Automatic Example Generation for Babylonian Programming to facilitate Human-centric Domain Exploration

About

Software is at the center of many sustainability efforts, yet the abstract nature of source code impedes collaborative design. This project uses concrete, live examples to explain software and demonstrate intended changes. The team led by Robert Hirschfeld (HPI) and Martin Rinard (MIT) examines the consequences of software changes by example and propose a novel approach where programmers can directly specify the expected outcome of a computation using example.

The team has achieved breakthroughs across three key areas: AI-generated coding examples that automatically adapt to real-world contexts, a visualization system for tracking how code modifications impact program behavior, and quality assessment frameworks integrated into professional tools like PyCharm. These innovations are already transforming sustainability software development by enabling true co-creation between technical and domain experts.