Anderies, John Marty

Combating environmental degradation requires global cooperation. Institutional designs for such efforts need to account for human behavior. In this talk, I will use the voyage of the Titanic as an analogous case to learn from, and use behavioral insights to identify critical aspects of human behavior that serve as barriers or opportunities for addressing the global environmental challenges we face. Based on an analysis of existing international organizations focused on how human behavior and institutions intersect, I will describe a set of public goods that may help us mitigate aspects of human behavior that act as barriers to collective action and leverage behaviors that promote it. Finally, based on insights from this analysis applied to existing institutional solutions for global environmental protection, I will present a set of institutional design features that, if adapted to better account for human behavior, could lead to more effective institutional solutions to global environmental problems.

The cryosphere is rapidly transforming due to climate change, with a shortening ice season, thawing permafrost, and glacial retreat becoming increasingly prevalent. These changes contribute to "Arctic amplification," causing the region to warm four times faster than the rest of the world. Ice has historically been a key indicator of climate change, and the Earth is becoming less 'frozen.' The cryosphere is critical for global climate stability and human populations. Rockstrom et al. (2024) recently identified the Arctic Cryosphere as a planetary commons, spanning national and supranational boundaries. Unlike conventional global commons, the Arctic has been populated for millenia, particularly by Indigenous Peoples. We urgently call for treating Arctic ice, snow, and permafrost as a critical planetary commons – a collectively governed common pool resource important at the planetary scale. Sustaining these frozen commons requires engagement from diverse actors and diverse ways of knowing at local, regional, national, and global levels, extending beyond nation-to-nation decision-making. Building on experience from Interior Alaska and northern Mongolia, this presentation highlights the importance of the concept of frozen commons for current discussions about sustainability and resilience of complex systems. We argue that bringing perspectives, amplifying voices, and elevating ethics of care for human and non-human relations in historically marginalized communities, such as Indigenous Peoples and nomadic herders in decision making and planning efforts is necessary to address climate change in a just and equitable way.

This research examines whether LLMs demonstrate environmental awareness, with particular focus on how they reflect and potentially bias human attitudes toward environmental issues as a result of their training data and processes.

With the widespread adoption of AI technologies, particularly large language models (LLMs) such as ChatGPT, questions arise about their training processes and inherent biases. While these models have become integral to daily life, the opacity of their training data and processes—protected as corporate trade secrets—makes it difficult to clarify their pre-training biases and potential toxicity. Even after developers implement fine-tuning measures aimed at neutrality, initial biases may persist in these systems.

We developed a comprehensive evaluation framework to assess major LLMs from leading AI developers worldwide. The framework examined environmental behavior, knowledge, attitudes, social norms regarding environmental issues, and the ability to predict human behavioral shifts toward pro-environmental behaviors. To establish a comparative baseline, we conducted a cross-sectional survey using the same evaluation framework with a stratified random sample of 385 participants representative of the U.S. population.

The results show that while LLMs possess extensive knowledge, this does not always lead to positive predictions about human environmental behavior. Significant gaps exist in environmental awareness between human participants and LLMs, with these gaps varying based on AI developer, national origin, and training language. Our study highlights the ongoing challenges in developing AI systems that can accurately interpret and represent complex human and societal issues. These findings have important implications for developing more culturally and environmentally aware AI systems, suggesting the need for more sophisticated approaches to AI training and development in the context of environmental understanding and human behaviors.