Abstract:

Literate programming emphasizes the importance of clear code documentation and explanatory text integration, and fosters a deeper understanding of engineering concepts, effective collaboration, and critical thinking skills. It has been at the forefront of data science practice in the form of interactive notebooks like Jupyter notebooks for over a decade. In fact, literate programming as a technique of structured documentation has been used in computational physics and engineering education and practice since its definition by Knuth in 1984.

This presentation critically examines the role of AI coding assistants and their impact on the relevance of literate programming in engineering education. As AI coding assistants advance, offering real-time feedback and automated code completion, questions arise regarding the necessity of literate programming principles.

The session explores the benefits of AI coding assistants, including enhanced coding efficiency and support for novice learners. However, it also considers the potential drawbacks and limitations of relying solely on these tools. We posit a symbiotic relationship between AI coding assistants and literate programming: both approaches can complement each other by combining AI automation with human-readable documentation.

Attendees will gain a nuanced understanding of the interplay between AI coding assistants and literate programming in data science and engineering education. They will consider the strengths and limitations of each approach and explore how to strike a balance to create a comprehensive educational experience.

Keywords:

Artificial Intelligence, Literate Programming, ChatGPT, Data Science, Large Language Model.