Rethinking Engineering Education Outcomes in the Light of Artificial Intelligence

Abstract

The deployment of LLMs (large language models) in AI (artificial intelligence) has led to a wide-spread increase in the use of AI in contemporary society, which will impact how teaching and learning takes place in higher education in the future. Since constructive alignment – which suggests that in curriculum design there should be a close alignment of teaching and assessments with intended outcomes – plays a crucial role in higher education teaching and learning, this paper examines how burgeoning AI technology reshapes such alignment considering both human and AI capabilities. The current paper investigates an undergraduate engineering curriculum to determine which tasks specified in module outcomes, as presented in the yearbooks (program portfolio), can 1) be delegated to AI, 2) are uniquely human, and 3) require a combination of human and AI capabilities. This categorization of outcomes will, be compared to the ECSA prescribed graduate attributes of the South African engineering student. The aim will be to determine which module outcomes and graduate attributes remain relevant in an AI-dominated pedagogical environment, and which have been usurped by the increasing competence and accuracy of AI systems, to future proof students in a changing technological landscape. We recommend that the outcomes and graduate attributes should be reformulated to incorporate the potential and risks of AI from the beginning.