AI's Environmental Footprint: Insights & Actions

As AI systems scale and their use grows, their environmental footprint is becoming impossible to ignore. The energy and water demands of large language models now rival those of entire cities, yet most organizations remain opaque about their actual resource usage.

Join the ML+X community on Tuesday, September 9th, 1–2pm CT, via Zoom for our first forum of the fall semester. Nidhal Jegham, lead author of How Hungry is AI?, will introduce a new benchmark that measures the energy, water, and carbon costs of LLM inference. His work develops clever methods to estimate consumption across both open-source and proprietary models despite limited transparency, and provides a framework that weighs accuracy against resource use—helping practitioners, researchers, and everyday users select models more mindfully.

To complement this work, we'll also preview the launch of our 2025 WattBot challenge, as part of the 2025 Machine Learning Marathon kicking off on Thursday, September 11th. This challenge invites our community to build retrieval-augmented generation (RAG) systems to extract credible, citation-backed emissions data from peer-reviewed research. Tutorials, advisor support, and cloud credits included!

Zoom: https://uwmadison.zoom.us/j/92639425571?pwd=Z0tCaWZxK0dDcWs2dm51dXZpcy9mQT09

Passcode (if non-UW account): 111195.

 This talk is part of a monthly forum hosted by the ML+X community at UW-Madison. Join our Google group to be notified of future events!