Abstract:
Artificial intelligence tools are democratizing programming, making computational research accessible to researchers who have little or no formal programming background. This seminar offers a practical introduction to programming with AI assistance, beginning with a brief history of how AI—and AI coding tools in particular—came to be. We then discuss practical considerations for programming with AI: how to work effectively with AI assistants, how to frame problems clearly, and how to evaluate the code they produce. The foregoing skills are essential in addressing “technical debt” in AI-assisted programming, where generated code does not generalize easily to new features. The talk should provide insights into what AI-assisted programming can and cannot do, and a foundation for using AI tools responsibly.
Joseph L. Hellerstein received his PhD in Computer Science from the University of California at Los Angeles. He has thirty years of experience in research and software engineering at the IBM TJ Watson Research Center, Microsoft Corp. and Google. He is a Fellow of the IEEE for contributions to the control engineering of computer system performance, and has approximately 200 peer-reviewed publications. At the eScience Institute, his primary focus is on AI-assisted software engineering. His role in the Department of Bioengineering is teaching and research related to mechanistic models of biological systems.
