Climate and AI
AI technologies and techniques are transforming climate action, from tracking deforestation to optimising logistics. At the same time, the Generative AI boom means that the energy requirements of training and deploying models may have an increasing climate impact. As AI becomes deeply embedded in daily life, how can we balance the benefits and the climate impact of AI?
The Conversation
“AI is accelerating the climate crisis… Once we have transparency, we can start legislating.”
Dr Sasha Luccioni, Hugging Face
“The most urgent need in this context is not to have more powerful AI but to become smarter at where and how we use AI.”
Lambert Hogenhout, OICT
“[…] expect the proliferation of AI technology, and the data centers necessary to feed it, to drive an increase in power demand the likes of which hasn’t been seen in a generation.”
Goldman Sachs, Top of Mind Report, June 2024
“As we further integrate AI into our products, reducing emissions may be challenging due to increasing energy demands from the greater intensity of AI compute…”
Google Environmental Report, 2024
“AI is supposed to make us more efficient – but it could mean we waste more energy”
Felippa Amanta, University of Oxford
Summary
AI technologies can be seen as both a tool for climate action and a growing environmental concern.
With some uses of AI, people are using innovative approaches to track deforestation, optimise energy use and help track conservation projects. The potential for using AI to fight the climate crisis is huge but not yet fully realised.
Yet, the main way in which people are using AI is via Generative AI models like GPT-4. Training these models emits significant amounts of CO2, with ongoing use adding to energy demands.
While the energy cost of Generative AI is heavy, that’s not the only concern. Data centres, which are crucial to the use of AI technologies, require large amounts of water for cooling, and the hardware that AI depends on requires mining for Rare Earth Elements, which can harm ecosystems.
Efforts to reduce AI’s footprint include building smaller, more energy-efficient models and developing more environmentally friendly green data centres. While these efforts will help, when it comes to how people use Generative AI – Chat GPT, transparency remains limited.
While the benefits of AI technologies in general are clear, the environmental costs of Generative AI raise questions about how and when it should be used.
Questions. Answers
Here are some of our most frequently asked questions.
-
The huge amount of energy needed to power the servers and data centres that AI technologies rely on generates heat as a by-product.
To prevent servers from overheating, data centres rely on cooling systems, such as large air conditioning units or cooling towers. These systems require substantial amounts of clean, fresh water.
As a result, the AI industry not only consumes vast amounts of energy, but it also competes for local water resources, impacting ecosystems.
-
The GPUs, servers and devices that power AI technologies aren’t just invisible data in a far-off cloud, they’re physical, tangible products. And just like all other electronics, they are made from raw materials extracted from the Earth.
At the heart of AI’s hardware is silicon, a material used in computer chips. While it’s essential for the tech we rely on daily, the extractive nature of silicon mining has serious environmental consequences. Mining for silicon is energy-intensive in and of itself, and can lead to groundwater pollution and ecosystem disruption, as well as posing health risks to those mining it.
AI also relies on a range of Rare Earth Elements. Metals like Neodymium, dysprosium, lanthanum and cerium are all important to the hardware needs of training and deployment of AI, and all are rare within Earth’s crust. This rarity means that the process of mining them can cause particularly severe environmental damage, including deforestation and soil erosion.
Cobalt is a key element withing lithium-ion batteries, which power many of our smart devices and laptops, as well as the servers in data centres that keep LLMs running. Other specialised components within AI hardware require platinum, palladium, tin, tantalum, aluminium and gold – sourced from South Africa, Zimbabwe, Rwanda, Indonesia, Australia and China. Mercury and cyanide are known to be used within mining these metals, which seep into surrounding ecosystems.
-
In Scotland, certain public bodies are currently required to report on their carbon emissions and action taken on the climate crisis. Part of this reporting is dedicated to the wider climate impacts of their digital services, including all “upstream” and “downstream” carbon impacts.
At present, public bodies report only on the impact of digital services generally, making no reference to AI technologies specifically.
With the impact of AI technologies potentially being greater than other digital services, making a differentiation within the reporting could be an opportunity to increase transparency around the impacts of AI.
-
According to various estimates, a single ChatGPT query uses anywhere from 10 to 60 times more energy than a Google search. This difference stems from the fact that AI models require more computational power to generate responses compared to traditional search algorithms. That’s why it’s so important to use Gen AI only when absolutely necessary.
-
When it comes to the climate, AI is a double-edged tool. An environmentally friendly approach could include using AI thoughtfully for tasks where it add real value, rather than for entertainment or to cut corners. Supporting responsible AI regulation and advocating for transparency from AI companies about their environmental impact will help everyone make better informed decisions.
What’s happening in Scotland?
What’s happening in the rest of the world?
Featured Resources
