Carbajo-Núñez Martín ,
Artificial intelligence and sustainability, (9.04.2025), (34),
in
Blog: www.cssr.news, 34-EN (2025) p. 2
.
The rapid development of digital technologies and Artificial Intelligence (AI) offers enormous possibilities for progress while also posing challenges to sustainability, as it can have adverse impacts on both the physical and social ecosystems. On the social realm, information sources are increasingly narrowing, “fostering a single approach, developed on the basis of an algorithm.”[1]
Environmental challenges
On the environmental level, AI has a considerable ecological footprint, largely due to the high ecological costs associated with the extraction of rare earth elements, which are essential for producing its technological components. The scarcity of these materials requires complex, slow, and costly extraction processes that significantly contribute to environmental degradation. Although more sustainable alternatives are being explored, these are still in the early stages of development.
Currently, China produces about 70% and processes more than 80% of the world’s rare earth elements. The European Commission has identified 17 of these rare metals, including antimony, beryllium, germanium, graphite, lutetium, promethium, tantalum, vanadium, and tungsten. The extreme scarcity of some of these materials necessitates the removal of vast amounts of earth to extract them. An extreme case is the extraction of one kilogram of lutetium, which typically requires the removal of 200 tons of rock.[2]
AI consumes also large amounts of water. It is estimated that, to respond to 20 to 50 questions, ChatGPT-3 uses about half a liter of water for cooling purposes. By 2027, AI's total water consumption could range between 4.2 and 6.6 billion cubic meters, more than four times the current water consumption of Denmark.
AI also requires enormous amounts of electricity. It is estimated that a single search on ChatGPT can consume between three and ten times more energy than a search on Google.[3] According to the International Energy Agency's 2024 report, the nearly 8,000 data centers worldwide (33% of them in the USA) will consume over 1,000 terawatt-hours (TWh) by 2026, an amount equivalent to Japan’s current electricity consumption. This rising energy demand is contributing to higher greenhouse gas emissions and forcing some countries to regulate AI usage. For example, Singapore had to temporarily restrict the construction of new data centers in 2020, after these facilities accounted for 7% of the nation's total electricity consumption.
New opportunites
On the positive side, AI offers new opportunities for promoting ecosystem sustainability. It can play a crucial role in monitoring methane emissions, a major contributor to the greenhouse effect, and in addressing other facets of the ongoing ecological crisis. Additionally, AI can predict energy consumption patterns, enabling more efficient responses, while optimizing the performance of existing power grids and enhancing renewable energy production. Beyond energy, AI can also improve the management of water resources, waste, and recycling processes. It also provides valuable tools for analyzing environmental data, such as monitoring deforestation, evaluating plastic pollution in oceans, and predicting potential natural disasters.
Too much power in a few hands
However, the greatest risk lies in the fact that AI is largely controlled by a small number of profit-driven multinational corporations. This concentration of power in the hands of a few threatens democratic life, the sustainability of ecosystems, and integral human development.
Martín Carbajo-Núñez, ofm
Blog Alfonsiana: ("Inteligencia artificial y sostenibilidad", online ); Blog cssr: Italiano (Online), English (Online); Español (Online)
[1] Francis, “Message for the 58 th World Communications Day” (Jan. 24, 2024), [ WCD], in OR 19 (Jan. 24, 2024) 8.
[2] Kettmajer Michele, “Il potere collettivo dei dati digitali,” in Rivista di Scienze dell’Educazione 62/1 (2024) 16-36.
[3] In Internet: https://www.europapress.es/portaltic/sector/noticia-consulta-chatgpt-consume-tres-veces-mas-energia-buscador-google-20230728164651.html?utm_source=chatgpt.com (Access: Feb 1, 2025).
(Attached file) |
