2026 ASEE Annual Conference & Exposition

Paper Type: The Tinkerer’s Triumphs and Tragedies

The rapid diffusion of generative artificial intelligence (GAI) and large language models (LLMs) into engineering education has disrupted traditional approaches to teaching numerical methods courses in engineering curricula. Core competencies that once required sustained cognitive engagement, such as programming algorithms in MATLAB, Python, or C, or executing standard numerical computations, can now be reproduced with minimal effort by prompting an AI system. This work proposes a shift in course purpose, objectives, and assessments towards durable, professionally valuable skills focused on computational decision making, including higher-order argumentation, defending model assumptions, framing and contextualizing problems within domain-specific contexts, and communicating results to diverse audiences. This shift challenges both the validity of long-standing assessment practices and the assumptions faculty hold about which skills are central to an engineering curriculum.

To affect the proposed changes, we redesigned a sophomore-level numerical methods course to emphasize higher-order engagement aligned with Bloom’s taxonomy, specifically targeting skills and dispositions at the evaluation level that current GAI tools do not reliably replicate. New course objectives focus on students’ abilities to make computational decisions, requiring justification of methods and models, data selection, validation of results, and the supporting of judgement with evidence-based argumentation. Rather than assessing students on code production or computation accuracy, redesigned assignments require students to communicate results persuasively to technical and non-technical audiences, requiring student decision making on information design and appropriateness. Deliverables include written memos and visual presentations, both of which foreground metacognition, epistemic cognition, and audience analysis.

This paper presents the rationale for these changes, outlines the new objectives and assessment structures, and discusses early insights into how this shift reshapes student engagement. By repositioning numerical methods instruction away from computer code creation and numerical results and toward the communication and justification of computational decisions, we aim to prepare students in the age of GAI with competencies that resist automation and remain central to engineering practice.