Using Evolutionary AI to Optimize Land Use for Climate and Economy
Scientists at The University of Texas at Austin and Cognizant AI Labs have developed an AI system that recommends optimal land-use policies to balance carbon storage, economic impact, and environmental health. By analyzing 175 years of global land use and carbon data, the AI identifies strategies that support the United Nations’ sustainability goals while minimizing trade-offs.
This approach, based on evolutionary AI, mimics natural selection by generating multiple policy scenarios, evaluating their outcomes, and iteratively refining them. Poorly performing policies are discarded, while the best combinations reproduce with variations, leading to increasingly effective solutions over many generations.
Balancing Carbon Storage with Practical Land Management
Contrary to the simple idea of converting as much land as possible to forests, the AI model revealed more nuanced strategies. For example, converting cropland to forest yields higher carbon storage benefits than converting rangeland, such as deserts or grasslands. The effectiveness of land-use changes also varies by latitude, emphasizing that targeted actions in key areas provide better results.
Daniel Young from Cognizant AI Labs highlighted the importance of balancing climate goals with preserving food supply, habitats, and urban areas. “We need to be smart about where we make changes,” Young said, underscoring the need for carefully considered policies.
Interactive Tools for Policy Makers
The research team has created an interactive tool that enables legislators and decision makers to simulate how incentives like tax credits might shift land use and reduce carbon emissions. Since agriculture and forestry contribute nearly 25% of human-caused greenhouse gases, optimizing these sectors is critical for climate mitigation.
This AI-driven approach offers alternatives that can be more acceptable to stakeholders hesitant about change, providing data-backed recommendations that balance multiple priorities.
Integrating AI into Climate Simulation Platforms
The team is collaborating with Climate Interactive to enhance the En-Roads climate simulator by adding their evolutionary AI method. While En-Roads allows users to manually adjust policy levers to see climate impacts, the new AI add-on, called Decision Making for Climate Change, can automatically generate optimized policy sets to meet specific temperature targets such as limiting warming to 1.5°C above pre-industrial levels.
This tool streamlines the policy optimization process, providing users with actionable recommendations rather than relying solely on trial and error.
Broader Applications Beyond Land Use
Previously, the same evolutionary AI framework was applied to optimize responses to COVID-19. The system evaluated numerous intervention combinations and identified effective policies, such as early lockdown timing, that were not initially anticipated. The researchers even advised Iceland’s government on school opening strategies during the pandemic.
These examples demonstrate the AI’s versatility in addressing complex decision-making challenges involving multiple competing factors.
Open-Source Platform for Wider Impact
The underlying technology of this AI system is available through the Project Resilience platform. This open-source initiative encourages researchers, policymakers, and organizations to develop models and solutions addressing sustainable development goals and other global challenges.
The growing interest in AI tools for socially beneficial applications is encouraging. According to project lead Risto Miikkulainen, the AI for Good movement is gaining momentum as more people recognize AI’s potential to assist in critical, real-world decision-making.
Learn More About AI Tools and Courses
For researchers and professionals interested in advancing their AI skills, Complete AI Training offers a selection of courses focused on applied AI techniques, including evolutionary algorithms and climate modeling applications.
These resources can help you explore how AI can support complex problem-solving in your field.
Your membership also unlocks:
