As the world struggles with climate change, will increasing use of artificial intelligence (AI) be a help or a hindrance? climate change, and public policy introduces a framework for understanding the complex and multidisciplinary AI relationship with greenhouse gas emissions, and proposes ways to better integrate AI with climate change goals. “AI affects the weather in many ways, good and bad, and many of these effects are not well measured,” said David Rolnick, Assistant Professor of Computer Science at McGill University and a member of the Core Academic Mila – Quebec AI Institute he is the co-authored the paper. “For example, AI is used to track and reduce deforestation, but AI-based marketing programs are likely to make climate change worse by increasing human consumption.”
The paper classifies the effects of AI on greenhouse gas emissions into three categories: 1) Impacts from computer power and computer systems used to develop, train, and use AI algorithms, 2) the immediate effects of AI use – as a whole. the use of energy in buildings (reducing emissions) or accelerating the exploration of fossil fuels (increasing gas emissions), and 3) systemic effects caused by the ways AI systems affect behavior patterns and society at large, such as marketing and stability systems. -driving cars.
“Climate change should be a key consideration when developing and testing AI technology,” said Lynn Kaack, Assistant Professor of Computer Science and Public Policy at Hertie School, and lead author of the report. “We find that those impacts that are easy to measure are not the ones that have the greatest impact. Therefore, exploring the impact of AI on a holistic climate is important.”
The effects of AI on greenhouse gas emissions – a matter of choice
The authors emphasize the ability of researchers, engineers, and policymakers to shape the effects of AI, writing that “… the ultimate impact on climate is not far from predetermined, and social decisions will play a major role in shaping all its effects.” For example, this paper notes that AI-enabled autonomous vehicle technology can help reduce pollution if it is designed to make public transportation easier, but it can increase carbon emissions when used in private cars and result in more people driving.
Researchers also note that machine learning technologies often focus on a limited set of characters. This raises potential challenges to managing and operating a learning machine in the context of climate change, as it may create or expand digital fragmentation, or drain power from large private private companies over who controls relevant data or intelligence.
“The decisions we make obviously as an expert can be very important,” said Prof Rolnick. “Ultimately, AI for Good should not only be about adding profitable applications over the business as usual, it should be about creating all AI applications to achieve the impact we want to see.”
Source Journal Reference: Lynn H. Kaack, Priya L. Donti, Emma Strubell, George Kamiya, Felix Creutzig, David Rolnick. Aligning artificial intelligence with climate change mitigation. Nature Climate Change, 2022; 12 (6): 518 DOI: 10.1038/s41558-022-01377-7
