September 21, 2024
There is great hope that artificial intelligence (AI) can help tackle some of the world’s biggest environmental emergencies. Among other things, the technology is already getting used to map the destructive dredging of sand and chart emissions of methane, a powerful greenhouse gas.
But when it comes to the environment, there is a downside to the explosion of AI and its accompanying infrastructure, according to a growing body of research. The growing data centers that house AI servers produce electronic waste. They are big consumers of water, which is becoming scarce in many places. They rely on critical minerals and rare elements, which are often unsustainably mined. And they use massive amounts of electricity, fueling the emission of planet-warming greenhouse gases.
“There’s still a lot we don’t know about the environmental impact of AI, but some of the data we do have is concerning,” says Golestan (Sally) Radwan, the Chief Digital Officer of the United Nations Environment Program (UNEP). . “We need to make sure the net effect of AI on the planet is positive before we deploy the technology at scale.”
This week UNEP has a issue note which examines AI’s environmental footprint and considers how the technology can be rolled out sustainably. This follows a major UNEP report, Navigate through New Horizonswhich also explored AI’s promise and perils. Here’s what those publications found.
First, what is AI?
AI is an umbrella term for a group of technologies that can process information and, at least superficially, mimic human thought. Rudimentary forms of AI have existed since the 1950s. But the technology has evolved at a rapid pace in recent years, in part due to advances in computing power and the explosion of data, which is essential for training AI models.
Why are people excited about the potential of AI when it comes to the environment?
The great advantage of AI is that it can detect patterns in data, such as anomalies and similarities, and use historical knowledge to accurately predict future outcomes. This could make AI invaluable for monitoring the environment, and helping governments, businesses and individuals make more planet-friendly choices. It can also improve efficiency. UNEP, for example, uses AI to detect when venting oil and gas installations methanea greenhouse gas that drives climate change.
Advances like these raise hopes that AI can help the world to at least some aspects of the triple planetary crisis of climate change, loss of nature and biodiversityand pollution and waste.
So how is AI problematic for the environment?
Most large-scale AI deployments are hosted in data centers, including those operated by cloud service providers. These data centers can take a heavy toll on the planet. The electronics they house rely on a staggering amount of wheat: the making of a 2 kg computer need 800 kg of raw materials. The microchips that power AI also require rare earth elements, which are often mined in environmentally destructive ways, noted Navigate through New Horizons.
The second problem is that data centers produce electronic waste, which often contains hazardous substances, such as mercury and lead.
Third, data centers use water during construction and, once operational, to cool electrical components. Globally, AI-related infrastructure could soon consume six times more water than Denmark, a country of 6 million according to one estimate. It is a problem when a quarter of humanity already do not have access to clean water and sanitation.
Finally, to power their complex electronics, data centers that house AI technology require a lot of energy, which in most places still comes from burning fossil fuels, which produce planet-warming greenhouse gases. A request made by ChatGPT, an AI-based virtual assistant, consumes 10 times the electricity from a Google Search, reports the International Energy Agency. While global data is sparse, the agency estimates that the rise of AI in Ireland’s tech hub could see data centers account for almost 35 percent of the country’s energy consumption by 2026.
Driven in part by the explosion of AI, the number of data centers has risen to 8 million from 500,000 in 2012, and experts expect the technology’s demands on the planet to continue to grow.
Some have said that when it comes to the environment, AI is a wildcard. Why is that?
We have a decent handle on what the environmental impacts of data centers can be. But it is impossible to predict how AI-based applications themselves will affect the planet. Some experts worry that they could have unintended consequences. For example, the development of AI-powered self-driving cars could cause more people to drive instead of cycling or taking public transport, pushing up greenhouse gas emissions. Then there are what experts call higher-order effects. AI can be, for example used to generate misinformation on climate change, which downplays the threat in the eyes of the public.
Is anyone doing anything about the environmental impacts of AI?
More than 190 countries have adopted a series of non-binding recommendations on the ethical use of AI, covering the environment. As well, both the European Union and the United States of America introduced legislation to temper the environmental impact of AI. But such policies are few and far between, says Radwan.
“Governments rush to develop national AI strategies, but rarely consider the environment and sustainability. The lack of environmental safeguards is no less dangerous than the lack of other AI-related safeguards.”
How can the world contain the environmental fallout of AI?
In the new edition note, UNEP recommends five main things. First, countries can establish standardized procedures to measure the environmental impact of AI; at the moment there is a lack of reliable information on the subject. Second, with support from UNEP, governments can develop regulations requiring companies to disclose the direct environmental consequences of AI-based products and services. Third, tech companies can make AI algorithms more efficient, reduce their demand for energy, while recycling water and reusing components where possible. Fourth, countries can encourage companies to green their data centers, including by using renewable energy and offsetting their carbon emissions. Finally, countries can weave their AI-related policies into their broader environmental regulations.
UNEP is focused on helping the world better navigate the environmental challenges of tomorrow. To do this, we marched we work on strategic foresightscan the horizon for emerging threats to the planet. This process culminated in the development of Navigating New Horizons – A global outlook report on planetary health and human well-beingwhich was published earlier this year. Produced in collaboration with the International Science Council, it examined eight global shifts that are accelerating the triple planetary crisis of climate change, loss of nature and biodiversity, and pollution and waste.