Quantifying the Environmental Impacts of Large Language Models (LLMs) in Digital Humanities through Carbon Reporting

This talk provides an overview of the environmental impacts of using AI, particularly large language models, in research. It begins by briefly addressing the debated question of whether personal or academic LLM use is environmentally problematic. It situates these concerns within digital humanities discourse and highlights how influential AI-critical works, such as the famous 2021 Stochastic Parrots paper, foreground environmental issues. The talk also contextualises the current AI boom, drawing on AI Snake Oil, and refers to Kate Crawford’s critical work, including Atlas of AI, Anatomy of an AI System, ImageNet Roulette, and recent reporting on AI resources demands in a New York Times video feature. It then introduces the concept of the “guesstimate” to explain the difficulty of accurately measuring energy and water consumption associated with AI and data centres, referencing an instructive video by Hank Green. The working paper Quantifying the Environmental Footprint of Curating Datasets with Large Language Models is discussed as an example that includes a detailed literature review and demonstrates how carbon reporting might be implemented in digital humanities projects, while also noting the methodological challenges involved. Finally, the talk argues for more consistent carbon reporting in research involving large language models, invoking the principle that “what gets measured gets managed” (Peter Drucker). At the same time, it questions whether increased awareness and visibility alone are sufficient, or whether they risk fostering complacency without leading to substantive change. You may also be interested in the related working paper: Lang, Sarah, Wishyut Pitawanik, Pascal Belouin, Emma Sevink, Jesse Olszynko-Gryn, Alfred Freeborn, and Etienne Benson. “Quantifying the Environmental Footprint of Curating Datasets with LLMs”. Zenodo, December 11, 2025. https://doi.org/10.5281/zenodo.17902822 .