Can AI Help Build a More Sustainable World?

As technology advances so quickly, it's becoming increasingly difficult to define what future we're building. Microsoft Research's Doug Burger, together with experts from various fields, examines the most challenging problems AI poses for technology, policy, and business, emphasizing that a shared understanding is needed for the AI transition to have a net positive effect.

Amy Luers and Ishai Menache view the relationship between AI and climate change as two-sided. datacenters account for a small but rapidly growing share of global emissions, increasing the burden on regional infrastructure, while at the same time AI can be a powerful tool for optimizing complex systems and accelerating climate solutions.

The key is data, not hype. Rather than viewing AI purely as a cause of or solution to climate problems, we need to carefully distinguish its actual impact from its potential applications and design accordingly.

• AI is a double-edged technology that can both increase and reduce emissions.
• Data centers account for a small share of global emissions, but their growth rate and impact on regional infrastructure cannot be ignored.
• Tools like OptiGuide and OptiMind show a direction for designing systems like supply chains more efficiently by enabling complex optimization algorithms to be handled through natural language.
• Ultimately, building a sustainable future requires an approach that calibrates to actual figures and operational realities, rather than being unconditionally optimistic or wary about AI.