Leveraging the Power of AI to Drive Climate Action
This article delves into the transformative role of artificial intelligence (AI) in addressing climate change. Focusing on Google's initiatives and the recent report with Boston Consulting Group (BCG), it explores the potential of AI in reducing greenhouse gas emissions and fostering a sustainable future.
With the impending COP28 conference in Dubai, global leaders strive to build momentum for climate action. The United Nations' Intergovernmental Panel on Climate Change (IPCC) has projected that the world needs to cut emissions by 43% by 2030. The key to achieving this ambitious target may lie in the combined power of artificial intelligence (AI) and collective action.
Recently, Google and the Boston Consulting Group (BCG) released a report revealing that AI could potentially help reduce global greenhouse gas emissions by 5-10% by 2030 — equivalent to the total annual emissions of the European Union. This article explores how Google is driving climate progress using AI and working to manage AI's environmental impact.
Accelerating Climate Action with AI
AI is beginning to address climate challenges in three critical areas: providing people and organizations with better information for sustainable choices, delivering improved predictions to help adapt to climate change, and optimizing climate action for high-impact applications.
Google has been at the forefront of using AI to address these climate challenges. For instance, Google Maps now uses AI for fuel-efficient routing, suggesting routes with fewer hills, less traffic, and constant speeds. Since its launch in October 2021, the fuel-efficient routing feature has prevented more than 2.4 million metric tons of CO2e emissions, equivalent to taking around 500,000 fuel-based cars off the road for a year.
Google's AI is also being applied to predicting climate-related events. Since 2018, Google Research has been developing a flood forecasting initiative that uses advanced AI and geospatial analysis to provide real-time flooding information. The Flood Hub platform, available in over 80 countries, provides forecasts up to seven days in advance for 460 million people.
Moreover, Google Research has collaborated with American Airlines and Breakthrough Energy to develop contrail forecast maps using AI. These maps help pilots choose routes that avoid creating contrails, significantly impacting climate. Post-test flights, pilots reduced contrails by 54%.
Managing the Environmental Impact of AI
While it's crucial to scale these AI applications and discover new ways to use AI for climate action, it's equally important to manage the environmental impact associated with it. The energy use and emissions from AI computing in data centers are challenging to predict as AI is at an inflection point. However, Google is making AI computing more efficient by using proven methods to cut emissions and exploring new ways to increase efficiency.
To minimize the carbon footprint of AI workloads, Google uses practices that can reduce the energy required to train an AI model by up to 100 times and cut associated emissions by up to 1,000 times. Google's Tensor Processing Units v4 is recognized as one of the world's fastest, most efficient, and most sustainable ML infrastructure hubs. Additionally, Google's data centers are designed to maximize efficiency, with an average annual power usage effectiveness (PUE) of 1.10, compared with the industry average of 1.55.
Working Together for a Sustainable Future
Building a sustainable future requires collective action from policymakers, government officials, city planners, business leaders, and individuals. Policymakers, in particular, are central in utilizing AI for climate action and ensuring its sustainable and equitable use. They can accelerate AI's deployment for climate by defining sector priorities, delivering on public sector use cases, and encouraging private sector action.