The Hidden Cost of AI: Visualizing Energy and Carbon Impact of Machine Learning

Description

Every time someone asks ChatGPT a question or generates an image with AI, servers somewhere are burning electricity, but how much, and what's the environmental impact? Most people have no idea that AI inference has a carbon footprint, let alone how their usage patterns or model choices affect it. 

You'll build an interactive dashboard that makes these hidden costs visible by generating synthetic AI traffic, running lightweight machine learning models locally, and measuring real computational loads. By combining these measurements with known power characteristics of CPUs and GPUs, your system will estimate energy consumption and carbon emissions over time, showing how different models, prompt lengths, or usage patterns impact sustainability. This isn't just about displaying numbers; you'll create visualizations that help users develop intuition about AI's environmental footprint and make more informed choices. 

You'll gain hands-on experience with performance monitoring, energy profiling, data visualization, and the emerging field of sustainable AI, while building a tool that addresses one of the most important questions facing AI adoption: how do we balance the benefits of these powerful technologies with their environmental costs?