The "Planetary Health Axis System" (PHAS), jointly developed by Peking University's Institute for Global Health Development and dozens of top research institutions both domestically and internationally, was recently officially released, becoming the world's first large-scale artificial intelligence model for planetary health.

Currently, the PHAS system comprises a planetary health coordinate system based on four main axes: "human health," "species health," "environmental health," and "social health." It enables dynamic visualization and calculation of 48,000 key variable indicators and utilizes open and accessible big data sources from around the world for modeling and data analysis through mechanisms such as machine learning, systems science, and expert feedback calibration.

With its comprehensive data foundation, adjustable model complexity, evolutionary architecture, and long-term optimization goals, PHAS has demonstrated unique superior performance: First, it breaks through the limitations of the traditional "selected data source" approach, utilizing all publicly accessible data (including over 48,000 variables) and artificial intelligence to significantly expand the scope of knowledge acquisition and the degree of technology integration.Second, by leveraging machine learning and AI training data, the continuity and high quality of the data are ensured, solving the problems of limited data and reliance on manual definition of quality in traditional methods. Third, it is no longer limited to sub-domains or limited cross-domain interactions, but rather symmetrically integrates data from all sub-domains to promote interaction and unleash the synergistic value of data across all domains. Fourth, it does not simply summarize the conclusions of experts from different fields, but generates conclusions from a system calibrated by feedback from a global expert group, making the conclusions more systematic and professionally calibrated from a global perspective.

Furthermore, as an AI-driven global public analytics platform, PHAS, through system modeling and continuous learning, can analyze and present the complex relationships between massive amounts of variables, and employ cooperative game theory methods (such as Shapley values) to fairly allocate planetary boundary responsibilities among nations. Currently, PHAS is primarily applied to environmental economics, and in the future, it will expand to fields such as epidemiology, water resources, and food security.

The Paper learned from Peking University that the PHAS project was initiated because of three important trends that Liu Guoren, a health economist and dean of the Peking University Institute for Global Health and Development, has long been concerned about: First, modern economic growth, although powerful and widespread, has driven large-scale poverty alleviation globally, but it is highly dependent on fossil fuels; second, "Great Divergence," economic growth is accompanied by the continuous widening and solidification of global disparities in wealth, resources, and carbon emissions; and third, "Future Discounting," that is, the economic system, due to inflation, uncertainty, and investment returns, focuses more on current gains and ignores intergenerational equity.

These three phenomena prompted him and his team to consider: if these three trends continue to develop, can human civilization maintain its sustainability, and how can these trajectories be altered?

As AI technology gradually matures, Liu Guoren and his team saw the possibility of developing an effective tool and officially launched the PHAS project in 2023. The project aims to establish a scientific and comprehensive framework that clearly and intuitively presents the current situation and risks, allowing everyone to see their own health and the development risks of the entire planet in real time and dynamically.

As a "digital compass" for tracking planetary health, PHAS has constructed a multi-disciplinary, multi-layered AI-driven digital model. For the first time, it has realized a planetary health system that covers all countries and regions around the world, is dynamic in real time, and has visual interaction. This system aims to map the panoramic process of the relationship between human development and the generalized planetary boundaries, thereby better understanding the complexity of planetary health and promoting the transformation of economic development paradigms from a single GDP dimension to a new type of planetary economy with multiple dimensions.

When discussing how to truly translate the scientific insights of PHAS into global decision-making and concrete actions, Liu Guoren emphasized that PHAS is an open global public good designed to empower various users, including governments, academia, the private sector, and civil society. By constructing a scientific and systematic framework, the public can understand their own health and the health and development risks to the planet in real time and dynamically. In the future, the team will promote the widespread adoption of the system by establishing a partner network and other means.