Making AI (truly) sustainable: from environmental costs to social impacts
Details
With the rapid spread of generative AI, attention has increasingly focused not only on its benefits but also on its environmental impacts, including massive consumption of electricity and water, as well as the growing volume of electronic waste. This lecture explores the environmental footprint of AI and data centers, while also examining their effects on local communities, employment, and human thinking and creativity.
The lecture further introduces practical approaches for promoting sustainable AI use, such as selecting models that are appropriately matched to specific tasks, measuring usage patterns and energy consumption, leveraging smaller-scale AI systems, and developing frameworks and evaluation methods that enhance transparency.
Drawing on environmental, economic, and social perspectives, the lecture considers how AI can become a truly sustainable technology and discusses the roles and actions required of developers, organizations, and each of us as individuals to achieve that goal.
Speaker
Dr. Sasha Luccioni
Related Videos
