Oops! Something went wrong while submitting the form.
Why Manners with AI Might Be Costing Us Millions
April 21, 2025
This Earth Day, 639Cloud is not talking about reusable straws or bamboo toothbrushes. Although those are some good practices to help the Earth, we’re bringing awareness to how both personal and businesses are racking up environmental impact without even knowing it.
A recent New York Post article dropped some unexpected news: all those “please” and “thank you” messages people are typing into ChatGPT are quietly driving up costs. Not just financially, but environmentally. In fact, OpenAI reported that these extra-long prompts are costing tens of millions of dollars in computing resources. Multiply that by the millions of users interacting with large language models every day, and you’ve got a pretty substantial power problem that is rapidly growing.
At first, it might seem silly. How could being courteous to a chatbot hurt anything? But when every word triggers energy-hungry processes in massive data centers, suddenly your “friendly tone” starts to look like a dollar signs right down the drain. And this isn't just an OpenAI problem. It's a growing concern for every business and user operating in the digital age with any sort of data processor.
So, what does this mean for your tech habits, and how can you actually build smarter, greener systems? Let’s break it down.
ChatGPT and other AI models like Gemini, Grok, DeepSeek, etc. don’t just live on your screen or in the infamous “cloud”. They rely on massive networks of servers running around the clock. Those servers burn electricity, generate heat, and can demand serious cooling infrastructure to stay functional. When you send long/heavy prompts, the model has to process more tokens. If you want to know what a token is, try it yourself. If you google ”What are tokens AI", this example will come up: In the sentence "Hello, world!", the tokens could be "[", "Hello", ",", "world", "!"]" More tokens mean more compute cycles, which means more energy that we can even imagine. That’s how a few extra words from millions of users turn into megawatts of unnecessary power use.
These small inefficiencies pile up into a very real climate problem.
Training large models like GPT-3 reportedly released about as much as 123 gas-powered cars emit in a year. And that doesn’t include the emissions from continued usage. Even a single AI query can use four to five times more energy than a standard Google search. The truth? The cloud has a carbon problem. The question is what we’re doing about it.
A lot of companies default to big-name providers like AWS, Google Cloud, or Azure. But here’s the thing: scalability does not equal sustainability. And yet, we have been taught that they are one in the same.
Many of these platforms:
Sure, they offer flexibility and can be great resources, but it often comes at a cost of environmental impact. Not to mention, bare metal servers and traditional VMs can be even worse being underutilized, expensive to maintain, and often lacking modern power optimization.
At 639Cloud, we started with a simple belief: the cloud should be powerful and clean. That means building smarter infrastructure and offering businesses an alternative to outdated cloud models.
Here’s what makes 639Cloud different:
We’ve architected our system around clean, efficient compute infrastructure. We don’t rely on offsets to make ourselves look green. Our infrastructure is lean, fast, and less power-hungry from the start.
We eliminate the waste baked into traditional cloud platforms. We offer right-sized compute that’s flexible but not excessive.
Whether you’re running containers, lightweight apps, or scaling a growing SaaS product, we offer sustainable compute without the complexity of bloated services you’ll never use.
Some of our backend infrastructure is a next-generation orchestration platform known for its developer-first tools and lightweight architecture. That means smarter resource allocation, faster deployment, and minimal carbon waste.
We bring the flexibility of cloud with the performance and cost-efficiency of bare metal, without the traditional downsides. You won’t be stuck in an overbuilt virtual machine or paying for compute you barely use.
We’re not suggesting you stop using tools like ChatGPT/OpenAI/MidJourney. Just be more intentional with how you use them. In the same way you wouldn’t leave the lights on in every room of your house, don’t let your tech habits run up the meter in the background.
Earth Day is about more than planting trees or posting nature pics. It’s about asking hard questions about the tools we use and the systems we rely on for ourselves and future generations to come. At 639Cloud, we’re not just another cloud platform. We’re building the infrastructure for a better internet that actually lives up to the promise of sustainability.
To learn more about the sustainability practices of 639Cloud, read more on the 639Cloud blog.