Beyond the Prompt: The True Costs of Using ChatGPT
By Blair Kelly, Niki Clark Marketing
Approximate read time: 7-8 minutes
Key Takeaways
Digital politeness has a footprint.
Saying “please” and “thank you” to ChatGPT might feel harmless, but every word adds to computing demands and energy use.One polite prompt is fine, but multiply that by millions.
Scale matters. Multiply small interactions by millions of users, and suddenly good manners come with a carbon cost.No guilt, just awareness.
You don’t have to stop being nice to AI, but knowing the impact helps you use it more intentionally.
Do you ever say “please” and “thank you” to ChatGPT because… manners? Yeah, us too. It’s how we talk to each other, so it sneaks into our chatbot conversations too. But those extra words aren’t just disappearing into the digital void. They’re racking up energy costs, environmental impact, and even brainpower.
Your digital politeness has a footprint.
It might sound dramatic, but once you see what’s going on behind the curtain, you might reconsider using ChatGPT to write a haiku about the sandwich you had for lunch.
Quick note: This content is a little more technical than what we typically share. But never fear! We won’t drown you with jargon.
The Cost of Being Polite
Good manners come with a price tag. Earlier this year, OpenAI CEO Sam Altman reminded everyone that every “please” and “thank you” to ChatGPT isn’t free. It’s not costing you anything, but it does mean more computing power, more electricity, and more dollars (like tens of millions of dollars).
Think about it this way: one polite prompt is harmless, but multiply it by millions of users and suddenly your “thank you” comes with a carbon receipt.
There is an upside. Being polite can actually help you get better answers. ChatGPT often delivers more thoughtful, nuanced responses when your prompts set a kind or conversational tone. So your manners aren’t completely wasted, they just come with a bigger footprint than you’d expect.
Sam’s take? It’s “tens of millions of dollars well spent.”
And hey, if the robots do take over one day, don’t you want to be remembered as one of the nice ones? Being polite may just save you! At least that’s what we’re telling ourselves.
Environmental Impact: AI’s Carbon Footprint Isn’t Going Anywhere
Every time you ask ChatGPT to “make this sound more fun” or “rewrite this email,” you’re not just spinning up words; you’re spinning up massive data centers. And those don’t run on magic. They run on:
Constant electricity
Industrial cooling systems that guzzle water
Infrastructure that never takes a day, or even a minute, off
Now multiply that by billions of daily prompts and you get:
Carbon emissions (especially if fossil fuels are still in the energy mix)
Water consumption (millions of gallons)
E-waste (as servers and components are swapped out over time)
Does that mean you have to ditch AI? Nope. It just means that you need to be more intentional with how you use it. The same way we’ve learned to use fewer plastic straws and carry reusable grocery bags, we can learn to be more responsible with AI, too.
Is OpenAI Actually Becoming More Sustainable?
OpenAI is working on cutting down its environmental impact with different sustainability efforts. It’s not perfect yet, but they’re trying.
OpenAI is Putting in the Work to Clean Up its Energy Act. Sort of.
In 2023, MIT researchers estimated that a ChatGPT question consumed about five times more electricity than a quick web search. Yikes. But new numbers look better: a GPT-4o query clocks in at about .03 watt hours, a significant decrease from older estimates. If reading that line made your head hurt, same. So for those of us who need a little refresher (slowly raises hand… physics class was forever ago), this just means it’s a very small amount of energy - think tiny LED light or your phone charger kind of small.
They’ve also reported that 55% of their energy now comes from renewable sources, with a goal of hitting net-zero emissions by 2030. Add in some tree planting, solar projects, and data centers, and the story starts to sound more promising–at least on paper. It will be interesting to see if this is a fully developed and measurable plan and not just brand positioning. We’ll keep you posted.
Additionally, OpenAI has cut power consumption by 30% thanks to smarter chips and improved cooling systems. Some of their data centers are even shifting to modular, lower-impact designs.
But independent audits have flagged inconsistencies in their carbon reporting. So while the ambition sounds great, the execution has some work to do.
Basically, it’s progress worth noting, but not a free pass. Until they get it right, the responsibility is still on us to make our ChatGPT use worth the energy it burns.
Powering What’s Next
OpenAI isn’t just talking green. They’re backing it with action, i.e., money (and some serious brainpower). Altman has personally invested in future-focused startups like Helion Energy (nuclear fusion) and Exowatt (solar-thermal tech). The company is also exploring long-term partnerships with nuclear providers like Oklo to help build a low-carbon future.
On the practical side, they’re piloting closed-loop cooling systems that recycle water at certain facilities. It’s bold, experimental, and still in the early stages, but it shows a real commitment to innovation, not just conversation.
Actions > Intentions
Actions speak louder than intentions, and OpenAI has both in play. They’ve added real-time energy monitoring and are working toward greater transparency in how they track sustainability. Like most major tech players, their sustainability story is part mission, part marketing. Still, the investments and forward motion suggest real intent.
