Wes Bethel, Associate Professor of Computer Science

Wes Bethel, associate professor of Computer Science, is the principal investigator of a research project that will study methods for automating the generation of software tools and processes for the purpose of constructing software that builds machine learning models.  

The work is in support of an effort funded by U.S. Department of Energy’s Office of Fusion Energy Sciences to leverage artificial intelligence (AI)/machine learning tools to reduce computational time-to-solution for specific physics calculations, with the ultimate objective of being able to predict plasma behavior in real time in fusion tokamak devices.  

Bethel, who joined SF State in 2022 after a career as a computer scientist at Lawrence Berkeley National Laboratory, has assembled a team of Computer Science graduates and undergraduates (V. Cramer, C. Pestano, A. del Rio and S. Verma), and other faculty (Computer Science Lecturer Lothar Narins) to study this problem. The SF State team is part of a larger multi-institutional effort led by J. Wright at the Massachusetts Institute of Technology and includes researchers from the Princeton Plasma Physics Laboratory and Lawrence Berkeley National Laboratory. 

The key idea behind the team’s approach is to leverage recent advances in cloud-based AI tools, such Large Language Model implementations such as OpenAI’s ChatGPT and GitHub’s Copilot to quickly produce code that builds other AI models and model validation processes. The intent is to reduce the time to solution for trained models from years to weeks.  

Bethel looks forward to reproducing this work on other computational challenges within the SF State scientific community.