Generative AI and the Environment: Balancing Innovation with Sustainability
The recent rise of generative artificial intelligence (AI) has created uncertainty among much of the public as questions have arisen about what the technology’s impact on humanity will be. The effects of AI’s usage are expected to have a massive impact on a wide range of societal aspects, one of which is the environment.
AI relies on data centers in order to function, and these centers utilize vast quantities of electricity, which in turn can generate carbon dioxide. On top of the emissions, the process of cooling these data centers takes in substantial water and evaporates it.
“You see new trends coming and I think the energy usage that’s going towards [AI] right now is pretty stunning,” Dargan Frierson, UW associate professor of atmospheric and climate science, said. “The increase in that has been way more rapid than I was imagining would happen and I think that’s true for many many folks.” AI’s environmental impact revolves largely around how electricity is generated for the data centers to utilize. For example, AI data centers that rely on sustainably produced electricity will have a significantly different environmental impact than ones that rely on electricity generated via carbon emissions.
“[Companies are] trying to build these big data centers in places with a lot of renewables,” Vikram Iyer, UW assistant professor of computer science and engineering, said. “And in some cases that also places a big strain on the power grid.”
AI can assess an enormous volume of information, which can have positive implications for understanding and responding to the impacts of climate change, according to Léonard Boussioux, UW professor of information systems and operations management at the Foster School of Business.
AI can identify areas where garbage is prominently located in the ocean so that it can be cleaned up more efficiently, analyze icebergs to determine how much they are depleting, and it can guide drones to drop seeds to rapidly and efficiently plant trees.
The intersection between the fields of AI and sustainability is explored at the Computing for the Environment (CS4Env) initiative at UW.
“The Computing for the Environment Initiative is something that we started a couple of years ago to try to get people from across UW collaborating on environment related research projects,” Iyer, who is a director of CS4Env, said.
Students have also been working on using AI as a way to create a more sustainable world. Computing for Environmental and Social Advocacy (CESA) has directly worked at the intersection of AI and the environment.
“Our projects include integrating [machine learning] algorithms into conservation data analytics and collection, modeling pollution in disadvantaged communities, and studying how attention mechanisms in large language models can lead to innovative solutions to climate change,” Shreyan Mayukh Mitra, a second-year student studying computer science and president of CESA, said in an email.
However, not everyone is using AI sustainably despite attempts to use it to help the planet.
“In addition to finding better environmental solutions, there’s also people who are using these new tools to find more fossil fuels in the ground and get [them] out as quickly as possible,” Frierson said.
With Seattle considered one of the largest growing regions for the AI industry, questions regarding the impact it will have on the environment are expected to continue to arise. Boussioux said that through conversations and education, he hopes to continue spreading awareness about AI’s impact.
“Just making people aware that those [AI] models actually consume tons of water and electricity is a very nice beginning. Not enough, but a beginning,” Boussioux said.
“[Companies are] trying to build these big data centers in places with a lot of renewables,” Vikram Iyer, UW assistant professor of computer science and engineering, said. “And in some cases that also places a big strain on the power grid.”
AI can assess an enormous volume of information, which can have positive implications for understanding and responding to the impacts of climate change, according to Léonard Boussioux, UW professor of information systems and operations management at the Foster School of Business.
AI can identify areas where garbage is prominently located in the ocean so that it can be cleaned up more efficiently, analyze icebergs to determine how much they are depleting, and it can guide drones to drop seeds to rapidly and efficiently plant trees.
The intersection between the fields of AI and sustainability is explored at the Computing for the Environment (CS4Env) initiative at UW.
“The Computing for the Environment Initiative is something that we started a couple of years ago to try to get people from across UW collaborating on environment related research projects,” Iyer, who is a director of CS4Env, said.
Students have also been working on using AI as a way to create a more sustainable world. Computing for Environmental and Social Advocacy (CESA) has directly worked at the intersection of AI and the environment.
“Our projects include integrating [machine learning] algorithms into conservation data analytics and collection, modeling pollution in disadvantaged communities, and studying how attention mechanisms in large language models can lead to innovative solutions to climate change,” Shreyan Mayukh Mitra, a second-year student studying computer science and president of CESA, said in an email.
However, not everyone is using AI sustainably despite attempts to use it to help the planet.
“In addition to finding better environmental solutions, there’s also people who are using these new tools to find more fossil fuels in the ground and get [them] out as quickly as possible,” Frierson said.
With Seattle considered one of the largest growing regions for the AI industry, questions regarding the impact it will have on the environment are expected to continue to arise. Boussioux said that through conversations and education, he hopes to continue spreading awareness about AI’s impact.
“Just making people aware that those [AI] models actually consume tons of water and electricity is a very nice beginning. Not enough, but a beginning,” Boussioux said.
International Research Hypothesis Excellence Award
Website Link: researchhypothesis.com
Nomination Link: https://researchhypothesis.com/award-nomination/?ecategory=Awards&rcategory=Awardee
Contact us: contact@researchhypothesis.com
#sciencefather#generativeai
#sustainability
#environment
#aiforgood
#greeninnovation
#sustainabletech
#ecofriendly
#aiethics
#climatechange
#greentech
#aiandtheenvironment
#innovation
#carbonfootprint
#renewableenergy
#aiimpact
#sustainabledevelopment
#cleantech
#environmentalimpact
#responsibleai
#techforsustainability
#researchawards
#researchinnovation
Comments
Post a Comment