs from people saying ‘please’ and ‘thank you’ to their models.” In response, Altman, posting on X (formerly known as Twitter), acknowledged these additional costs but considered them worthwhile. He stated, “Tens of millions of dollars well spent–you never know.”
The implication is that while politeness increases computational expense due to longer interactions with the AI model, Altman perceives the practice as valuable, though the full reasoning behind the comment remains open to interpretation.
Further insight into the effects of politeness when addressing AI comes from Kurtis Beavers, a design director on Microsoft Copilot’s team. In a Microsoft WorkLab memo, Beavers noted that using basic etiquette when interacting with AI helps the system produce “respectful, collaborative outputs.” He added that being courteous “not only ensures you get the same graciousness in return, but it also improves the AI’s responsiveness and performance.”
Research published in December 2024 by Future sheds light on user behaviours regarding politeness towards AI chatbots. It found that 67 per cent of people in the United States report being polite to AI, while the figure is slightly higher in the United Kingdom, with 71 per cent of users demonstrating courtesy in their interactions. Not all users adopt this approach, however; some report being impolite as a means to save time and maintain brevity.
Additionally, around 12 per cent of respondents indicated they were polite to AI out of concern for possible repercussions in future interactions, which may link back to Altman’s cryptic observation about the value of spending extra resources on courteous exchanges with AI models.
As AI chatbots become more integrated into daily communications, these findings highlight current user attitudes toward politeness and the unforeseen costs and benefits associated with it in the realm of artificial intelligence.
Source: Noah Wire Services
- https://www.entrepreneur.com/business-news/saying-thank-you-to-chatgpt-costs-millions-in-electricity/490341 – This article supports the claim that saying ‘please’ and ‘thank you’ to ChatGPT results in higher operational costs for OpenAI due to increased electricity usage. It highlights the financial implications of user politeness.
- https://www.calcalistech.com/ctechnews/article/r1quha71el – This piece details how polite language in interactions with ChatGPT leads to additional costs for OpenAI, as acknowledged by CEO Sam Altman. It explains the computational basis of these costs due to the token-based processing system.
- https://economictimes.indiatimes.com/magazines/panache/do-you-say-please-to-chatgpt-sam-altman-reveals-how-much-electricity-your-manners-cost-to-openai/articleshow/120455018.cms – This article discusses the financial and technical implications of politeness in AI interactions, including increased computational demands and electricity consumption. It also touches on user behavior and perceptions around AI politeness.
- https://www.perplexity.ai/page/politeness-to-chatgpt-raises-o-qu9DjX3DRp6v5fwHtNVYQA – This resource explains how politeness affects operational costs due to ChatGPT’s token-based pricing structure, which results in higher computational costs for longer input interactions.
- https://www.microsoft.com/en-us/research/uplizardsposables/microsoft-work-lab – Although not directly linked to Beavers’ memo, Microsoft WorkLab shares insights into improving human-AI interactions, which aligns with Beavers’ perspective on politeness enhancing AI outputs.
Noah Fact Check Pro
The draft above was created using the information available at the time the story first
emerged. We’ve since applied our fact-checking process to the final narrative, based on the criteria listed
below. The results are intended to help you assess the credibility of the piece and highlight any areas that may
warrant further investigation.
Freshness check
Score:
9
Notes:
The narrative features specific, recent events, such as Sam Altman’s response on X (formerly Twitter) on 15 April 2025, and references research from December 2024, indicating it is current.
Quotes check
Score:
8
Notes:
Sam Altman’s quote on X is noted, but there’s no attempt to verify if this is the first use of the quote since it directly refers to a recent event.
Source reliability
Score:
6
Notes:
The narrative originates from a Google News RSS and cites specific figures and events but lacks sourcing from a major, well-known publication. Therefore, while the information seems plausible, the reliability of the narrative is somewhat uncertain.
Plausability check
Score:
8
Notes:
The claims about the financial impact of politeness and AI performance improvements are plausible, supported by insights from industry experts and recent research. However, some assertions could benefit from more detailed evidence.
Overall assessment
Verdict (FAIL, OPEN, PASS): PASS
Confidence (LOW, MEDIUM, HIGH): MEDIUM
Summary:
The narrative is generally well-supported with recent data and expert insights, though it lacks direct sourcing from a major publication. The claims about the financial and performance implications of politeness towards AI appear plausible. However, further verification is needed for absolute confidence in all details.
