Date Posted: 5/02/2024

Ralitsa Todorova is a neuroscientist who intends to decode the neural activity in the brain during mental imagery. Zhongmou Chao is a chemical engineer developing a chip to sniff out cancers. Felipe Pacheco is an ecologist using state-of-the-art analysis to inform where best to site hydropower dams to produce the most power while minimizing environmental impacts.

While varied in their goals, each of these Cornell scientists shares a key commonality in their methods: through radical collaboration, they are leveraging artificial intelligence to develop solutions that were unimaginable just a decade ago and unearth scientific insights for the greater good.

These three are among the 20 Eric and Wendy Schmidt AI in Science Postdoctoral Fellows who presented their ongoing research projects during an end-of-year showcase for the AI for Science course (CS 6703) held April 19 in Gates Hall.

A color photo showing a woman pointing at a poster

Hosted jointly by the Cornell Ann S. Bowers College of Computing and Information Science, the College of Agriculture and Life Sciences, the College of Arts and Sciences, the College of Veterinary Medicine, and Cornell Engineering, the event entailed one-minute flash talks from each fellow and a poster session. The event highlighted AI’s broad application across the sciences – from sustainability sciences to astrophysics – and the critical role it will continue to play in scientific discovery.

“The next 20 to 30 years is going to be a golden era for science, and a lot of that work will be driven by what all of you are doing,” said Kavita Bala, dean of Cornell Bowers CIS and lead dean of the Cornell AI Initiative, in her opening remarks. “As scientists, we are fortunate to live in this time.”

The 20 fellows are part of the Cornell-Schmidt Sciences partnership established in 2022. Cornell is one of nine universities in the Eric and Wendy Schmidt AI in Science Postdoctoral Fellowship program, which provides access to AI tools and training to accelerate scientific innovation. The Cornell University AI for Science Institute was established to help guide postdoctoral research in this effort. 

"AI is revolutionizing our approach to science, empowering researchers to tackle scientific challenges in various fields," said Carla Gomes, the Ronald C. and Antonia V. Nielsen Professor of Computing and Information Science, the AI for Science course instructor, and the institute’s co-director. 

A color photo showing the AI for Science poster session

AI can help us better understand the neural circuitry behind mental imagery, said Todorova, who received a Ph.D. in neuroscience from PSL Research University and came to Cornell’s Department of Neurobiology and Behavior as a postdoctoral researcher in 2021.

“I’m interested in how our mental imagery takes place. When you visualize something, how does that work?” she said. This understanding could lead to better treatments for disorders that hinder internal recollection and imagination, like Alzheimer’s disease, schizophrenia, post-traumatic stress disorder, anxiety, and depression.

Elsewhere, research suggests dogs can sniff out certain odor molecules in cancers. According to Chao, with AI, we can create a chip to replicate the 300 million scent receptors in a dog’s nose and use it to detect molecules like hexanal and styrene found in samples from cancer patients. 

“Different smells trigger different pattern responses,” said Chao, who received a Ph.D. in chemical engineering from the University of Pittsburgh and joined Cornell’s School of Chemical and Biomolecular Engineering as a postdoctoral researcher in 2021. We can use machine learning to decode these pattern responses and predict the presence of odor-emitting molecules, he said. 

AI can also inform decisions to help with the world’s low-carbon energy transition. Felipe Pacheco, a postdoctoral researcher in the Department of Ecology and Evolutionary Biology, heads a large research team composed of scholars from Cornell and 10 other universities, institutes, and centers. The team is using AI to inform where best to site hydropower dams to produce the most power while minimizing environmental impacts. Global hydropower capacity will need to double to reach 2050 decarbonization goals, and AI can help analyze trade-offs between hydropower development and environmental conservation, researchers said.

A color photo showing a man pointing at a poster

"Our fellows are at the forefront of integrating AI into scientific discoveries and research, driving innovations that were once beyond our imagination," said Fengqi You, the Roxanne E. and Michael J. Zak Professor in Energy Systems Engineering and the institute’s co-director.

"AI is transforming how we address environmental challenges, enabling us to achieve sustainability goals with unprecedented precision," added Alexander S. Flecker, professor in the Department of Ecology and Evolutionary Biology and an associate director at the institute. 

These projects show the critical role that universities will have in AI’s development and deployment, Bala said. 

“Universities are the only places that have this breadth of scientists,” she said. “What we have at Cornell is magical. We’re bringing together this set of great scientists together with AI researchers.”

Louis DiPietro is a writer for the Cornell Ann S. Bowers College of Computing and Information Science.