OpenAI CEO Claims Saying ‘Please’ and ‘Thank You’ to ChatGPT Costs ‘Tens of Millions of Dollars’ — Here’s Why (2025)

• Sam Altman — the CEO of OpenAI, the company that owns ChatGPT — claimed that using "please" and "thank you" in ChatGPT queries costs "millions of dollars" in electric bills
• Altman's response is in reference to the fact that queries on generative AI platforms use massive amounts of electricity, and additional words require additional computation
• A single query in ChatGPT uses enough energy to power a lightbulb for “about 20 minutes,” a researcher told NPR

Saying “please” and “thank you” to ChatGPT costs “millions of dollars” — at least according to OpenAI CEO Sam Altman.

Altman recently responded to an X user who posed the question, “Wonder how much money OpenAI has lost in electricity costs from people saying ‘please’ and ‘thank you’ to their models.”

“Tens of millions of dollars well spent,” replied Altman, whose company owns ChatGPT.

Altman’s answer refers to the fact that powering the data centers that generate AI responses requires massive amounts of energy — a.k.a electricity. According to researcher Jesse Dodge, a single query in ChatGPT uses enough energy to power a lightbulb for “about 20 minutes,” per NPR.

“So, you can imagine with millions of people using something like that every day, that adds up to a really large amount of electricity,” Dodge told the outlet in 2024.

The additional niceties that many users include in their queries when using AI platforms add additional language computation to the request, which increases the energy used by the model.

In 2024, Google released an emissions report that revealed its greenhouse gas emission levels had risen by 48% since 2019 — and it directly attributed the increase to the energy demanded by AI data centers.

Microsoft, which launched Copilot, its own generative AI platform, in 2023,reported similar findings, with a 29% increase in its greenhouse gas emissions since 2020, per NPR.

Many experts have voiced their concern over the amount of energy expended by these models — as well as the ensuing environmental impact — over the last several years.

“There's a lot of people out there that talk about existential risk around AI, about a rogue thing that somehow gets control of nuclear weapons or whatever,” Alex Hanna, the director of research for Distributed AI Research Institute, told NPR for the same piece.

“That's not the real existential risk. We have an existential crisis right now. It's called climate change, and AI is palpably making it worse,” she added.

However, a number of AI researchers also stress the importance of using respectful etiquette when communicating with AI models like ChatGPT and Copilot, in large part because the models are actively learning from our communication cues.

“When you ask it to do something for you, the quality of what you get back is a function of how you make the request,” notes Microsoft WorkLab, an information hub with tips and advice for best harnessing the power of generative AI.

“Using polite language sets a tone for the response,” explains Kurtis Beavers, a director on the design team for Copilot, in the same WorkLab memo. This also means you are more likely to get a genuinely more helpful and useful response from generative AI when you approach it in a collaborative manner.

According to a December 2024 survey from Future, the publisher that owns the technology site TechRadar, 67% of Americans who use AI reported that they are polite to AI chatbots.

Additionally, 82% of the respondents who said they prefer to use good manners when communicating with AI said they do so simply because it is the “nice” thing to do — while the remaining 18% jokingly responded that they don’t want to upset AI in case of a future “robot uprising.”