Over the past decade, tomato farmers in central India have faced severe droughts, resulting in significant crop losses and directly impacting the livelihoods of local farmers.
Conversely, on the other side of the globe, Silicon Valley-based company ClimateAi has been developing an artificial intelligence platform designed to assess the vulnerability of crops to rising temperatures over the next 20 years.
This innovative tool utilizes climate data, soil information, and water supply specifics for a given region to forecast the viability of cultivation in the coming years.
In 2021, Maharashtra, India, served as one of the first case studies. Farmers in the region accessed ClimateAi’s app, where they could input details about the seeds they planned to cultivate and their desired planting locations.
Using this data, ClimateAi ran simulations and concluded that extreme heat and drought would lead to an approximate 30% reduction in tomato production in the region over the next two decades. The company then advised farmers to adjust their cultivation strategies.
Adaptations
The results proved crucial: tomato producers modified their business strategies, opting for more climate-resilient seed varieties and adjusting the timing of their tomato planting schedules.
Traditionally, finding new cultivation sites was a time-consuming process for farmers affected by climate change. However, according to Himanshu Gupta, CEO and co-founder of ClimateAi, “this can now be done in minutes, significantly reducing costs.”
Gupta, who grew up in India, emphasizes that artificial intelligence is seen as a multiplier of time and efficiency in seeking solutions to climate change.
Enhanced risk assessment for future agriculture is just one of many ways AI technologies are being deployed to address the climate crisis.
AI in Combating Climate Change
For years, researchers have pondered how AI could help understand and address climate change. Now, experts indicate that AI is poised to accelerate a variety of initiatives, from reducing pollution to improving weather modeling.
“Artificial intelligence is a system endowed with robust predictive abilities that can be tremendously useful across various domains, from understanding molecules at a small scale to broader climate systems, aiding us in the fight against climate change,” said Fengqi You, a professor at Cornell University’s School of Engineering.
However, despite the promise of AI, the infrastructure supporting this technology—such as data centers filled with rows of energy-intensive supercomputers—can, in itself, place pressure on the environment.
Technology Accelerating Discovery
Artificial intelligence is a broad term encompassing various digital tools trained to perform a wide range of tasks that previously required human intervention.
What these technologies often have in common is their ability to process and quickly identify connections across vast volumes of diverse data, making AI particularly effective in activities like forecasting and simulation.
Unlike traditional computer programs, AI tools can continue learning over time as new data becomes available or as systems receive additional feedback on the quality of their results.
While scientific discovery used to depend on human capacity to gather, observe, and analyze evidence, computers can now process large datasets, identify patterns, and conduct digital experiments in a fraction of the time it would take humans.
How Is AI Being Used?
AI has the capability to identify in real-time which renewable energy sources are available in areas where consumers demand them, optimizing the match between consumer demand and renewable energy supply.
Additionally, in other fields, AI is being employed to research materials capable of effectively recapturing carbon from the atmosphere and to model and predict large-scale floods, helping local government agencies better prepare for and respond to potential emergencies.
The Cool Down, a media company dedicated to helping consumers better understand the climate crisis and its potential solutions, plans to launch an AI tool in early 2024. This tool will answer user questions about adopting a more sustainable lifestyle, explained Anna Robertson, co-founder and head of content and partnerships at the company.
The tool will use data from the website to identify the most sought-after climate topics among consumers, answering questions like: “What can I do with my old jeans?” or “I want to switch my laundry detergent—where should I start?”
“Part of the challenge is that the problem itself has become so overwhelming and often dominated by doom and gloom rather than focusing on the solutions within our reach,” Robertson told CNN. “Our goal is to make it easier for people to make more sustainable choices.”
Addressing the Environmental Costs of AI
A significant downside to this growing computational power is that running AI models consumes a substantial amount of energy. Many data centers operate in regions still heavily reliant on fossil fuels, warned You of Cornell.
Furthermore, data centers often require water for cooling—a resource increasingly scarce in some areas where these facilities are located, including the western United States.
While the amount of energy used to power AI is currently relatively small compared to energy consumption by transportation or buildings, You noted that this consumption is growing rapidly and requires careful attention before it becomes exponential.
A study conducted in October by Dutch researcher Alex de Vries estimated that, in a “worst-case scenario,” Google’s AI systems could eventually consume as much electricity annually as the entire country of Ireland, assuming widespread adoption of current AI hardware and software.
Developers are therefore advised “not only to focus on optimizing AI but also to critically consider the need to use AI in the first place,” as not all applications benefit equally from AI, and the benefits do not always outweigh the costs, the study concluded.
Sustainable Initiatives for Data Centers
Some data center operators are beginning to address these concerns.
Amazon Web Services (AWS), the cloud computing division of the e-commerce giant, has pledged to be “water positive” by 2030, meaning “returning more water to the communities where our data center infrastructure is located than we withdraw,” CEO Adam Selipsky said in an interview with CNN in October.
In regions such as Oregon, where droughts have worsened in recent years, AWS provides water used for cooling its data centers to local farmers for irrigation at no cost.
Companies building and operating data centers for AI workloads can also consider strategically locating these centers in areas where natural resource usage is minimized, You observed.
For instance, if data centers are established in colder regions of the world, there would be less demand for water for cooling. Scandinavia has been a popular choice for such data centers due to its availability of renewable energy sources.
Regulatory and Global Considerations
Lawmakers worldwide, who have recently turned their attention to regulating AI, should consider both the potential benefits of this technology in combating climate change and its environmental impact when drafting regulations, You added.
“Regulators, decision-makers, and policymakers really need to think about this when considering the growth of the [AI] industry,” You emphasized. “The growth of the industry isn’t just about software, tools, and so on, but also about how these data centers are operated.”
Technology experts also stress the importance of making AI accessible in low-income nations, particularly those in the Global South that are on the frontlines of the climate crisis yet contribute less to global pollution. Gupta hopes to address this issue as ClimateAi expands its program.
“When it comes to applying AI to climate change, I believe we’re just scratching the surface of its potential—both in terms of the impact it can have for businesses and the impact it can have on the world at large,” Gupta said.
