top of page

Artificial Intelligence and Sustainability : Companions or Contradictions?

Writer's picture: sreeshachakrasreeshachakra

In December of the year 2018, Microsoft had launched its 15 million dollar “AI for Earth” Program, as a forerunner for the leadership position of the AI usage for ethical concerns. Microsoft President Brad Smith wrote in a blog post that they intended to, “put artificial intelligence technology in the hands of individuals and organizations who are working to protect our planet.” In other words, they aimed to use the extensive corpus of knowledge that artificial intelligence possesses and convert it into “actionable intelligence,” finding and funding specialists in critical environmental research areas, such as climate change, water and snow policies, agriculture and biodiversity. 


The Proverbial Silver Bullet : AI as an Accelerator


AI has been metaphorically hailed as a “silver bullet” for addressing the precarious position that the global climate crisis stands at currently. With 2023 recording some of the hottest temperatures around the world to date, it is immensely clear to the researcher as well as the layman that the climate crisis is not just something that can be regaled to light reading or a concern of another time. While AI alone will not solve the problem, its advanced capabilities can allow businesses around the world to navigate and manage complex sustainability issues. 


AI can track, predict and enhance system behaviours, thus swifty analysing thousands of materials to select the one best for creating sustainable products. The World Economic Forum has also reported on recognised methods in which AI has aided in tracking deforestation in companies such as Space Intelligence, making waste management significantly more efficient in companies such as Greyparrot, and help industries decarbonise their operations in mining, oil, metal and gas with companies such as Eugenie.ai


AI has also been proved to help in better resource allocation through pattern recognitions, and healthcare institutes such as the Mayo Clinic have used AI to accelerate the identification of drugs and diseases in patient systems from 45 minutes to seconds, which aid in better management of treatment, conclusively decreasing the amount of experimental or support treatments or medicines that would be used till the correct diagnoses are manually made. Through such pattern recognition software, AI is also able to predict natural disasters fueled by climate change, therefore proving itself useful in the environmental and sustainability context.


The Recoil : Return On Investment Dilemma


The pace of change of AI development is staggering, and it is leaving behind a stench trail of double the global energy consumption of 2022 by the year 2026, and it is not even clear whether our current resources can manage that kind of acceleration. Tech giants such as Google and Microsoft are already immensely missing their sustainability goals as of July 2024, reporting that their carbon emissions have increased stupendously in the last four years alone. CNBC reports that Microsoft’s carbon footprint has been increased 30% when compared to 2020 due to emissions from the construction of their new data centers which support artificial intelligence. BBC reports something similar about Google, implying that the explosive growth of their artificial intelligence resources has driven up their greenhouse gas emissions by 48%. 



Construction of Amazon Mid-Atlantic Region data center in Northern Virginia, Loudoun County, USA as of February 10, 2024 due to the growing demands of generative AI.
Construction of Amazon Mid-Atlantic Region data center in Northern Virginia, Loudoun County, USA as of February 10, 2024 due to the growing demands of generative AI.


Both these companies “aim” to remain as carbon neutral as possible, but that is not the reality of the situation at the moment. Thus, this prompts a familiar conversation around return of investment - is a product that costs the earth thirty to fifty percent more than its older, reliable counterpart a good product? From an isolated economic perspective, that is a seemingly bad investment on the part of humanity. However, AI proponents argue that if artificial intelligence technology can optimize efficiency and facilitate sustainability by around 45%, then the ROI justifies the cost. Many look into the future, predicting that AI may increasingly optimize its own energy usage, and could consequently make the whole energy grid far more efficient.


However, such revolutionary projects and research that could redirect AI energy consumption to power 1.1 million U.S homes by 2026, is currently severely affecting water and other critical environmental factors in the Global South, such as South American, Caribbean and African regions. The mining and production for hardware that supports the gluttonous power consumption of artificial intelligence has led to accelerated soil erosion and pollution in these countries. The electronics that large-scale AI plants house require 800 kgs of raw material for a single 2 kg computer, the United Nations Conference on Trade and Development 2024 Report informs. A prompt given to AI assistants such as ChatGPT consumes 10 times the amount of energy as a Google Search. AI data centers have surged from half a million to almost eight million in the last two years, and experts expect them to only keep growing to match the demand from here on now.


Just Google It! A Call to Action - From Prompting to Searching


When a quarter of humanity is already lacking access to clean water and sanitation, is it really worth it to prompt an AI assistant instead of just googling it and scrolling through a few web pages, like we did in the good old days? Experts also worry that the development of AI technology in various fields may have unprecedented consequences, such as more people buying and using personal vehicles due to the rise in AI-powered self driving technologies, instead of walking or cycling or using public transit - thus undoing the efforts and education that climate activists worked to inculcate within the common man in the last decades. AI can also be easily manipulated to spread misinformation about climate change and sustainability at large, which can be a higher-order effect, downplaying the significance of the entire movement - the Stockholm Resilience Center reports. 


While AI can be a catalyst for scalable environmental change, it is also the current cause of accelerating sustainability and pollution issues, especially in the subaltern. The sustainable future being talked about by proponents is also concentrated within regions and powers who are already removed from the climate crisis by ways of privilege. Is the summarisation of that article that you could have read in ten more minutes really worth the consequences that AI poses to global climate and sustainability issues?


Comments


bottom of page