AI’s Energy Consumption: A Double-Edged Sword 

Sam Altman, ChatGPT Founder exclaim: “Don’t say ‘thank you’ and ‘please’ to AI—it’s costing us electricity.”

Artificial Intelligence (AI) consumes a significant amount of electricity, even for simple tasks like generating text messages. According to a recent study, creating a 100-word email using ChatGPT requires approximately 14 kilowatt-hours (kWh) of electricity—enough to power 14 LED bulbs for an hour. This raises important questions about AI’s sustainability

Understanding Where Energy Goes 

The intricacies of AI’s high electricity consumption stem from its database, a vital component of quantum computing. These databases are responsible for storing vast amounts of information and generating diverse outputs. For instance, the emotional tone that ChatGPT demonstrates in its responses is built upon meticulously programmed computer languages, all powered by energy-intensive processes.

To test this hypothesis, I posed emotionally charged questions to ChatGPT. Here’s what I discovered from this experiment that AI is charged with different information through database, often replicating our own interactions.

AI and Sustainability 

Electricity is the backbone of modern life, vital for almost every domain of human existence. In a world grappling with global warming, the increasing energy demands of AI present new challenges. The moral and ecological questions raised include: 

– Is AI predominantly powered by renewable or non-renewable energy sources? 

– Could this electricity consumption compromise the needs of future generations? 

– Does producing vast amounts of energy elevate pollution levels?  – How is the toxic waste, generated as a byproduct of energy production, managed?

These concerns highlight AI’s potential to disrupt progress on the United Nations’ 17 Sustainable Development Goals (SDGs). Addressing these challenges is crucial to ensure AI aligns with global sustainability efforts.