Project Overview

The IEEE Computer Society at German Jordanian University (GJU) released its second post in the “Research Summaries” series on November 21, 2024. This summary explores the environmental impact of artificial intelligence (AI), particularly focusing on the energy and water resources required for training and operating generative AI (GAI) models. The post was shared across the society’s official platforms, including Instagram, LinkedIn, and Threads.

Objectives

The primary objectives of this research summary were to:

      1. Highlight the significant environmental costs associated with AI technologies.

      2. Discuss AI’s potential for improving energy efficiency across various sectors.

      3. Foster awareness and dialogue about sustainable AI development.

Content Highlights

1. Energy and Water Consumption

– Training and operating AI models require substantial energy, with data centers contributing heavily to global electricity demand.
– Cooling systems for these centers consume significant water resources.

2. Environmental Footprint

– AI model training, such as for GPT-3 and GPT-4, results in high CO₂ emissions.
– The environmental costs of AI systems include both direct energy usage and the broader implications of their deployment.

3. Potential for Efficiency Gains

– AI-driven solutions, like Google’s DeepMind, have reduced cooling costs in data centers.
– AI applications are enhancing energy efficiency in telecommunications, transportation, and other industries.

Relevance to IEEE Computer Society’s Mission

This topic aligns with the society’s mission to address critical challenges in technology and engage the community on pressing issues. The environmental impact of AI is particularly relevant as members work towards developing responsible and sustainable technological solutions.

Conclusion

The summary not only raises awareness about the environmental costs of AI but also emphasizes its potential for driving sustainability. By bridging the discussion between environmental challenges and technological opportunities, this post supports the society’s efforts to educate and inspire future leaders in responsible AI development.