It’s impossible to go online these days without hearing about developments in the AI space. It seems like you can’t even Google something now without the AI Overview as your first result—as of this writing, anyway. (Thanks to savvy coders, there is a way around that if you want it.)The use of AI in various professions could lead to job losses, but it could also lead to the creation of new roles that work alongside the technology. AI automates some tedious tasks so workers can be free to focus on other ones, but it’s also only as effective as how it’s been trained. So, you can’t really claim AI as a societal good or evil yet—or maybe ever. This is especially true when it comes to AI’s effects on the environment. So far, we’re seeing lots of evidence that AI is both good and bad for the planet. And we’ve known this for a while.
ENVIRONMENTAL IMPACTS: NEGATIVE
In 2020, Payal Dhar wrote the following in Nature Machine Intelligence:
On the one hand, [AI] can help reduce the effects of the climate crisis, such as in smart grid design, developing low-emission infrastructure, and modelling climate change predictions. On the other hand, AI is itself a significant emitter of carbon. This message reached the attention of a general audience in the latter half of 2019 when researchers at the University of Massachusetts Amherst [analyzed] various natural language processing (NLP) training models available online to estimate the energy cost in kilowatts required to train them. Converting this energy consumption in approximate carbon emissions and electricity costs, the authors estimated that the carbon footprint of training a single big language model is equal to around 300,000 kg of carbon dioxide emissions. This is of the order of 125 round-trip flights between New York and Beijing, a quantification that laypersons can visualize.
PCMag’s Emily Dreibelbis confirms that the excessive consumption didn’t start with new tools such as OpenAI’s ChatGPT—the tech industry’s water consumption was increasing before its launch in November 2022. For example, “Microsoft and Google reported a 34% and 21% spike in water consumption, respectively, in 2022 compared with 2021. The main culprit? Data centers and the water used to keep their temperatures in check,” Dreibelbis writes. “Water is the cheapest method for tech companies to cool their servers, CNBC reports, making it the unexpected bedrock of Big Tech.” Microsoft, Google, Meta, and Amazon are trying to offset their water consumption with strategies such as rainwater collection and recycling municipal wastewater, she shares.
Matt O’Brien and Hannah Fingerhut report for the Associated Press (AP) that OpenAI used water “pulled from the watershed of the Raccoon and Des Moines rivers in central Iowa to cool a powerful supercomputer as it helped teach its AI systems how to mimic human writing. … Few people in Iowa knew about its status as a birthplace of OpenAI’s most advanced large language model, GPT-4, before a top Microsoft executive [mentioned it] in a speech. …”
It’s obvious why Big Tech isn’t advertising its impact on the environment. Mariana Mazzucato, a professor in the economics of innovation and public value at University College London, writes in a piece for The Guardian, “[I]t’s critical that we turn the spotlight on its environmental footprint. Large language models such as ChatGPT are some of the most energy-guzzling technologies of all. … The industry’s environmental impact is a key issue, yet the companies that produce such models have stayed remarkably quiet about the amount of energy they consume—probably because they don’t want to spark our concern.”
AP’s O’Brien and Fingerhut cite Shaolei Ren, a researcher from the University of California–Riverside who is looking into AI’s environmental impact: “Ren’s team estimates ChatGPT gulps up 500 milliliters of water (close to what’s in a 16-ounce water bottle) every time you ask it a series of between 5 to 50 prompts or questions. The range varies depending on where its servers are located and the season. The estimate includes indirect water usage that the companies don’t measure—such as to cool power plants that supply the data centers with electricity.” For example, O’Brien and Fingerhut write, “In some ways, West Des Moines is a relatively efficient place to train a powerful AI system, especially compared to Microsoft’s data centers in Arizona that consume far more water for the same computing demand.” Microsoft and OpenAI both issued statements saying they recognize their high consumption levels and that they’ll be looking into ways to address them, say O’Brien and Fingerhut.
But environmental impact is inevitable. “While there may be some efficiency gains because of AI, we can anticipate a net increase in electricity demand, particularly as the entertainment industry and others develop new and creative uses for AI. This AI-driven increase will likely begin within the next several years, well before the power network has had the time to convert from the current fossil fuel-based system to a low-emissions renewables-dominated one,” Philippe Benoit writes for The Hill. Research from the Brookings Institution finds the following:
The proliferation of tech applications such as generative AI and the expansion of ubiquitous cloud services remain tethered to digital infrastructure and supply chains. This elevates the challenge of stimulating economic growth without comparable increases in emissions—a task that University of Florence Assistant Professor Roberto Verdecchia has said is ‘physically impossible.’ …
Technological innovation is often promoted as a mechanism to decrease carbon emissions, with digital operations and infrastructure in particular cited as ‘cleaner’ advancements from industrial production reliant on fossil fuels. Yet this relationship remains unclear and has lacked a compelling demonstration. Despite their operation in a virtual space, AI and the cloud have considerable tangible effects. They will intensify greenhouse gas emissions, consume increasing amounts of energy, and require larger quantities of natural resources. This emerges, in one form, through rising energy demands.
IMPACT ON MISINFORMATION
There’s another aspect of AI’s environmental impact to consider when using the technology: climate change misinformation.
Boston University’s Center on Emerging and Infectious Diseases (CEID) posted about a recent university-hosted talk by Melissa Fleming, United Nations undersecretary-general for global communications. “Fleming’s lecture discussed the challenges and opportunities of AI in disseminating accurate global public health communication, particularly in the areas of vaccines, climate change, and well being of women and girls,” CEID states. Fleming said she worries that AI “content can be produced at scale and [can be] far more easily personalized and targeted.” CEID continues:
‘Since Elon Musk took over X (formerly known as Twitter), all of the climate deniers are back, and [the platform] has become a space for all kinds of climate disinformation,’ said Fleming. ‘There is a connection that people in the anti-vaccine sphere are now shifting to the climate change denial sphere.’ …
As is the case with the vaccine sphere, AI makes misinformation about climate change easier and cheaper to produce, which in turn reaches a wider audience and diminishes trust in public institutions.
