We are reader-supported. When you buy through links on our site, we may earn affiliate commission.
Artificial intelligence chatbots have become extremely common in everyday life, with ChatGPT among the first to gain widespread marketplace appeal and accessibility. Eco-conscious people have understandably wondered about the ChatGPT carbon footprint, and many researchers have sought more details through in-depth investigations. Even those don’t necessarily find conclusive answers.
Many people have had the frustrating experience of finding one study that says ChatGPT uses a tremendous amount of energy and other resources, only to come across another with a contrasting finding minutes later. That doesn’t necessarily indicate a fault with the researchers or their methods. It is the reality faced by people digging into the specifics of a relatively new and rapidly emerging technology. What should those who care about the planet and its future do when faced with these circumstances?
The first thing someone should do when they come across a news article or study offering new information about the ChatGPT carbon footprint is to look beyond the initial headline. Anyone who has ever taken a news writing class knows that editors excessively sensationalize many headlines, designing them to grab people’s attention and incite emotions.
Even if a headline contains a wholly true statement, it cannot tell the whole story. That’s why people should go into the content and read it completely instead of just checking out the headline — and possibly the comments others make about it on social media. It’s not always easy or possible to do this, primarily because paywalled content has become so common.
Sometimes, the only option is to read the headline and the first paragraph or so. After that, readers see a pop-up window that asks them to register at the site or purchase the article, and sometimes both.
Most don’t want to do those things to get the full content of something they randomly encountered online. But if the ChatGPT carbon footprint is a significant concern to them due to personal interests or occupation, they should look into ways to read the full content. In other cases, the best thing to do is to keep a balanced perspective about the content based on the available information. People should also think carefully before sharing partially accessible content, especially if the parts they can see may not accurately represent the complete study.
Another thing to remember is that ChatGPT does not represent AI as a whole. It is only a single tool within the wider category of generative AI. OpenAI, the company behind that product, has also released several versions and continues to do so. If enough users pressure its leaders to take environmental sustainability seriously, one can reasonably expect that developers will continually seek options with progressively smaller carbon footprints.
It’s easy to feel frustrated by the progression of AI when people go online and constantly see content commonly called “AI slop” for its low quality. But that material is only a tiny part of AI’s capabilities. But that material is only a tiny part of AI’s capabilities. For example, some experts are applying the technology to weather forecasting for more accurate results. If that approach can save lives by providing more reliable information about when severe weather will hit, most people would find a larger carbon footprint more justifiable.
Eco-conscious individuals should consider evaluating their usage, if applicable. If they find themselves regularly using AI chatbots to address boredom, they should consider filling their time in other ways, such as going for a walk.
However, the larger point is not to start feeling guilty, even if someone’s main use of ChatGPT and similar tools is to have fun. Even if someone is a heavy user, their decision to stop using the tool will not solve ChatGPT’s carbon footprint issue. Instead, tech decision-makers should take responsibility by continually developing more environmentally responsible products.
Even people opposed to ChatGPT for environmental or other reasons may not have a choice but to use it. Employers are progressively forcing workers to use AI, and some even track productivity that way.
People in that situation can consider complying with their employers’ requests while deciding to avoid using the technology for personal reasons whenever possible. Again, this could be harder than it looks. Most who interact with ChatGPT know it’s AI, but the underlying tech isn’t as obvious when using an email program or maps app.
Artificial intelligence is becoming so prominent in a wide assortment of products and services, so it is not always easy to avoid it, even when people have the best of intentions. Even so, some may deem that doing their best is better than remaining passive.
Each person has individual circumstances that could shape whether they use AI. Most would understandably not want to risk their jobs, but some would think twice before using ChatGPT or a similar tool for a frivolous task, especially if they take issue with its carbon footprint.
The primary way to interact with ChatGPT is to type requests into an interface. Experts suggest being as specific as possible when making these prompts. Doing that increases the likelihood of useful output. It also lowers the carbon footprint because people do not have to repeatedly ask for things in different ways after chatbots fail to respond usefully to their initial efforts.
That may mean asking the chatbot to respond as if it were in a specific occupation or role, or requesting that it provide the content in a particular tone. If people are asking a question, similarly to how they might use Google, they should be as detailed as possible about what they want to know to increase the chances of the chatbot giving them everything needed on the first try.
On a more general note, people should think carefully before using any chatbot to learn important information, especially if it relates to their health, finances, or another critical topic. Generative AI in its current state cannot understand someone’s question and context and respond accordingly. It also bases its answers on its training data, which may have substantial gaps, especially regarding recent events.
ChatGPT’s carbon footprint has attracted significant attention in recent years. Still, people should aim to view the information in context and not let themselves get drawn into headlines designed to capture attention.
Even if they find that the chatbot does use gigantic amounts of resources, they should remember that many other things they do every day have the same problem. Rather than getting stressed out over things they cannot immediately influence on a large scale, they should instead learn about and support efforts to make AI and other technologies more environmentally friendly. Those successes will convince other researchers that those products are worthwhile and have marketplace relevance.
AI company Hugging Face has a public-facing AI tool that allows people to calculate the energy usage based on tool and task. More brands will likely develop similar resources, especially if public sentiment warrants it.
That’s a sign they should seek advice from a medical professional about the next steps. Someone unable to participate in daily tasks or responsibilities due to these concerns may need mental health support.
As people learn more about the ChatGPT carbon footprint, they will likely find information about water usage. The consumption mainly comes from the cooling technologies needed to keep data centers operational. Those facilities keep AI and similar technologies running.
