Everything discussed in AI Labs content is experimental, in progress, and not guaranteed to become part of the Degreed product suite.
Learning courses could never be completely personalized… right?
We’ve been talking about personalized learning for a long time. Yet, when most people log into a course today, they get the same content as everyone else, irrespective of job function, department, or knowledge base. It doesn’t matter whether they’re a beginner or a highly trained expert.
That gap between what we talk about and what we actually deliver in terms of personalization in development is what sent me down this particular rabbit hole: What if a course was designed instantly with you, the learner, at the forefront? What if it built itself, in real time, around who you are, what you already know, and where you’re trying to go?
We wanted to try doing this in an AI Labs experiment.
Why Static Courses Break Down
Think about the last time you slogged through a course that ended up being mostly a review of material you already knew. Or one that jumped right into the deep end, leaving you feeling like you missed some critical, foundational learning before you started it. Both feel like a waste of precious learning time. One is a slow walk when you are ready to jog; the other is sprinting away while you’re trying to get your running shoes on.
Pre-built courses are designed for an imaginary average learner that, let’s be honest, probably isn’t you.
Personalization has been the theoretical answer to this for years. Adaptive learning, branching scenarios, curated pathways: all attempts to solve the same thing. But they’re still built on a fixed foundation that someone, somewhere, decided on in advance.
In this experiment, there is no pre-built course.
Building the Course in the Moment for the Need
So, where do you start, if not with a pre-authored content module? You enter into an AI learning conversation. The AI persona will provide options for topics you can talk about that are already established areas for dynamic learning. Then, you can choose your own adventure.
As that conversation unfolds, the course builds around you: visuals, quizzes, text, reflection questions—all generated in real time based on where you are in the conversation and what you actually need to learn.
Plus, it’s adaptable, in the moment and as you build it. You can ask it to skip sections or focus more in a certain area. If you’re confused, it changes its approach. If you want to go deeper on something, it goes there. If you need reinforcement, it generates questions to help you cement the concept. The course is less a destination and more of a continuous dialogue.
In a way, this mimics the new cycle for learning, where upskilling is continual, rather than episodic.
A Live Example: Learning About Model Context Protocol (MCP) on the Fly
To give you a concrete feel for it, let me walk you through what this looked like when I used it myself.
I decided to use this to explore Model Context Protocol (MCP), which is something Degreed has been building directly. We have our own Degreed MCP server, so I have background and context, but I’m genuinely curious about where this technology is going. It’s also a good test because it’s complex and has a lot of layers to understand.
First, the AI prompted me with a few topic options. I chose MCP fundamentals. As the first diagram loaded, it asked me what I already knew about connecting large language models with external tools.
That’s a key moment. It wasn’t starting from zero. It was calibrating where I already was.
I shared some context about Degreed’s work with MCP. The response picked up on that immediately. It didn’t re-explain the basics I’d already implied I understood. Instead, it moved to the mechanism: how MCP creates structured communication channels, separates intent from execution, and enforces schema in a way that raw API calls don’t.
When I pushed back with a real question (Why use MCP rather than just having the model call APIs directly?), it answered clearly and specifically. Type safety. Separation of concerns. Predictability. Then it moved into the next layer of depth without me having to ask.
After the technical content, it checked in: Would I like to go deeper? Watch a more detailed video? Take a quiz? Or shift to a reflection question?
The course didn’t just keep running. It asked what I needed next to progress.
How Are Dynamic Courses Different from Traditional, Adaptive Learning?
Most adaptive learning still works from a pre-authored map. The system decides which branch you take based on your answers, but the branches were written in advance. The experience is personalized in the sense that you might take a different path, but the paths exist before you arrive.
What we’re exploring here doesn’t work that way. This is real-time generated content created in response to you, and the visuals appear naturally as the conversation progresses. The questions are written for your specific thought processes and moments of confusion. Traditional adaptive learning can’t do that.
What This Could Mean for Learning
I want to be careful not to overstate what we’ve built here. This is an experiment. A lot of what comes out of AI Labs stays in AI Labs.
But I do think this points to something important: The most common complaint about workplace learning isn’t access to content. It’s relevance. Almost 80% of employees say they want learning that’s directly correlated to their responsibilities. Yet, how can they get that without the learning having specific, personal context into their existing skills and their goals? Learning teams, talent directors, and managers don’t always have time to close that gap manually, for every person.
That’s not scalable.
A course that builds itself around the person experiencing it is one answer to the problem. Not because it’s novel (novelty fades), but because it’s more likely to be useful and effective, quickly. Workforces need a way to keep pace with the changing world, and that means thinking beyond traditional learning methods to mobilize relevant, personalized learning experiences faster.
That’s what real personalization at its best could actually offer. It’s a fun future to imagine.
If you’re curious about what we’re building, reach out to me on LinkedIn. I’m always happy to talk through it or even give you access to test some of our experiments yourself.
Watch the recording of my full LENS 2026 session on our AI experiments.
Unser heißer Draht zu Ihnen.
Exklusive Einblicke, Infos zu bevorstehenden Veranstaltungen, Updates zu Degreed-Lösungen – das alles erwartet Sie in unserem monatlichen Newsletter.