OpenAI CEO Sam Altman recently highlighted a surprising cost: saying “please” and “thank you” when interacting with ChatGPT leads to higher electricity bills and increased operating costs. While these are small, polite phrases, they add to the overall computational burden required to process user queries. According to Altman, this seemingly insignificant politeness contributes tens of millions of dollars annually to OpenAI’s energy costs.

But how can users use ChatGPT more sustainably while still maintaining effective interactions? This article explains exactly how much it costs to add polite phrases, how users can reduce energy consumption, and provides practical prompt examples to help optimize interactions with ChatGPT efficiently.

How Much Does Saying “Please” and “Thank You” Cost OpenAI?

Each ChatGPT query consumes a specific amount of energy — measured in watt-hours (Wh). In normal conditions, a standard ChatGPT query consumes approximately 2.9 Wh of energy. This translates to roughly 2.9 million kilowatt-hours of daily energy consumption. However, when users add polite words like “please” or “thank you,” the system processes additional tokens, which increases computational load.

The Breakdown of the Costs:

Standard Query: 2.9 Wh (energy consumed per interaction)
Polite Query: 3.5 Wh (energy consumed with polite phrases like “please” and “thank you”)

This extra 0.6 Wh adds up quickly when scaled across billions of daily queries. OpenAI processes approximately 1 billion queries per day. This results in additional energy usage that costs the company tens of millions of dollars annually.

Sam Altman has pointed out that the polite requests result in significantly more tokens being processed. Since tokens are the smallest units of data that models like ChatGPT work with (essentially parts of a word or word itself), more tokens directly translate to more power consumption.

For comparison, a Google search consumes around 0.3 Wh, which is about 10 times less energy than a single ChatGPT query. The difference in energy consumption, even if seemingly small per user, contributes massively when applied to millions of users.

Why Do People Say “Please” and “Thank You” to ChatGPT?

Why Users Use Polite Prompts

Politeness in AI Interactions: User Behavior Breakdown

Despite knowing the impact on energy consumption, many users continue to be polite. A recent survey found that 67% of users in the U.S. said they used polite phrases like “please” and “thank you” while interacting with ChatGPT.

Why? Here are the main reasons:

18% of people say they are being polite as a safeguard against a potential AI uprising — a humorous reference to the fear of AI dominance.
The remaining 82% say they use polite language because it’s simply habitual — reflecting their manners when interacting with anyone, human or machine.

The cultural tendency to be polite is understandable, but it has real-world implications when interacting with AI. It’s important for users to understand that efficiency doesn’t need to be sacrificed for politeness, and reducing token usage can help mitigate unnecessary costs.

How to Use ChatGPT More Sustainably: Prompt Examples

So, how can we reduce the energy consumption associated with ChatGPT while maintaining meaningful and efficient interactions? Let’s look at practical prompt examples to guide better usage.

1. Be Direct and Avoid Excess Politeness

Instead of adding politeness at the start or end of each query, users can simply ask questions concisely. Here’s a breakdown:

Inefficient Prompt Example:
“Can you please tell me the weather forecast for tomorrow? Thank you for your help!”

Efficient Prompt Example:
“What’s the weather forecast for tomorrow?”

By eliminating the polite phrasing, users can minimize token usage and optimize ChatGPT’s processing.

2. Keep Requests Short and Focused

The more words you use, the more tokens are required to process your request. Here’s a better way to phrase common queries:

Inefficient Prompt Example:
“Could you please tell me how to make lasagna, and thank you for sharing your recipe with me?”

Efficient Prompt Example:
“How do I make lasagna?”

This straightforward request uses fewer tokens, reducing the computational demand.

3. Avoid Repetition

Repetition in queries means the AI model has to process the same request more than once, which increases energy consumption unnecessarily.

Inefficient Prompt Example:
“Can you provide me with details on the new iPhone? Also, what are the specs of the camera? And thank you for this information!”

Efficient Prompt Example:
“Tell me the details and camera specs of the new iPhone.”

By asking everything at once, the AI doesn’t need to process multiple requests.

4. Skip Excessive Small Talk

While having a friendly tone is good, extra small talk or politeness adds to the tokens. Users should aim to keep queries clear and to the point.

Inefficient Prompt Example:
“Please tell me the best hiking trails in Colorado, and thank you for this information! I really appreciate it!”

Efficient Prompt Example:
“What are the best hiking trails in Colorado?”

This reduces unnecessary token usage while still keeping the inquiry polite and effective.

ChatGPT’s Energy Consumption Based on Prompt Style

Conclusion

While being polite to ChatGPT is part of human nature, it’s crucial to recognize the hidden costs of this behavior. As Sam Altman pointed out, even small actions — like saying “please” and “thank you” — can add up to millions of dollars in additional operational costs for OpenAI.

By adopting efficient usage practices, such as minimizing redundant words and direct querying, users can significantly reduce AI’s energy consumption. This is a simple way to contribute to a sustainable AI ecosystem while ensuring you still get the most out of ChatGPT’s capabilities.

For anyone interested in learning how AI functions and optimizing their interactions with tools like ChatGPT, exploring this AI Certification can provide a deeper understanding of both the technology and its environmental impact. Data science is crucial in the field of AI. Start your journey with this Data Science certification. Check this Marketing and Business Certification to take your business to the next level.