Prioritize environmental sustainability in use of AI and data science methods [Comment]

Jay, Caroline; Yu, Yurong; Crawford, Ian; Archer-Nicholls, Scott; James, Philip; Gledson, Ann; Shaddick, Gavin; Haines, Robert; Lannelongue, Loïc; Lines, Emily; Hosking, Scott ORCID: https://orcid.org/0000-0002-3646-3504; Topping, David. 2024 Prioritize environmental sustainability in use of AI and data science methods [Comment]. Nature Geoscience. 3, pp. https://doi.org/10.1038/s41561-023-01369-y

Full text not available from this repository. (Request a copy)

Official URL: https://rdcu.be/dxv7y

Abstract/Summary

Artificial Intelligence (AI) and data science will play a crucial role in improving environmental sustainability, but the energy requirements of these methods will have an increasingly negative effect on the environment without sustainable design and use. Against the backdrop of an implicit assumption that computational resources will continue to increase in availability and reduce in cost, it is rare for researchers to explicitly consider environmental impact when designing or choosing analysis methods. We believe there is an opportunity for the environmental science community to drive a change in approach, optimizing energy usage when conducting their own computational research, and advocating for other research domains to do the same. Considering environmental sustainability in computational research will both accelerate innovation and democratize it: regions most affected by climate change — and where local research could have huge benefits — are less likely to have access to significant computational resources. Making energy efficiency and sustainability a primary consideration will also catalyse innovative methodological approaches to scientific research. By aligning these changes with domain-based understanding of the scientific need, we can set standards for best practice in a strategic way.

Item Type:	Publication - Article
Digital Object Identifier (DOI):	https://doi.org/10.1038/s41561-023-01369-y
ISSN:	1752-0894
Date made live:	01 Feb 2024 10:03 +0 (UTC)
URI:	https://nora.nerc.ac.uk/id/eprint/536826

Item Type:

Publication - Article

Digital Object Identifier (DOI):

https://doi.org/10.1038/s41561-023-01369-y

ISSN:

1752-0894

Date made live:

01 Feb 2024 10:03 +0 (UTC)

URI:

https://nora.nerc.ac.uk/id/eprint/536826