Artificial Intelligence’s Not-so-Artificial Impact on the Climate
Fury Basso-Davis | September 18, 2024
As Artificial Intelligence’s (AI) popularity rises, so does the temperature. The constant running of and training of AI systems require large amounts of electricity and as a result, carbon dioxide emissions are affecting the climate.
The effect that AI currently has and will have in the future is hard to foresee because there are multiple types of AI, ranging from machine models that help with data, vision programs in self-driving cars and Large Language Models (LLM) such as chatbot like ChatGPT, My AI, Meta AI and many others. When OpenAI trained an LLM called ChatGPT-3, it released 500 tons of carbon dioxide into the atmosphere. However, simpler models produce fewer emissions.
In 2019 Microsoft partnered with ExxonMobil to improve its analyses and enhance operational efficiencies at its Permian Basin acreage, making it the largest-ever oil and gas acreage to use cloud technology. ExxonMobil would use Microsoft's cloud computing platform and gain by using technology that relied on AI which would be optimal for mining operations. Microsoft AI is directly used to add more fossil fuels, releasing greenhouse gasses when burned.
Fossil-fuel extraction is not the only AI application that could be environmentally harmful.
“There’s examples like this across every sector, like forestry, land management, farming,” said Emma Strubell, a computer scientist at Carnegie Mellon University.
For example, consumerism and fast fashion ads are another large factor in fossil fuel emissions. The practice of shoveling specific adds boosts over consumptive action. Fast fashion ads push cheap mass production of clothes to consumers, resulting in more demand for clothes and more emissions the fast fashion industry produces up to eight percent of global emissions. Fast fashion companies produce even more emissions from shipping and cause clothing to build up in landfills. “The development of the next generation of AI tools cannot come at the expense of the health of our planet,” said Massachusetts Senator Edward Markey.
However, there are AI applications that can help minimize climate change and environmental problems such as xView2, a program that implements machine learning models and computer vision with satellite imagery to find buildings damaged in natural disasters. The Defense Innovation Unit, a U.S. Department of Defense Organization launched the program. Assess damaged infrastructure which reduces the danger and saves time for first responders to make quick decisions when it comes to their safety and the safety of civilians.
Another example of what AI can look like while helping the climate is WattTime Climate TRACE which uses satellites and AI to independently track greenhouse gas emissions from all sectors globally. It was founded in 2019 when WattTime and TransitionZero were given a grant to watch emissions that came from power plants using satellites from space.
In 2024, the European Union’s (EU) Artificial Intelligence Act was signed into law, which is the first law of its kind.
“The use of artificial intelligence in the EU will be regulated by the AI Act, the world’s first comprehensive AI law. Find out how it will protect you,” said European Parliament.
The law calls for regulation of AI to better conditions for the development of its use in technology. According to the European Parliament’s website, their strategy is to make sure that the use of artificial intelligence is safe, transparent, traceable, non-discriminatory, and environmentally friendly. They believe that AI systems should be watched and looked at by people rather than more computers to prevent potentially harmful outcomes. On their website, they list different rules for different risk levels, the transparency requirements, supporting innovation and the next steps they want to take.
AI could be a valuable tool when it comes to looking at and solving environmental issues if the AI industry can reduce the negative impact AI has on the planet.
Photo by Vivian Bryan.