I was delighted to be asked to share my thoughts on producing convincing, engaging bespoke e-learning in the age of AI. Rather than go into theory, or list the various tools now available, I wanted to share some candid thoughts from current projects.
AI is like a keen junior researcher. It helps me locate content, interrogate a topic and structure my thoughts. As someone who is neurodiverse – whose teachers said I would never be able to write – I find that very useful.
But I have also never been busier. I am increasingly called in to help with projects that began with great enthusiasm, where the sheer volume of material produced by AI created a feeling of progress. A lot of content is not the same as a well-designed course.
Before creating content, start with a simple question: what should someone know or do differently after completing this course? From there:
- Map the learning outcomes.
- Interview the people who understand the subject in practice.
- Storyboard the content and decide what can be cut.
- Choose what should be filmed, animated, written or made interactive.
- Write scenarios in which the answers are nuanced and plausible. Treat learners as adults.
For instance, I am currently working on a short course about consultative sales skills. The initial ideas and experience had to be turned into filmed sections, animation, questions and feedback. The extract below is from the real working storyboard. It shows how the instructional designer sets out both the words the learner will hear and what should happen onscreen.
Extract from the working storyboard for a consultative sales course, written by instructional designer Daniel Whiston and reproduced with permission.
This is why bespoke e-learning is a team effort. The subject expert provides the experience. The instructional designer and writer give it structure. A filmmaker or animator decides how best to show it, and the developer builds the final experience. It is difficult for one generalist e-learning developer to do all of those jobs equally well.
AI can help throughout that process. It can organise interview notes, suggest questions, check for gaps, tidy writing and help us explore visual approaches or different backgrounds. But the most useful parts of a course still tend to come from real experience: the phrase an expert actually uses, the difficult decision they have faced, or the example that learners recognise from their own working lives.
A real person discussing a topic also brings an emotional connection. Even where budgets are limited, filming a colleague on a phone can be more convincing than using an AI avatar.
The use of AI that interests me most is question and answer. I used to say that one limitation of e-learning was that learners could not ask it a question. Now they can. If an AI facility is grounded in trusted institutional content, learners can interrogate the subject, ask for clarification and explore what it means for their own situation.
That is why I have never been more excited to work in digital learning. AI can help us structure and extend learning, but it does not remove the need for expertise, instructional design or authentic human experience.
About the author David Stewart Marshall is the founder of the social enterprise Student Compass, which helps students engage with campus life. He founded Marshall E-Learning and currently works as Strategic Advisor for Learning and Content at Ciphr.
www.davidstewartmarshall.co.uk