In This Story
A computer program so adept at playing the game Go that it could beat the world's best players changed the career plans of George Mason University computer science assistant professor Simin Chen.
Watching advances in artificial intelligence (AI), especially its ability to outperform human experts at the ancient strategy game, convinced Chen that he wanted to explore opportunities the technology offered, so he left his bachelor's and master's degrees in civil engineering behind.
Today, his research encompasses three interconnected areas: AI, computer programming, and trustworthiness. He examines AI systems at multiple levels, from the data used to develop models to the underlying computing infrastructure that supports them. His goal: identify weaknesses and develop approaches that improve security and reliability throughout these systems.
One particular interest is agentic AI and the considerable promise and risk it brings. AI-generated programs can contain errors or security vulnerabilities, potentially creating problems when the technology is used without adequate oversight.
For Chen, ensuring that these increasingly capable systems produce dependable results is a central research challenge. At George Mason, his immediate priority is establishing a research lab, recruiting students, and developing projects centered on AI, programming, and security. Over the next several years, he hopes to create fundamental frameworks that help researchers and developers build more trustworthy AI agents. His longer-term ambitions extend beyond software engineering to the broader use of AI across engineering disciplines.
Beyond research, Chen also sees important implications for computer science education. As AI becomes increasingly capable of handling routine programming assignments, educators must reconsider how they teach students to think critically and solve problems independently.
"How do we teach the first generally AI assisted computer science students to do human reasoning but not purely rely on AI's answer to finish assignments or tasks?" he said.
Chen compares AI to a powerful calculator. Just as calculators allow people to tackle more complicated mathematical problems, AI can help researchers, engineers, and students accomplish things that would otherwise be difficult or time-consuming.
"People who know how to use AI better can achieve better or higher achievements," he said.