The world is focusing on accelerating the zero-carbon transition. Applications in artificial intelligence (AI), data science, have been created with the goal of salvaging the environment. AI uses machine learning to spot trends and find patterns in data across different sectors. Experts say that AI can be costly in fighting climate change.
Researchers believe that AI should be used to solve climate problems. However, it is important to first understand the growing carbon footprint. AI can be used for climate prediction, smarter decisions in decarbonizing industries, and allocating renewable energy.
AI’s potential use as a tool to combat climate change is being challenged by ethical concerns. Machine learning applications, which use data for predictions and decisions, are causing more ethical concerns. Data usage could lead to bias, discrimination, and intentional misuse.
AI is a major emitter of carbon. Researchers at the University of Massachusetts Amherst studied several online models for natural language processing to understand the energy costs involved in training them.
A Nature report stated that training one big language model would result in a carbon footprint of 300,000 kilograms of carbon dioxide. The total emissions equal almost 125 round trip flights between New York City and Beijing.
Due to the increasing demand for AI, more processing power is required to support larger AI models. Indian Express reported that digital technologies accounted for 1.8 and 6.3% of global emissions in 2020.
Only a few countries have the resources, strengths in research and development, skilled workforce, and wealth to invest in AI. According to the Indian Express report alone, North America and East Asia are responsible for three quarters of the world’s private investments in AI, patents, and publications. The report stated that climate change could worsen this inequity.
Experts believe that AI must be developed and applied responsibly in accordance with environmental sustainability principles.
It is important that AI is not just a tool, but a resource seeker.
There are many steps that can be taken to ensure cleaner AI practices. Industry processes that promote green AI development could be awarded green AI certifications.
It is important to set up a framework and guidelines for organisations and companies to support environmentally-friendly AI practices, Gabriela Prata Dias, head of the Copenhagen Centre on Energy Efficiency, and Xiao Wang, programme officer, told Nature.
Governments should also work to establish legal frameworks and legislations that address transparency and sustainability in AI-related development.
Governments should reevaluate their technology-based growth priorities in light of AI’s climate impacts. Deepika Sandeep, an AI scientist at Bharat Lighting & Power in Bengaluru, said to Nature that deep-learning technologies can only be used with a thoughtful approach. Sandeep stated that not every problem can be solved by machine learning.
In November 2021, UNESCO adopted the Recommendation on Ethics of Artificial Intelligence. This urged global actors to reduce the environmental impacts of AI systems, including its carbon footprint.
Microsoft, Amazon, Google parent Alphabet, and Meta have announced net zero policies and initiatives. Although these initiatives are positive, they are only the beginning.
(Edited By : Shoma Bhattacharjee)
