Accelerating climate action through AI

As I write this, almost 200 nations are meeting in Baku, Azerbaijan, to discuss how to address the perils of climate change. While there are early signs of positive movement, the 29th edition of the Conference of the Parties (COP29) seems to be struggling, especially on climate finance negotiations. Quite fittingly but worryingly, the meeting is happening amidst the backdrop of an expected record level of greenhouse gas emissions this year, contrary to the aims of the Paris Agreement.

One approach that is creating a lot of buzz in accelerating climate action is through generative artificial intelligence (AI). Sans the hype, there are promising new applications that could massively boost climate solutions. Indeed, recent scientific literature suggests that AI could be a game changer, and sky’s the limit to its applications.

For one, future climate projections are based on computer models that require enormous computing power. This is why only wealthy countries can afford to develop and run these models. With AI, it may be possible to accelerate the modeling process while increasing its precision. Hopefully, such AI-based modeling may also become more affordable for developing countries for them to run country-specific climate models to better guide their decision-makers.

AI could expand climate mitigation action (reduction of greenhouse gases) in various ways. According to the MIT Technology Review, AI could improve predictions of energy demand, help discover new materials for more efficient energy storage, make transport routes more efficient, and make buildings more efficient, among other things. A global coalition of academics and AI experts has built the Transient Climate Evolution (TraCE) model at the planetary scale. This AI-powered model uses satellite images and ground data to track worldwide greenhouse gas emissions in ways never done before.

AI applications on climate and disaster resilience are even more intriguing for vulnerable countries like ours. For example, Vonich and Hakim (2024) demonstrated that AI can predict the occurrence of heat waves around three weeks in advance in the western part of the United States. On the other hand, Almikaeel and co-workers (2024) have developed an AI flood model that can predict floods about 12 hours in advance. Lastly, Cano and others (2024) used AI to predict the onset of diarrheal disease in Nepal, Taiwan, and Vietnam using climatic data such as ENSO events. Their work could alert public health officials of an impending health crisis months in advance if further refined.

Admittedly, many of the above applications are in their infant stage and need further fine-tuning and testing. But there is enough evidence to show that there is more to come.

However, just like other technological marvels, there is no free lunch. One of AI’s most cited impacts is its contribution to the rise of greenhouse gases. For example, recent media reports have pointed out that Google emitted about 50 percent more greenhouse gases than in 2019 because of its increased use of AI. Microsoft had a similar, albeit a more minor, increase in emissions. A study by Columbia University warned that emissions from data centers could surpass that of the aviation industry shortly. These increased emissions will have to be balanced by AI’s potential to reduce emissions, as discussed above.

Several ethical issues must also be resolved to prevent AI from being hijacked for nefarious purposes. These issues include protecting privacy and intellectual property, among others. There is also a considerable gap between the “haves” and the “have-nots” nations in terms of using AI. These issues and more should be discussed within and among countries to ensure a more equitable sharing of AI benefits.

As a highly vulnerable country, the Philippines must explore how AI could make us more resilient to climate hazards and long-term climate change. Our policymakers should increase their support to our scientists and developers so that we can optimize the use of AI toward this end. This should include capacity building for our research and development people to accelerate further the integration of AI into our climate action and disaster risk reduction programs. Partnerships between the government, the academe, and civil society with the private sector can unlock hitherto untapped resources.

Whatever happens in COP29 in Baku and in successive COPs, the Philippines must exert all efforts to enhance our people’s coping capacity and resilience to climate change. AI could be one powerful tool in this quest.

—————-

Dr. Rodel Lasco is the executive director of The Oscar M. Lopez Center, a foundation devoted to discovering climate change adaptation solutions (http://www.omlopezcenter.org/).