Large language models in the engineering workplace and associated curricular implications: An exploratory study
2024
The disruptive appearance of ChatGPT in November '22 has spurred considerable debate and research on student use of Large Language Models (LLMs) in higher education.Although divergent viewpoints exist, many institutions of higher education have gravitated towards a policy that can be succinctly characterized as "informed and responsible use".A categorical ban on generative AI tools would fail to adequately prepare students for the future workplace where such tools are anticipated to be harnessed for their productivity enhancement, so the argument goes.Conversely, an all too uncritical embrace would undermine indispensable learning objectives of higher education curricula.This exploratory research paper seeks to identify the extent -and purposes -to which professional engineers are currently using LLMs and to examine associated curricular and pedagogical implications.In a student-driven action research project, engineering students of the second bachelor of the industrial engineering technology program at University of Leuven (Belgium) surveyed 249 engineers in October 2023.Results show that about half of respondents do not make use of LLMs for professional purposes and have no immediate intention of doing so, with some engineers referring to prohibitive corporate policies.About one-third of respondents are currently not employing LLMs, but they state the intention to explore their potential in the (near) future.The remaining respondents state that they already make use of LLMs for professional purposes, with over half using them for content generation.Other common modes of usage are instructing an LLM to revise a self-written text to optimize phrasing, spelling and grammar or to repurpose it for different audiences or media; to summarize texts; to write computer code; to explain technical concepts; to provide references or sources; and as a search engine.It is worth noting that several of these types of usage do not fit within the commonly accepted boundaries of "informed and responsible use", underscoring the need for didactic interventions in higher education that raise student awareness of how LLMs actually function, what their inherent limitations are and which ethical concerns they entail.This paper describes how such interventions can be designed and integrated within an engineering program.Furthermore, it suggests ways in which higher education programs can monitor the fast-evolving landscape of AI workplace practices to ensure students are well-prepared to navigate the opportunities but also the challenges presented by LLMs.