AI & Training

Your Best People Know the Job. That Doesn't Mean They Know How to Teach It.

The gap between expertise and instruction is real — and it's why most internal training falls flat. Mereon's AI bridges that gap by turning what your people know into lessons that actually teach.

Jonathan TowellClaude
Jonathan Towell·Claude Opus 4.6
March 31, 20267 min read
Your Best People Know the Job. That Doesn't Mean They Know How to Teach It.

Think about the best person on your team. The one who's been doing the job for fifteen years. The one everyone goes to with questions. The one who can look at a problem and know exactly what's wrong before anyone else has finished reading the error message.

Now ask that person to train the new hire.

What you'll get — almost every time — is a brain dump. Two hours of everything they know, delivered in the order they happen to think of it. The new hire nods along, takes some notes, and three days later can't remember half of it.

This isn't a knock on your expert. They're brilliant at the work. But teaching the work is a completely different skill.

Knowing and teaching are different crafts

A master electrician can wire a panel in their sleep. But wiring a panel and teaching someone to wire a panel are two different jobs.

Teaching means making a series of decisions that have nothing to do with the subject matter itself:

  • What order do you present things? You can't teach someone to troubleshoot a system before they understand how the system works. Foundations first, then application.
  • Where will learners get confused? An expert has forgotten what it's like to not know something. A teacher anticipates the sticking points.
  • How do you make it stick? Not just say it — make it memorable. The right analogy at the right moment. A mnemonic for the steps that matter. A visual that shows what words can't.
  • How do you know they actually learned? Not "did they sit through the video" but "can they apply this on the job?"

This is instructional design. It's a real discipline with decades of research behind it. And asking your subject matter expert to also be a great instructional designer is like asking your best accountant to also be a great graphic designer. Different skill. Different training. Different craft.

What most companies do instead

Here's the typical approach: record the SME explaining things. Maybe record their screen while they walk through a process. Upload the video to your LMS. Write a few multiple-choice questions. Done.

The result is a video library, not a training program.

Learners watch the videos (or skip through them). They take the quiz (and guess on half the questions). The LMS shows a green checkmark. Everyone moves on. Three weeks later, the new hire makes exactly the mistake the training was supposed to prevent.

The problem isn't the content. Your SME's explanation was probably excellent. The problem is that explaining and teaching aren't the same thing. Explaining is one-directional — information flows from expert to learner. Teaching is structured — it sequences concepts, builds understanding in layers, checks comprehension along the way, and uses specific techniques to make information stick.

What good teaching actually looks like

Good training does two things that a recorded explanation doesn't.

It structures the material for learning

An expert explains things in the order they think of them. A teacher structures content so each concept builds on the last:

  • Foundational concepts first. You don't teach someone to diagnose a hydraulic failure before they understand what hydraulic pressure is.
  • One idea at a time. Not three concepts packed into one paragraph — one concept, explained clearly, with a chance to absorb it before moving on.
  • Application after understanding. Practice scenarios come after the learner has the building blocks, not before.

This sounds obvious, but it's the step that gets skipped most often. Most training content is organized by topic (the way an expert thinks about it) rather than by learning sequence (the way a new learner needs to receive it).

It uses learning devices

A lecture is just words. A lesson uses specific tools designed to help people understand and remember:

  • Definition callouts that isolate key terms so learners know exactly what matters
  • Analogies that connect unfamiliar concepts to things they already understand — "think of electrical resistance like water flowing through a narrow pipe"
  • Visuals that show what's hard to describe — a diagram of the system, a photo of the correct assembly, an illustration of the process flow
  • Flash cards for the things that need to be memorized — safety codes, equipment specs, step sequences
  • Comprehension checks after each concept — a quick question to confirm understanding before moving on, not a quiz at the end when it's too late
  • Application scenarios that put the learner in a realistic situation and ask them to apply what they just learned

These aren't nice-to-haves. They're the difference between "I watched a video about it" and "I actually learned it." Every teacher knows this. Most training software ignores it.

That's what Mereon's AI does

Here's where we come in. Mereon doesn't ask your SME to become a teacher. It takes what they already know — their video, their document, their SOP — and does the instructional design for them.

The AI reads the source material and does what a professional instructional designer would do:

It breaks the content into teachable chunks. Not arbitrary segments based on time or page count — concept-sized pieces that each teach one thing.

It reorders them for learning. Foundations first, then application. If concept B depends on concept A, concept A comes first — even if the SME mentioned it second in their video.

It generates the learning devices. Definition callouts for key terms. Analogies that map unfamiliar ideas to familiar ones. Memory aids for sequences and procedures. Comparison tables that clarify what's similar and what's different. Visuals that illustrate what words struggle to convey. Flash cards for the facts that need to be memorized. Comprehension checks woven in after every few concepts. Application scenarios that test real understanding, not just recall.

It adapts to the learner's level. A new hire learning the basics gets clear explanations, concrete examples, and straightforward recall questions. An experienced team member brushing up gets deeper application scenarios and critical thinking exercises. The same source material, taught at the right level for each person.

The result isn't a summary of your content. It's a structured lesson that teaches it.

The SME stays the expert. The AI becomes the teacher.

This changes the job. Your subject matter expert no longer needs to spend days building a training program. They do what they're already good at — explaining the work — by recording a video or writing up their process. Then they review what the AI built, adjust anything that needs adjusting, and assign it.

The shift is from content creator to content curator. The SME brings the expertise. The AI handles the instructional design. The learner gets a lesson that's actually structured to teach them, not just inform them.

Upload a safety training video. Get back a lesson with sequenced concepts, definition callouts, visual aids, practice questions, and application scenarios — all grounded in the actual content of that video. Not generic filler. Not AI hallucinations. Structured teaching derived from what your expert actually said.

That's the difference between a training platform and a teaching platform.

See it with your own content

The best way to understand what Mereon does is to try it. Upload a training video or document — something your team actually uses — and see what comes out.

Not a transcript. Not a summary. A lesson.

Start your free trial and see the difference.

Share this post