As the world battles climate change, will the increasingly widespread use of artificial intelligence (AI) be a help or a hindrance? In an article published this week in Natural climate changea team of experts in AI, climate change and public policy, presents a framework for understanding AI’s complex and multifaceted relationship with greenhouse gas emissions, and suggests ways to better align AI on climate change goals.
“AI affects the climate in many ways, both positive and negative, and most of these effects are poorly quantified,” said David Rollick, assistant professor of computer science at McGill University and senior academic member of Mila – Institut québécois de l’IA, co-author of the article. “For example, AI is used to track and reduce deforestation, but AI-based advertising systems are likely making climate change worse by increasing how much people buy.”
The paper divides the impacts of AI on greenhouse gas emissions into three categories: 1) Impacts of computational energy and hardware used to develop, train and run AI algorithms, 2) impacts immediate caused by AI applications – such as optimizing energy use in buildings (which decreases emissions) or accelerating fossil fuel exploration (which increases emissions), and 3) system-level impacts caused by how AI applications affect behavior patterns and society more broadly, such as through advertising systems and self-driving cars.
“Climate change should be a key consideration when developing and evaluating AI technologies,” said Lynn Kaack, assistant professor of computer science and public policy at the Hertie School and lead author of the report. “We find that the easiest impacts to measure are not necessarily the ones with the greatest impacts. It is therefore important to assess the effect of AI on climate holistically. »
The impacts of AI on greenhouse gas emissions – a matter of choice
The authors point to the ability of researchers, engineers, and policymakers to shape the impacts of AI, writing that its “…ultimate effect on the climate is far from predestined, and societal decisions will play an important role in shaping it.” formation of its global impacts”. For example, the paper notes that AI-based autonomous vehicle technologies can help reduce emissions if designed to facilitate public transport, but they can increase emissions if used in personal cars and incentivize people to drive more.
The researchers also note that machine learning expertise is often concentrated among a limited set of actors. This raises potential challenges for the governance and implementation of machine learning in the context of climate change, as it can create or widen the digital divide, or shift power from the public to large private entities in under whom controls the relevant data or intellectual capital. .
“The choices we make implicitly as technologists can matter a lot,” Prof Rolnick said. “Ultimately, AI for Good shouldn’t just be about adding beneficial applications on top of business as usual, it should be about shaping all AI applications to achieve the impact we want to see.”
Reference: Kaack LH, Donti PL, Strubell E, Kamiya G, Creutzig F, Rolnick D. Aligning artificial intelligence with climate change mitigation. Nat Clim Chang. 2022;12(6):518-527. do I:10.1038/s41558-022-01377-7
This article was republished from the following materials. Note: Material may have been edited for length and content. For more information, please contact the quoted source.