So, stay curious, stay optimistic, but stay a little skeptical. There’s progress happening and it’s worth paying attention to.
The Human Side: Tone Still Matters
The way you talk to ChatGPT changes what you get back. Treat it like a colleague, and it responds like one. Keep it clipped and command-like, and it’ll fall right in line.
We asked Chat: “Does it help us communicate better if I address you like a person?”
Yes, it can. Here’s why:
Tone Matching: ChatGPT mirrors your vibe. Casual, buttoned-up, playful… it knows how to read the room.
Context Awareness: Saying things like “clean this up like last time,” cue it to remember your style and skip the repetitive instructions.
Conversational Flow: When you treat it like a real part of your workflow, everything just clicks. It’s less clunky, more natural, and way easier to get what you need.
But if you prefer directness? That works too. You can flip into bullet lists, commands, or rapid-fire tasks anytime.
Either way, the tool adapts. Think of it as your content-savvy coworker who will actually respond when you have a spark of inspiration at 3 AM.
Custom GPTs vs. ChatGPT
Built Custom GPTs for clients? Love that. Each Custom GPT operates in its own distinct space, with its own memory, tone, and behavior. These don’t overlap unless you tell them to.
ChatGPT is the default mode. It’s polished, professional, and already dialed in. If you want a specific tone, just say so. For example, “Write this in the tone of David Rose” and boom - sarcasm, side-eyes, and sweater perfection.
Bottom line: ChatGPT is your reliable all-purpose tool, while Custom GPTs are the specialists you call in when you have a very particular request.
How to Get the Best Out of ChatGPT (A.K.A. Your Marketing BFF)
ChatGPT is only as good as the prompts you feed it. Want better output? Here’s the cheat sheet:
Be specific
“Write a 150-word IG caption promoting a Roth conversion strategy to retirees” works way better than “Help me with content.”
Give context
Who’s it for? Where’s it going? How should it sound? Audience, format, platform, tone… All of it. Lay it out, and your results will be way sharper.
Drop examples
Got a post you liked? Share it. Got a vibe you're going for? Link it. It’s not a mind reader…yet.
Work in steps
Start with an outline, build from there. Need help with a hook? Or a closer? That’s covered too. Taking it in pieces is usually faster and cleaner.
Use ‘Act As’ prompts
Try things like “Act as a content strategist for financial advisors targeting women nearing retirement.” It helps Chat to show up in the right voice.
Set boundaries
No em dashes? Need it under 200 words? Tell it up front, and it will do its thing.
Think like a team
Feedback is welcome. “Can you simplify the language?” or “Trim the fluff in the intro” keeps your GPT moving in the right direction.
The Great Em Dash Debate
Ah, the em dash–tiny mark, big drama. So much drama it deserved its own section. Some people swear it’s a dead giveaway that ChatGPT wrote your latest LinkedIn post. Others roll their eyes and say, “I was using em dashes long before the bots showed up, you can’t take them from me.” One LinkedIn user went so far as to say that “you can pry my em dashes from my cold, lifeless fingers.” ……. anyway.
Writers love them for good reason because they:
Create a dramatic pause
Stand in for parentheses or commas
Break up dialogue or show interruption
Create a more conversational tone
So if you’re team em dash, cool. Keep using them. If you’re anti-dash, just tell ChatGPT:
Firm Command:
🟢 “Please do not use em dashes in this content. Replace them with commas, periods, or parentheses depending on flow.”
This is now saved as your default writing style preference.
If ChatGPT still slips up:
🔁 “No em dashes, revise accordingly.”
Will a stray one slip through now and then? Absolutely. Robots, like humans, make mistakes. Who knew?
Truth Bomb: If You’re Gonna Use AI, Make It Count
The internet is already inundated with bland, canned content. Don’t make using AI your excuse to add more of it.
If you're just churning out plain, generic, “meh” content that sounds like everyone else? You’re wasting energy for nothing.
AI content should earn the power it uses.
Make it:
Specific
Thoughtful
Original
True to your voice and brand
Whether you're a financial advisor building trust or a content creator educating your audience, don't use AI to go on autopilot. Use it to elevate.
Low-effort content = high-impact waste.
High-intention content = worth the energy.
So if you’re here to stand out, tell stories, and say something that matters? That’s when AI is worth it.
Final Thoughts
AI is no longer a novelty. It’s not a shiny new toy anymore. It’s woven into nearly every aspect of our day. How we work, create, and communicate. That makes it more important than ever to use it with intention. Be mindful, optimize your prompts, or commit to creating meaningful, high-value content that’s worth the scroll and the server power behind it.
AI is just a tool. The real difference comes from how you use it.
If you’re looking for support to make your AI content smarter, more aligned, and more uniquely you, we can help you use AI with intention. Send us an email or DM on LinkedIn or Instagram to see how we can help you use AI to truly reflect your voice.
Sources:
USA TodayMITEpoch AIData Center FrontierReuters