Instead of trying to deny every false claim about climate change, Fleming and her team focus on promoting accurate and fact-based information in order to give institutions like the UN more agency. This approach also creates more positive interactions with communities on social media platforms.
Michael Khoo, climate disinformation program manager at Friends of the Earth, has a similar take: “In the wrong hands, AI could forever undermine factual climate discourse with its ability to tailor-make stories, arguments, even realistic images. By scraping social media posts and other digital activity, AI has the potential to create billions of pieces of disinformation and then personalize them and disseminate them—which could make it extremely difficult to tell fact from fiction. This could not only hinder fact-based climate action, but it could also pose serious danger around extreme weather events—when clear and accurate information is critical.” Khoo calls for regulations like those for pharmaceuticals and car safety to help combat this problem, as well as for transparency from AI companies about how their algorithms work.
POLITICO’s Arianna Skibell reports on an AI model that is being trained specifically to fight misleading information about climate change, which “could help repurpose AI from an agent of misinformation to a policing force. But it’s a tall order: According to one third-party tracker, fake news websites that use AI have ballooned in recent months. Compounding the problem, social media giants such as Facebook and X have weakened their misinformation safeguards.” In addition, Skibell asserts, “Misleading claims about climate change are particularly hard to weed out. Fossil fuel proponents and conservative groups have spent decades sowing public confusion around climate science by using false, misleading or cherry-picked research.”
ENVIRONMENTAL IMPACTS: POSITIVE
Let’s circle back to the idea that AI can actually be a tool for alleviating the effects of and even helping to combat climate change. The United Nations’ UN News issued an explainer stating, “As extreme weather events unfold with more frequency and intensity, AI can help communities around the world to better brace for climate disasters.” For example, areas susceptible to landslides can use AI-based mapping to help local officials ensure resident safety, and urban areas with poor air quality can use AI to track pollution and monitor dangerous levels. Other key aspects of dealing with climate change via AI include the following:
Leveraging AI’s benefits is also part of the UN Secretary-General’s groundbreaking Early Warnings for All initiative. Launched earlier this year, … its action plan aims to ensure everyone on Earth is protected from hazardous weather, water or climate events through early warning systems by the end of 2027. …
In terms of realizing the global goal for affordable and clean energy for all by 2030 … , AI can optimize grids and increase the efficiency of renewable sources. Predictive maintenance using AI can also reduce downtime in energy production. That can mean reducing the planet’s carbon footprint.
The World Economic Forum posted “9 Ways AI Is Helping Tackle Climate Change.” Read the article for all of the details, but in brief, they are:
- AI has been trained to measure changes in icebergs 10,000 times faster than a human could do it.
- AI, satellite images and ecology expertise are also being used to map the impact of deforestation on the climate crisis.
- In Africa, AI is being used in a United Nations project to help communities vulnerable to climate change in Burundi, Chad and Sudan.
- Another AI system is helping to tackle climate change by making waste management more efficient.
- In the Netherlands, an environmental organization called The Ocean Cleanup is using AI and other technologies to help clear plastic pollution from the ocean.
- In São Paulo, Brazil, a company called Sipremo is using AI to predict where and when climate disasters will occur, and what type of climate disasters they will be.
- Google DeepMind, Google’s AI research laboratory, says it is applying AI to help fight climate change in a number of areas.
- AI is being used to help companies in the metal and mining, oil, and gas industries to decarbonize their operations.
- AI-powered computers are pairing up with drones in Brazil to reforest the hills around the coastal city of Rio de Janeiro. …
Here’s an outside-the-box way that AI can help with climate change. Earth.com’s Eric Ralls profiles an AI tool called SLEAP, writing, “[S]cientists are leveraging AI to optimize plants’ natural ability to draw carbon dioxide out of the atmosphere, a crucial step in combating climate change and limiting global temperature rise.”
For Earth.Org, Chris Hocknell boils down the importance of AI to this: “Ultimately, climate change is a technical problem. As with any problem, to solve it, you must first understand it. AI will grant us a far better understanding of exactly what is causing climate change, and how.” Hocknell calls headlines that criticize AI’s energy usage “short-sighted,” because AI is integral to transitioning to green energy in the long run. He continues:
While AI’s energy and water usage are cause for attention, we cannot allow that to overshadow the enormous benefits the technology will bring. Indeed, it is the one technology that allows our other technologies to reach their full potential. …
To have any real chance of mitigating climate change, we must take a long-term, strategic view. Short-term emission reductions are essential. But we mustn’t cow in fear at scary-looking emission profiles of new technologies, while ignoring their enormous emission-abating potential.
DUALITY
In the Journal of Information, Communication and Ethics in Society, Anders Nordgren analyzes “ethical issues raised by the dual role of artificial intelligence (AI) in relation to climate change, that is, AI as a contributor to climate change and AI as a contributor to fighting climate change.” Research such as this shows that AI is a double-edged sword. We as a society need to be aware of its environmental impact (“substantial emissions,” writes Nordgren) while also understanding the role it can play in how climate change threatens humanity (“mitigation and adaptation,” per Nordgren). And it’s important for the AI stakeholders to acknowledge and understand this dichotomy too if they want to move the needle on climate change. “Given this dual role of AI, ethical considerations by AI companies and governments are of vital importance,” Nordgren concludes.