AI for a Better Planet

By using artificial intelligence, researchers are gathering good information on the sources of greenhouse gas emissions, which is essential for responding to climate change. (Photograph by NONGASIMO/Shutterstock.com)

Artificial intelligence (AI) is everywhere. From booking a cab and checking GPS-enabled maps for traffic diversions to choosing an outfit based on weather predictions, AI facilitates everyday decision-making. But, if AI can simplify our daily lives, can it not be used for greater public good?

This was the question the panelists at “AI for Good: Climate of Change,” a U.S. Embassy New Delhi-sponsored session at the Jaipur Literature Festival (JLF) 2024, attempted to answer. Among the panelists was David Sandalow, an inaugural fellow at the Center on Global Energy Policy, Columbia University, who has previously served in senior positions at the White House, the U.S. Department of State and the U.S. Department of Energy.

Investments in AI across the world benefit social wellbeing in areas like medicine, environmental sustainability, education and public welfare. For example, AI helps scientists process information from NASA satellite images, revealing patterns in human activity over a long period of time.

Experts are now working to improve these technologies, with the aim of deploying them on a larger scale, with more advanced processing power. Their goal is to find solutions to pressing global problems like the climate crisis.

AI and climate

“AI can, broadly speaking, do three things with large data sets,” explained Sandalow during the session at JLF. “It can predict, optimize and simulate. If you apply that framework, there are so many opportunities for using AI for good, particularly in the area of climate change.”

Sandalow led the team that published a report titled, “Artificial Intelligence for Climate Change Mitigation Roadmap,” which studies the advantages and barriers of using AI to mitigate climate change. The report, released in December 2023 by the Innovation for Cool Earth Forum, notes that concentrations of heat-trapping gasses in the atmosphere are at their highest and are changing Earth’s climate. “Good information on the sources of greenhouse gas (GHG) emissions is essential for responding to climate change,” says the report. “AI is helping to significantly improve such information by analyzing vast amounts of data from Earth-observation satellites, airplanes, drones, land-based monitors, the Internet of Things (IoT), social media and other technologies,” it notes.

(From left) Indian author Anirudh Suri, David Sandalow and British scientist Marcus du Sautoy at the JLF panel discussion. (Photograph courtesy Samsung Galaxy Tab S9 Series Jaipur Literature Festival 2024)

One key area of impact, according to the report, is providing advance warning for extreme weather conditions. “AI is beginning to improve the weather forecasts associated with extreme events, providing accurate, near-term advanced warnings in critical contexts,” the report explains. “This work has made major strides in the past two years and could ultimately transform climate adaptation responses. Some of the most crucial areas for applying this AI-enabled ‘nowcasting’ (forecasting within six hours) capability are extreme precipitation and extreme wind speeds. Additionally, research is ongoing to predict extreme heat over timescales of days to weeks,” it says.

Sandalow believes the ability of AI to find different solutions through permutations and combinations of existing data will be crucial in driving an efficient response to the climate change crisis. “When Thomas Edison came up with a new version of the light bulb 150 years ago, he physically took dozens of different kinds of materials and ran electric charges through those materials to find out how much light and heat they would produce,” explained Sandalow during the JLF panel discussion. “Today, AI can simulate millions of these interactions using chemical structural constraints in a second. This allows us to both select from a wider range of options more rapidly and expand the universe of materials we’re testing.”

Looking forward

AI holds promising potential in tackling climate change but it comes with barriers and risks. Some key barriers, according to the Sandalow-led report, would revolve around conflicting datasets from different sources, data access, equitable tech resources and AI literacy.

However, there is a need for “dedicated effort” to ensure AI-led technology is used for public good. “For example, one of the things we recommend in our report,” explained Sandalow, “is that every institution with a role in climate mitigation have an AI office, or at least a senior adviser on AI. This is because AI is transformational technology, and if you care about climate change mitigation, you need to pay attention to it and think about how AI can help reduce greenhouse gas emissions.”

Machine learning tools, a type of AI that uses algorithms to make predictions with minimal human input, can make a big difference in several areas related to climate change mitigation, according to Sandalow. These tools are currently being used in solar and wind farms to help predict electricity generation patterns and facilitate their better integration into electric grids. As Sandalow says, “That is just the beginning of what we can do with AI when it comes to climate change.”

---

Click here to sign up for the free SPAN newsletter: https://bit.ly/SubscribeSPAN

---

---