Do Robots Dream of Passing a Programming Course?

Torres, Nicolas

Abstract

Programming typically involves humans formulating instructions for a computer to execute computations. If we adhere to this definition, a machine would seemingly lack the capability to autonomously design algorithms. However, recent generative Artificial Intelligence models, such as GPT, have demonstrated an impressive ability to perform complex human tasks with remarkable precision. In this paper, we initially showcase how an AI model can successfully complete an entire college-level programming course, akin to one of the top-performing students in the class. We then put forward strategies for crafting programming exercises that enable educators to effectively integrate these innovative technologies into their teaching methods. Lastly, we illustrate how these models can transition from being perceived as a potential threat to educators to becoming a valuable opportunity when employed judiciously.

Más información

Título según SCOPUS: ID SCOPUS_ID:85178999122 Not found in local SCOPUS DB
Título de la Revista: 2018 37TH INTERNATIONAL CONFERENCE OF THE CHILEAN COMPUTER SCIENCE SOCIETY (SCCC)
Fecha de publicación: 2023
DOI:

10.1109/SCCC59417.2023.10315744

Notas: SCOPUS