# Teaching Students the Hidden Energy Cost of Artificial Intelligence
Most high school students know how to use ChatGPT or generate images with AI tools. Few understand the electricity consumption powering those systems.
A new framework from TeachThought emphasizes that AI literacy must extend beyond prompt engineering and code fundamentals. Students need to grasp the environmental footprint of the technologies they use daily. This approach treats energy awareness as a core component of digital citizenship.
Training large language models demands extraordinary amounts of electricity. OpenAI's GPT-3 consumed an estimated 1,287 megawatt-hours during its training phase. Google's BERT model required similar energy-intensive processes. Each time a student queries ChatGPT or runs an image generator, data centers consume power to process and respond. Over billions of queries, that energy accumulates into measurable carbon emissions.
The TeachThought framework proposes ten teaching strategies that embed environmental awareness into AI curricula. These strategies move beyond abstract discussions of climate impact. Instead, they ground AI literacy in calculation, comparison, and decision-making.
One approach asks students to calculate the carbon footprint of their own AI usage. Students access real data about energy consumption per query and convert kilowatt-hours into equivalent carbon emissions or other tangible metrics. Another strategy involves comparing the energy cost of AI solutions to traditional methods. Does using AI to grade essays actually save time and resources compared to traditional grading? What about using AI for weather prediction versus human meteorology?
A third strategy examines data center locations. Students research where companies like Microsoft, Google, and Meta locate their servers. They discover that energy costs and environmental regulations shape these decisions. A data center in Iceland benefits from geothermal power. A facility in Texas may rely on natural gas. Geography matters.
The curriculum encourages students to research corporate sustainability commitments. Has Amazon met its pledge to power data centers with renewable energy? Students become researchers, evaluators, and critical consumers of corporate claims.
Other teaching angles include studying the supply chain for hardware. Mining lithium and other rare earth materials carries environmental costs. Students trace the journey of a GPU from extraction to deployment to disposal. They explore e-waste management and the circular economy.
The framework also prompts discussions about equity. Wealthy nations and wealthy companies can afford renewable energy infrastructure. Developing nations often host data centers powered by coal. Which populations bear the environmental burden of AI innovation?
These lessons prepare students to make informed choices. When should a student use AI? When should they resist? The answer depends on context. Writing a grocery list hardly justifies the energy cost. Training a climate model using AI might justify substantial power consumption. Students graduate with frameworks for evaluation rather than blanket rules.
This approach aligns with growing attention to environmental literacy across K-12 curricula. Schools increasingly weave sustainability into science, social studies, and even mathematics. Adding AI and energy use to that mix reflects the reality that technology shapes environmental futures.
Educators integrating this framework begin with a simple premise: students should understand the systems they depend on. Opacity around energy and environmental impact serves nobody. Making these costs visible empowers students to demand efficiency, hold corporations accountable, and design technologies with environmental constraints in mind.