By John Mullin, Enlearn CEO
A friend recently shared a music playlist with me. When my response was “meh…,” he suggested I put it on shuffle — that somehow a new sequence would overcome my disinterest. Still “meh”.
The issue wasn’t the order of the songs in the playlist. It just didn’t contain music I wanted to hear.
Unfortunately for students, this “shuffle the playlist” approach has the same limitations when applied to curriculum. The standard models for adaptive curricula today typically start with a single, fixed set of content (a text, a course) which is then re-sequenced based on a student’s progress and demonstrated understanding. This approach probably makes good economic sense by “personalizing” existing content for as many users as possible. But what if that content doesn’t contain what the student needs in the first place?
This is the underlying problem and limitation of current adaptive efforts. The content is limited to the concepts, content, student exercises, etc., that a specific set of authors, editors, and their publisher chose. To my knowledge, no single text or courseware has been able to cover the myriad learning paths and progressions best suited for every unique student, regardless of how the content is sequenced. If such texts existed, we wouldn’t have math wars, or battles between whole language and phonics, etc. Every student would already have the content they need for optimal learning. Currently, no matter how you slice it, no finite text will serve every learner’s needs.
Here’s an interesting illustration of this limitation…. Recently our partners at the Center for Game Science at the UW found through their Algebra Challenges that nearly 95% of students participating were able to reach concept mastery of linear equations. Some students needed as much as 6 times the practice to reach those mastery levels. The additional practice material that any 2 students required was different — it was specific to each student’s unique learning pathway and progression. This means providing 6 times the content, tailored specifically to the moment-by-moment learning challenges of each student. A fixed text is incapable of generating any new content, let alone 6 times the content.
What if instead we could start from the perspective of each unique student, and then create the content that works best for them, continuously? A virtually limitless text that learns each student, instead of vice versa — one that creates and then scaffolds the content with laser precision for each learner? A playlist that doesn’t just shuffle, but one that could generate new music for each listener?
This is what we’re doing at Enlearn. It’s called Generative Adaptation, and we’ll focus on it in future posts.