Designing AI-supported learning for real moments of work

For a long time, learning has been packaged into classrooms and course content  even though we’ve always known that’s not where most learning actually happens.

Some of the most meaningful learning happens in the flow of real work:

• when someone is about to have a difficult conversation
• when they are writing an important email
• when they are making a decision under pressure

It happens in moments. The challenge is not lack of experience.

It’s that these moments often pass without being used as learning opportunities.

Yet decades of research in learning science point to something important:

• John Hattie highlights feedback as one of the highest-impact influences on learning and performance
• David Kolb reminds us that experience alone isn’t enough. Learning happens when we reflect on it

For years, we’ve known learning should be:

• embedded in real work
• timely
• contextual
• personalized

But we were constrained by logistics, complexity, budgets, and capacity.

Now those constraints are shifting,  largely because of AI.

We’re already seeing this inside organizations:

• AI tools giving real-time feedback on communication and tone
• AI coaches supporting practice of conversations and decisions
• Digital assistants guiding employees at the moment of need
• Systems analyzing real interactions and generating coaching insights

This is no longer theoretical. It’s learning happening in the flow of work. But this is where I see a critical distinction.

This is not about:

• personalizing learning paths
• or making content delivery more efficient

It’s about something deeper:  supporting learning in the moment it is needed

Because that’s when learning is most powerful.

Imagine this:

You’re about to have a difficult conversation with a colleague.

That moment carries more learning potential than any case study.

Now imagine support at that exact moment:

• a prompt that helps you clarify your intention
• a reminder about tone or framing
• a quick reflection question before you speak
• a follow-up nudge after the conversation

That’s not a course. That’s learning embedded in experience. This requires us to rethink content design.

Not just:

• What content should we create?
• How do we deliver it?

But:

• Where does this content become relevant in real life?
• What support would help in that exact moment?
• How do we guide people to use it when it matters?

For a long time, we’ve been sharing content and hoping people would apply it later.

But most people were never really taught how to learn efficiently  (including many of us in L&D.)

At the same time, information is changing rapidly, and the pressure to reskill is increasing.

We don’t have the luxury of waiting for people to figure this out on their own.

Because of AI, learning is starting to show up inside the work itself:

• writing support inside email tools (tone, clarity, structure)
• real-time coaching for leadership and communication
• AI assistants helping employees navigate tasks
• practice environments with immediate feedback

So here’s the question :

Are we using AI to make content delivery easier or are we using it to support learning in the moment it’s needed?

And even more importantly:

Are we designing AI agents to support real learning moments or just more efficient content?

One simple way to start:

Take a real situation your learners face (for example: a difficult conversation)

Instead of sharing content, design a small support mechanism for the moment itself:

• prompts
• reflection questions
• AI-assisted feedback
• decision guides

We’ve always known experience is the best teacher.

Now we finally have the tools to design learning inside the experience.