Research Article
Capturing Elusive Technology: Designing a Course on AI for Learning and Development Practitioners
Synthesis: Boise State's Organizational Performance and Workplace Learning program needed a course on AI for working learning-and-development practitioners, and da Silva built one as a five-week asynchronous elective, later extended to ten weeks. The design combined constructivist project work with a Learning Engineering cycle, and a team of eight handled design, media, accessibility and quality assurance. Four iterations followed student feedback: cheaper tools, ethics before sign-up, clearer project options, fresher tool guidance. The case reports the design decisions and the change log, not learning outcomes.
Key Findings
- The audience forced a broad entry point. Learners ranged from career changers to chief learning officers, in a program that serves the United States, Europe and Asia. The course had to work for people whose jobs, tools and AI experience differed widely, which pushed it toward a project each learner could shape to their own workplace.
- Learning Engineering supplied the iteration machinery. The design followed Kessler's cycle of challenge, creation, implementation and investigation, and used the LEED Tracker to record every requested change, its justification and its status. That record is what makes the four iterations legible.
- Cost was the first barrier students named. Several pointed out that the AI tools the course demonstrated were paid products. The fix was to foreground a free option through the university account and to add guidance for it, which removed the barrier for that iteration.
- Ethics had to come before the first prompt. Feedback from the instructor and subject-matter expert moved privacy and acceptable-use guidance into the kick-off session, so students knew what not to paste into a tool before they signed up for one.
- Institutional timelines, not pedagogy, set the pace of change. Students asked for a longer course after the first iteration; the change could only land in the third, because major program changes follow a calendar. Two iterations ran at a length the designer already knew was too short.
- The content aged between offerings. Each iteration updated tool guidance and readings, and the case treats that maintenance as a standing cost of teaching a moving technology rather than a one-off revision.
Designing a course for people whose jobs differ
The hard design question, for a workplace learning audience, was not which tools to cover but what a chief learning officer and a career changer could both do in five weeks. The answer was a single project, chosen by the learner, with the course supplying the workflow: analyze a performance problem, design an intervention, build a piece of it, and reflect on the AI tools used along the way. That structure let the same assignment carry different levels of ambition, and it made the tool guidance a support rather than the content.
What this means for practice
- Instructional designers. Keep a change log with a reason attached to each change, as the LEED Tracker does. Four iterations of a fast-moving course are only interpretable if the reasoning survives the term.
- Instructional designers. Check the cost of every tool you demonstrate, and name a free route through the institution. A course that requires a subscription selects its students by budget.
- Faculty developers. Put privacy and acceptable-use guidance in the first session, before students create accounts. Retrofitting it after the first assignment is too late for the data already shared.
- Administrators. Approval calendars, not evidence, decide when a course can change length. If a design needs a structural change, start the paperwork in the first offering rather than after the second.
Limitations
- No outcome data: no pre/post measure, no comparison section, no analysis of student work. The account is a design narrative with a change log.
- Student feedback is reported in summary form, without counts, sampling or an instrument, and it drove design decisions that a formal evaluation would have tested.
- One program, one institution and one instructor-designer, with a team of eight supporting the build; the resourcing is not typical of an individual course revision.
- The course is asynchronous and elective, so the students who enrolled were already interested in AI, which limits what the feedback can say about the wider practitioner population.
Citation
da Silva, R. L. (2025). Capturing elusive technology: Designing a course on AI for learning and development practitioners. International Journal of Designs for Learning, 16(2), 124–134. https://doi.org/10.14434/ijdl.v16i2.42150