The Shift in Teaching Programming: Using LLMs in Introductory CS Courses

Dr. Leo Porter and Dr. Daniel Zengaro discuss the decision to incorporate LLMs (Large Language Models) into the introductory CS1 courses they teach. They explain that the increasing impact of LLMs on programming projects led them to realize the need for changes in how programming is taught. Instead of focusing extensively on syntax, they highlight the importance of teaching skills such as reading code, writing effective test cases, debugging, and problem decomposition. LLMs assist in automating syntax-related tasks, freeing up time for students to concentrate on higher-level problem-solving skills. They also note the potential for LLMs to change the interface and workflow of programming, as students could rely more on prompts and regenerate code as needed. However, they emphasize the importance of teaching principles and a systematic approach to using LLMs to ensure long-term success.