Prof. Willie E. May, Vice President for Research and Economic Development & Professor of Chemistry. Morgan State University and President, American Association for the Advancement of Science (AAAS)

The pace of scientific progress is accelerating, bringing the potential of artificial intelligence (AI) and other emerging technologies to both sustain and enhance our planet’s health. However, these advancements also pose significant environmental challenges. The need for innovative technologies that can address the pressing issue of climate change while benefitting both people and the planet is essential. How these technologies are developed and by whom really matters. Rapid advancements in technology present both opportunities and challenges. It is crucial to balance the drive for innovation with careful consideration of its long-term environmental and social impacts. How we craft policies to ensure these technologies are being developed sustainably, equitably, and responsibly will be the focus of my discussions.

AI and other emerging technologies have the potential to play a pivotal role in addressing climate change and promoting sustainability. These technologies can be leveraged to develop more efficient energy systems, optimize resource utilization, and enhance our understanding of complex environmental processes. Additionally, AI can support interdisciplinary research to create technologies that are both sustainable and equitable – this collaboration is vital in finding comprehensive solutions that can be adopted with efficacy. At the same time, it is critical that we develop models that utilize small-data techniques, such as few-shot learning and transfer learning, because training and running AI models on large datasets require vast amounts of electricity, leading to increased environmental impacts. Sustainable AI is essential in reducing the environmental impact of AI systems themselves.

But there are challenges:

• Reducing AI’s carbon footprint (renewable energy sources for data centers).
• Ensuring equity – AI systems need to be environmentally sustainable and socially equitable.
• Regulatory frameworks – there need to be robust policies and frameworks that encourage the development and use of sustainable AI tech (this should include standards for energy use, resource conservation, and ethical considerations).
• Global collaboration – global buy-in on shared standards and practices for achieving sustainable AI.

How Do We Get There?

• Hardware and software innovations – continued research and development is required.
• Community engagement and collaboration – promote educational initiatives around the development of sustainable AI for users and developers.
• Accomplishing these goals requires a sustainable AI vision and an unwavering resolve to get there.
• Every nation must answer the call to action.