Peace Department Reflections: AI & Sustainability - A Double Edge Sword
By now, most people have tested the waters of AI in one form or another, whether using it as a springboard for a new activity or weaving it into their day-to-day workflow and lives. AI promises to increase efficiency and augment human potential to tackle the pressing challenges of our time. But its meteoric rise also has real-world implications that exist outside of the cyberverse. Which begs the question: what effect will AI have on sustainable development? Whether AI becomes a tool to promote sustainability or set it back remains to be seen. Opening our eyes to both the dangers and opportunities is critical at this crossroads. Here, we look into the dynamics of AI’s unstoppable wave from a sustainability lens.
AI is its own best advocate. When we prompted ChatGPT to identify ways AI has already contributed to furthering sustainable development, a list of fifteen items organized by SDG popped up within seconds. It included deforestation tracking through AI-powered satellite imagery, wildlife protection via AI-assisted tracking drones, and energy-efficient building designs, among many others. AI can play a significant role in decarbonizing economies, reducing pollution, preserving biodiversity, increasing energy efficiency, protecting the oceans, and ensuring humanity operates within planetary boundaries. With its ability to develop models to track biodiversity, predict extreme weather patterns, optimize crop yield, and expand education, AI has the potential to accelerate new and existing sustainable development solutions faster than ever before. Clearly, in the words of UN Secretary-General António Guterres, “AI could be a game-changer for the SDGs.”
On the flip side, a growing body of research is focused on the sustainability challenges posed by the rise of AI. Experts caution that the AI explosion and its associated infrastructure demands not only require enormous levels of electricity and water, but will generate unprecedented electronic waste and pressure on existing energy systems. The rate of development of AI is far outstripping the renewable energy sector’s electricity production capacity. The danger of undoing years of green transition progress is real.
The rise of AI will likely spell huge increases in carbon emissions associated with the electricity demands of powering data centers, and embolden the oil and gas industry with tools for boosting fossil fuel extraction. ChatGPT, which we used to generate our SDG list above, produced the same carbon emissions during one “training phase” as three roundtrip flights from San Francisco to New York, with models being retrained thousands of times while under development. With water scarcity on the rise globally, the projected water requirements for AI cooling systems are another cause for alarm: just 10–50 responses in a single AI bot session can consume a half-liter of fresh water.
The tension between AI’s promise and peril plays out in the place-based approach to sustainability, which many, including The Peace Department, agree is the best mechanism for advancing just transition. AI has the potential to decentralize energy systems and develop complete transition plans for entire communities, integrating solutions and stakeholders with heightened efficiency. Implementation, of course, poses its own set of challenges, but the opportunity for using AI to maximize data’s usefulness, identify cross-sectoral synergies, and innovative collaborative approaches is undeniable. Yet, AI raises serious equity concerns. Discriminatory bias and misinformation are baked into data and algorithms. The threat of job displacement looms large. And, the likelihood of increased fuel production and new data center construction taking place in structurally marginalized communities risks exacerbating long-entrenched environmental racism.
In the face of AI’s rapid advancement, it’s important to create conditions and regulations that enable AI’s potential benefits for equity and planetary health to outweigh the risks of backsliding and further environmental degradation. Movers and shakers in business, government, and intergovernmental institutions are joining forces and asking important questions.
Here’s something we all can do to ensure the scales tip in sustainability’s favor:
Evaluate your personal use of AI. From utilizing smaller models to demanding transparency, as with many movements, change starts at home. Understand your use of AI and what sustainable practices can be adopted to minimize its harm.
Join discussions and coalitions on the future of AI and its use for sustainability. Launched in February, the Coalition for Sustainable AI brings together more than 100 partners from tech, academia, government, and civil society with the goal of “unlocking and leveraging its capabilities to address global [environmental] crises, while better understanding and mitigating its environmental footprint.” This represents just one of several initiatives geared towards ensuring that the future of AI is open, inclusive, ethical, safe, and conducive to the sustainability of both People and Planet.
Familiarize yourself with Green AI technologies and companies. From its application to its computing, the dynamic and evolving sector of AI presents an opportunity to bring green AI to the forefront of its use and development.
AI is a rapidly unfolding phenomenon that will inevitably and dramatically change the world as we know it. Working together to set up responsible guardrails today will enable this gamechanging technology to protect our planet and secure a better tomorrow for all.
