Eco-Conscious GenAI Use for Language Teachers – GenAI and All of That: Part 2

This post was intended to be titled, “GenAI and All of That: Part 2” as a follow up to my previous post. Since May, so much has changed in the artificial intelligence sphere that I thought posting additional common features of GenAI sessions delivered at IATEFL in April was not as important as providing my peers with the opportunity to become aware of consequences of using GenAI tools. More importantly, share advice on lessening  their GenAI carbon footprint moving forward into the new academic year.

At the international IATEFL conference, Vicky Saumell, 2025 TESL Ontario conference keynote speaker, shared a GenAI reality check related to the ecological consequences of using GenAI.   Many in the audience were surprised by these facts as up to that point most of the AI sessions were focused on what AI could do and how language teachers could exploit these new technologies.   

GenAI and the Ecosystem 

GenAI offers great potential for innovation, efficiencies and enhancements for our personal and professional endeavors. In a recent blog post, Sam Altman, the CEO of OpenAI, the creators of ChatGPT, wrote, “The average query [Prompt], uses 0.34 watt-hours of energy.”  He continued, “This is the same energy a high-efficiency lightbulb would use in a couple of minutes.” (Altman, 2025) 

You ask yourself, “So what?” This is such a small amount of energy. In our education centres, we consume energy for lighting, cooling, heating, IT services and devices, clocks, copiers, security systems and more. 

In July, ChatGPT reported 800 million active, weekly users. (Singh, 2025) This leads to approximately a billion queries a day. Also, not all prompts are the same, some consume more energy than others.  Generation of images consumes approximately twenty times the energy than a simple text prompt, and consider the explosion of GenAI integration into new online tools as well as existing tools such as Quizlet, Google search, Microsoft Word.  This energy is consumed at data centers around the globe.  Energy usage and carbon production and water consumption are beyond the scope of this post; however, below are some tips to consider in using AI responsibly.  

Eco-friendly practices 

Consider the following suggestions to reduce energy consumption while using GenAI tools. Consider exploring the links in the resources section below to learn more. 

1. Ensure that automatic AI responses on search engines are turned off. Chrome defaults to no AI assistance. To remove a Google AI function add “-ai” at the end of a search string.  For example, “What is the best pizza in Melville? -ai”. For Bing searches, while logged in to a Microsoft account, disable AI Copilot responses in the settings.
2. Turn off 3rd party tools such as Quizlet and Word 365 by locating the ‘disable AI’ toggle or setting on a tool-by-tool basis.
3. Prompt precisely with specific prompts to receive accurate responses.  This also eliminates iterative prompting which consumes more energy that a single prompt.  This is very important when generating images to save energy. 
4. Reuse your outputs by,  
   • Managing an organized chat history in ChatGPT. 
   • Storing prompts and responses in NotePad documents, folders or spreadsheets.  
   • Creating a personal prompt bank of reusable prompt or join a cooperative prompt bank.
5. Prompt directly without courtesy. Politeness does not improve the response quality, but it does consume more energy.  
6. Do not idle your GenAI tool.   The data centres poll your browser to see if there is a prompt waiting for an answer. While this does not consume a great deal of energy, it moves the needle, so close the tab or window if you are not engaging with AI.
7. If you are working with GenAI to create script or code, specify which output format you require. 
8. Look for lightweight GenAI versions.  For example, GPT-4.o uses more energy than GPT-3.5.
9. Do not assign learners to work with GenAI chatbots.  It is not entirely safe letting a student discuss freely with a GenAI chatbot and the number of exchanges would lead to an enormous amount of energy consumed. 
10. Consider typing prompts rather than speaking to a GenAI in voice mode. 
11. Limit learner facing GenAI activities in your lessons and set focused tasks that don’t do a great deal of exploration.
12. Ask yourself, is this a task I can do without GenAI?  We all did our jobs without GenAI before, why not now? 
13. Share these and other low-impact GenAI practices with your peers, learners, and friends. 

Final Thoughts 

As you have read through this list, you might reflect on overall online digital activities with the knowledge that data centres consume energy, water, and contribute to the global carbon issue. Consider raising awareness of this issue with your students and peers as well.  

Relevant Links 

Allan, J. (2025). Gen-AI and All of That: Part 1, https://blog.teslontario.org/gen-ai-and-all-of-that-part-1  

Altman, S. (2025). The Gentle Singularity. Sam Altman Blog. Retrieved July 23, 2025, from https://blog.samaltman.com/the-gentle-singularity  
 
Saumell, Vicky. Developing strategies to address GenAI issues. TESL Ontario Keynote. https://site.pheedloop.com/event/EVEGMCSAXLRSY/program/Keynotes 
 
Saumell, Vicky. (June 2025). Developing strategies to address AI issues. Avenue webinar. https://youtu.be/bdeyxknfcRI?si=64NcYO9eyNAW1IAO&t=141 

Singh, Shubham. (July 2025). ChatGPT Statistics 2025 – DAU & MAU Data [Worldwide] https://www.demandsage.com/chatgpt-statistics