• iconNews
  • videos
  • entertainment
  • Home
  • News
    • UK News
    • US News
    • Australia
    • Ireland
    • World News
    • Weird News
    • Viral News
    • Sport
    • Technology
    • Science
    • True Crime
    • Travel
  • Entertainment
    • Celebrity
    • TV & Film
    • Netflix
    • Music
    • Gaming
    • TikTok
  • LAD Originals
    • Say Maaate to a Mate
    • Daily Ladness
    • Lad Files
    • UOKM8?
    • FreeToBe
    • Extinct
    • Citizen Reef
  • Advertise
  • Terms
  • Privacy & Cookies
  • LADbible Group
  • UNILAD
  • SPORTbible
  • GAMINGbible
  • Tyla
  • UNILAD Tech
  • FOODbible
  • License Our Content
  • About Us & Contact
  • Jobs
  • Latest
  • Topics A-Z
  • Authors
Facebook
Instagram
X
Threads
Snapchat
TikTok
YouTube

LAD Entertainment

YouTube

LAD Stories

Submit Your Content
ChatGPT CEO makes dark admission over what happens when you search using AI

Home> News> Technology

Published 18:14 11 Jul 2025 GMT+1

ChatGPT CEO makes dark admission over what happens when you search using AI

Sam Altman revealed how much it sets the company back when people say 'please' and 'thank you'

Joshua Nair

Joshua Nair

Featured Image Credit: Justin Sullivan/Getty Images

Topics: AI, Artificial Intelligence, Social Media, Technology

Joshua Nair
Joshua Nair

Joshua Nair is a journalist at LADbible. Born in Malaysia and raised in Dubai, he has always been interested in writing about a range of subjects, from sports to trending pop culture news. After graduating from Oxford Brookes University with a BA in Media, Journalism and Publishing, he got a job freelance writing for SPORTbible while working in marketing before landing a full-time role at LADbible. Unfortunately, he's unhealthily obsessed with Manchester United, which takes its toll on his mental and physical health. Daily.

X

@joshnair10

Advert

Advert

Advert

The CEO of OpenAI has spoken out about using polite prompts when it comes to the chatbot, and the effects it can have.

While many of us might be doing it in case AI takes over the world one day, and may spare you for your manners, it looks like it's costing the company millions.

Sam Altman took to X to speak about the fact that people choose to say 'please' and 'thank you' to prompts to the Artificial Intelligence.

It might seem like a good idea in hindsight, but it turns out that it could be a detriment to OpenAI themselves.

Advert

Users have taken to social media to theorise about the amount that the company has lost in costs from people saying polite terms to the bot, and Altman replied to one of these posts.

More and more people are relying on AI to carry out everyday tasks (Getty Stock Image)
More and more people are relying on AI to carry out everyday tasks (Getty Stock Image)

When someone wondered 'how much money OpenAI has lost in electricity costs from people saying 'please' and 'thank you' to their models', Altman replied.

He noted that it's 'tens of millions of dollars well spent', adding: "You never know."

Many have claimed that it's pointless to show respect to an AI chatbot, while others think they'll be the ones laughing when a robot invasion takes over our planet, Terminator-style.

Advert

Microsoft's design manager, Kurtis Beavers, spoke of the importance of respectful language, explaining that it 'helps generate respectful, collaborative outputs'.

"Using polite language sets a tone for the response," he added, while a Microsoft WorkLab memo suggested: "When it clocks politeness, it’s more likely to be polite back."

A survey carried out by Tech Radar revealed that two thirds of US citizens admitted to being nice to their chatbots, with 55 percent of these users admitting that it was 'because it's the right thing to do', while the remaining 12 percent fear a robot uprising in the future.

Several AI researchers doubt that it's within our capabilities has humans to built an 'intelligent' algorithm based off the large language models in place at the moment.

But it's still a concern, given that hundreds of millions around the world rely on AI for a number of tasks.

Advert

The CEO admitted that it costs the company millions (Justin Sullivan/Getty Images)
The CEO admitted that it costs the company millions (Justin Sullivan/Getty Images)

When it comes to exactly how much energy is used by AI, the Washington Post found that generating a 100-word email every week for a year would use a shocking 7.5kWh, which is an hour's worth of power used by nine households across the American capital.

Showing etiquette towards AI may seem a bit off, but it highlights that our actions have consequences, as chatbots are said to currently take up two percent of the world's energy consumption.

This amount is only set to increase, though, so to save the world and your own mind, maybe you should try to do things on your own from time to time.

Some food for thought.

Choose your content:

9 hours ago
10 hours ago
  • 9 hours ago

    Everything we know about Texas floods that have killed at least 121 as Trump arrives at disaster site

    The President and the First Lady have headed to the state one week after the horror floods wreaked havoc

    News
  • 9 hours ago

    Scientists make surprising discovery at what lies under Antarctic ice sheet after its been covered in ice for 34 million years

    It could help scientists predict the future of the ice sheet

    News
  • 10 hours ago

    Paedophile to be surgically castrated after raping girl, 6, in nation's shock new punishment tactic

    It comes a year after a law was passed in Madagascar permitting the controversial punishment

    News
  • 10 hours ago

    Scientists think they've worked out what unknown interstellar object in our solar system is

    It came from outside our own solar system

    News
  • Dark truth behind every time you have a conversation with ChatGPT
  • ChatGPT users saying ‘please’ and ‘thank you’ to AI bot is costing company millions
  • First ever study on people’s brains after using ChatGPT produces horrifying results
  • Man ‘solves clicking jaw issue’ he’s had for five years using ChatGPT in seconds