Visual Programming

Visual programming uses visual elements to represent code, aiming to make programming more accessible and intuitive. Current research focuses on leveraging large language models (LLMs) to generate and debug visual programs, often within specific domains like audio processing, assistive technologies, or scientific figure creation, and employing techniques such as program synthesis, metaprogramming, and self-supervised learning. This field is significant because it promises to lower the barrier to entry for programming, improve code generation efficiency, and enhance the interpretability of complex AI systems, impacting both software development and educational applications.