As AI Grows, Data Centres Face Soaring Energy Demand
AI models consume a lot of energy, increasing energy demand from data centres.
However, it is being used in transport, utilities, agriculture, and green innovation to reduce carbon emissions and mitigate the climate crisis.
Its environmental impact will hinge on how it is developed and used in the future.
This is the era of artificial intelligence (AI), which is changing the tone and tenor of the tech industry. But AI also has a downside: It has a reputation for being energy-intensive and carbon-heavy.
Training a single AI model is equivalent to the electricity consumption of more than 100 US homes per year. As AI grows rapidly over the next decade or two, energy demand—particularly in data centres—is expected to double by 2030.
To cap it all, most of this extra energy will come from unsustainable sources like fossil fuels, mounting pressure on global climate goals.
READ MORE: Study Finds TikTok's CO2 Emissions Exceed Greece's
Notwithstanding its pitfalls, AI can be a powerful tool in combating climate change. Computers have always fared better than humans since the nineteenth century; they can efficiently calculate and analyse enormous amounts of data, spot patterns, and provide immediate solutions in ways humans cannot. For instance, it can process images from fire-prone areas to detect wildfires much earlier than humans could sense or predict.
Despite these benefits, it is not quite sure whether AI's climate solutions outweigh its environmental costs. A large number of AI developers do not track emissions in their systems, making comparisons more challenging against those who help avoid them.
Some real-world examples where AI is used to cut emissions include: 1) Waymo's driverless electric cars reducing transport emissions, and 2) application in ships and trains to optimise fuel use. For example, a Canadian startup has developed an app for loco pilots to save diesel, and robots clean ship hulls to reduce drag and fuel use.
ALSO READ: NVIDIA-Powered AI Models Cut Ship Fuel Use and Emissions
In utility companies, AI is being used to manage risks from climate-related events. Fire-detection AI cameras are now helping such firms supervise large areas, avoiding the need for humans to monitor them 24/7. Also, AI analyses satellite images to determine where overgrown trees might come in contact with power lines, circumventing frequent power disruptions and saving a lot of maintenance costs.
Now, coming to agriculture, AI tools are used to speed up the development of climate-resilient crops. And, in apiculture, it is used to monitor and care for honeybee colonies that are facing a huge threat from climate stress.
Eventually, its application has also extended to developing green and climate technologies at lower cost. For example, Kaio Labs, a climate startup, is using AI to make sustainable fuels from carbon dioxide (CO2) in cheaper ways, where, by and large, producing sustainable fuel is a costly affair.
Though AI has many advantages and uses in today's world, the onus is on its developers; their approach will determine whether new AI models have an environmental impact and harm our planet or not.
Are you looking to scale up sustainability in your business operations? Explore our ESG Marketplace, where you can find more than 650 ESG and sustainability providers offering an array of services.
Follow KnowESG's Environment News for regular news and views.
Check out our latest ESG Event updates
Source: Bloomberg
