Do we still need subject matter experts to build learning?

The short answer is yes. The long answer is hell yes.

For years, the standard process has been straightforward: The SME provides the knowledge, examples, processes and expertise. The learning designer turns that into something people can actually learn from.

AI is starting to disrupt this model, and not necessarily in a good way, where LLMs are being positioned as the new SMEs. Some learning design platforms even offer an “AI as SME” capability. Give the AI your learning outcomes, audience and subject, and it can research and generate the knowledge that would traditionally have come from a subject matter expert.

But should AI become the source of the knowledge from which we build the course? It can research unfamiliar topics, synthesise information, generate explanations and identify key concepts incredibly quickly. Of course, it can also produce information that is plausible, confident … and wrong.

So how can we take what AI is good at and use it in our courses? Could we use it to build the initial knowledge base, then bring in the SME to challenge it, correct it and add the nuance that AI may miss? Instead of spending hours briefing the learning designer on everything they know, the SME could increasingly become the validator, contextualiser and source of expert judgement.

💡 The answer to whether we need SMEs is still a resounding yes. But now for a follow-up question: Can we hand over parts of their traditional role safely to AI?