The real question isn't whether to use AI. It's how.
There's been A LOT of talk about AI-generated eLearning courses recently. Some people are happy (“we can save money and time!”), some are not (“my job is on the line!”).
I’m not sure it’s so binary.
I get why people are happy - the idea of going from brief to published course in minutes is genuinely appealing.
But what do those courses actually look like, and how effective are they?
AI can only work with what it has. In most authoring tools, that means standard content blocks - text, images, accordions, quizzes. The result is usually a course that's content-heavy, visually predictable, and designed to be completed rather than to change behaviour.
For some organisations, that's genuinely a good place to start. A small company with no L&D team, with SMEs under pressure to get something out fast? An L&D team drowning in requests that need a first draft to react to? There are real use cases there.
But at Learnopolis, we design bespoke learning for our clients - which means an AI-generated draft isn't a starting point for us. It's a different product entirely.
What we're focused on is what happens when an instructional designer uses AI as a tool, not a replacement. For me, that’s what AI helps me build within a course: code blocks in Rise … something that an AI generated course won’t include.
Recently, we built an interaction for a client around a five-step process. The learner works through each step, writes their action plan as they go, and downloads it at the end to use back on the job.
Previously, that would have required Storyline … and many hours of build. Now, I can describe exactly what I want, and AI will code it for me, ready to pop into a code block in Rise. (We all know it’s not quite that simple and quick … but you know what I mean!) Add in Mighty Studio, and I have a template that can be adapted for a completely different process with a different client within minutes.
To me, that's the shift … and why we instructional designers aren’t dead yet. It’s not about AI replacing the designer, but tools giving designers the ability to build interactions that genuinely serve the learning - whatever that looks like for that content and that audience. And for some, that starts with an AI generated course that they can improve, rather than a blank slate.
AI on its own can produce a course. But a course that changes behaviour? That still needs a human being who understands what the learner actually needs to do differently when they walk away.
What's the most useful thing AI has done in your design process - and what's the thing it still can’t do?
Nicola




Comments