Rising sea levels and storm surges increasingly threaten bridges and roads. Extreme heat strains energy grids and leads to the buckling of roads, airport runways, and the warping of train tracks. As these events become more frequent and severe, public assets face increased risk of depreciation and obsolescence. Governments are facing a dual challenge: they must reassess investment strategies to prioritize climate resilience and ensure these strategies contribute to reducing emissions.
A study conducted by the MIT Concrete Sustainability Hub reveals that allocating funds towards climate-resilient construction allows for a return on investment within just two years in areas susceptible to hazards. World Bank research corroborates the economic value of such investments, showing that Beyond these economic returns, resilient infrastructure serves as a conduit to enhanced health, education, and livelihood opportunities for people. It has a direct impact on their well-being, economic potential, and overall quality of life.
Currently, there are multiple sources of climate data that can be leveraged by governments – many of which are freely available. For example, However, processing and interpreting this data requires specialized skills, which are not always readily available. Without public sector capacity to process this data, the benefits of integrating climate considerations into government decision-making cannot be fully realized. And all else being equal, increased volumes of available data can lead to an information overload, further decreasing the government’s capacity to effectively leverage that data, creating a disconnect between data availability and actionable insights.
through two main mechanisms. First, AI can sift through vast datasets, identify actionable insights, and presenting them in intuitive, user-friendly interfaces. Second, governments can feed their own structured and unstructured datasets into AI solutions to generate insights that are more contextualized and targeted to government needs. This tailored approach can significantly enhance the public sector's ability to leverage available climate data, lowering the barriers to climate-smart public investment and asset management strategies.
With the objective of advancing climate-smart development, we initiated the integration of Artificial Intelligence (AI), within the Green Economy Diagnostic (GED) prototype under the Climate-Smart Development Initiative. The GED holistically maps the economic and environmental performance of sub-national regions within a country, to help them make informed, climate-smart investment decisions. Harnessing multiple datasets (e.g., on night light, extreme weather, air quality), structured through a rigorous methodology, and using OpenAI’s application programming interface (API), GED can identify trends such as variations in air quality and temperature volatility and offer recommendations such as stricter standards for limiting emissions and enhancing infrastructure resistance to extreme temperatures.
Although still nascent, the fusion of AI with the GED has already demonstrated its potential by amplifying the accuracy and comprehensiveness of analysis and hastening the process of generating actionable insights for climate-resilient economic planning at the subnational level. In future stages, one could envision leveraging AI in GED for modeling of climate impacts, optimizing resource allocation for climate mitigation and adaptation, and supporting stakeholder collaboration. Each of these avenues opens doors to more informed, effective, and timely climate action.
But as we conduct this work, two key lessons emerge. First, it's important to bring together teams capable of translating user needs into tangible solutions. While technological solutions can help bridge the gap in government capacity for data interpretation, the need for multidisciplinary teams with specialized skills to build these tools remains crucial – which requires adequate funding and support which, at least in early stages, might not be sufficiently provided by governments or the private sector. Second, the creation of practical, actionable solutions that can produce early demonstration results and be incrementally refined, requires a shift in approaches to technology in development, moving away from extensive theoretical discussions, reports, and memos, towards a more 'demos-driven' approach. This approach emphasizes the value of hands-on experimentation and iterative learning in creating effective solutions for climate and development challenges.