Can AI Be a Force for Environmental Sustainability?

Artificial Intelligence (AI) is swiftly inundating industry and society, dragging us all along like a rip tide out to uncharted seas. Along with the societal, behavioral and economic changes that face us, there are concerning environmental implications due to AI infrastructure’s enormous demand for energy and water.

Although we almost find ourselves coerced into this brave new world, we must ask: Is it possible to leverage this revolution to set our ecosystems and environments on a course towards a thriving future?

This was the central question at this spring’s Innovation Summit hosted by the University of Colorado’s Environmental Data Science Innovation & Impact Lab (ESIIL).

The summit brought together experts from an array of disciplines across academia, industry and nonprofit organizations, including myself and Associate Curator of Mycology Andrew Wilson. We spent the week learning new methods and approaches for applying AI to environmental sciences, sustainability and the biodiversity research that Andrew and I do here at Denver Botanic Gardens.

After a whirlwind of workshops and discussions, we formed a group interested in using large language models (LLMs) to gather data that are distributed across the internet. Bringing related, but often difficult to find, datasets together can aid researchers in their ability to pool together the data from similar studies to uncover larger trends. We call these metanalyses.

Utilizing ESIIL’s access to high-performance computing and LLMs, we created an AI agent that scours the internet for relevant datasets. The process was surprisingly simple, and within the course of a few hours we had put together an imperfect but functional product. Our group plans to continue meeting this summer to refine the project and possibly create some use-case studies for publication.

The summit concluded in a most unique and personal way. All 150 of us walked out into the sun, removed our shoes and socks and formed a circle on the grass. We then proceeded in a self-looping line to look each participant in the eyes, shake hands and exchange salutations. It was a deeply humanistic approach to remind us that people working together, not technology alone, can solve the biggest problems we face.