Measuring What Generative AI Actually Saves in Instructor-Led Training Development: A Matched-Deliverable Time-and-Quality Comparison
Abstract
Instructor-led training (ILT) is a core delivery mode for workforce development, safety, equipment familiarization, and procedural instruction, and producing it is expensive. Generative Artificial Intelligence (AI) promises to cut that cost, and organizations want a number they can trust. The obstacle is measurement. The metrics that are easiest to capture, software-recorded editing time and time per slide, are unreliable, and the apparent advantage of AI can appear or vanish depending on what is counted. ILT slides are deliberately sparse because the instructor carries the content, so word count per slide says little about completeness, and recorded editing time counts only work done inside the slide file. This paper fixes the deliverable instead: a complete teachable module of defined instructional scope, including slides, instructor talking points, activities, and assessment items, sufficient for a qualified instructor to deliver. Development time is measured to that standard and normalized per learning objective and per minute of instruction. Three workflows produced the same module from the same source material: an experienced designer, a junior designer, and a designer working with generative AI followed by Subject Matter Expert (SME) review. The AI-assisted workflow reached a complete module in 5 person-hours against 32 and 56, but drafting fell 94 percent while review and revision rose to 65 percent of what remained. The contribution is a matched-deliverable method that lets organizations state AI productivity gains defensibly rather than rhetorically, with guidance for safe adoption in safety-critical domains.
Keywords: Generative AI, Instructor-Led Training, Instructional Design, Productivity Measurement, Safety Training, Human-AI Teaming
DOI: 10.54941/ahfe1008495
Cite this paper
More from this volume
- AI-Enhanced Live-Streamed English Language Learning: Designing and Evaluating an Intelligent Platform for Chinese Learners
- "They Will Move": Confirmation Bias, Expectation Frames, and Premature Monitoring Drop Off in COLREG Encounters
- Automation transparency communication does not improve supervisory control performance, but shapes operator trust
- The influence of neuroticism on startle and surprise response: a flight simulator study
- EEG-Based Classification of Reward, Penalty, and Movement Events Across Four Video Games of Varying Complexity
- The Effects of Task Similarity and Time Pressure on Task Performance Following Interruptions in Complex Human-Computer Interaction Systems
- Trusting the Chain, Not the Black Box: A Provenance-Based Architecture for Human-Machine Teaming in Military Autonomy
- Design of a Simplified Humanoid Robot Face: Cultural Differences and Emotional Synchronization
- From Visual Observation to Auditory Interaction: A Human-Centered Tangible Music System Based on Physical Objects
- The Impact of Scarcity-Based Dark Patterns on User Performance, Perceived Darkness, and Workload in Subscription Services
- A Behavioral Observation Tool for Diagnosing Human–AI Interaction Profiles in Engineering Design Education
- The Mediated Moon: Semantic Mediation Between Gesture and Generative Image


AHFE Open Access