Michael Horn argues that educational technology only works when schools actively manage distraction and align incentives. Research shows personalization can boost achievement, but only 5% of students use ed-tech tools as intended. Alpha School demonstrates the alternative: monitoring systems and earned rewards for passion projects push students toward effortful learning instead of endless scrolling.
So personalization, mastery learning, and some uses of education technology. They do show strong research results. And yet schools clearly are not improving at the scale you'd imagine. David Deming at Harvard in his substack recently highlighted compelling evidence that personalization with tech can drive some big learning gains. But he also noted some real world barriers, distraction and cognitive offloading, when devices effectively do the thinking for students. And this points to me to a much deeper structural issue. As Daniel Willingham always reminds us, learning is effortful. And with technology, you're handing students a device that, yeah, it's got some cool learning tools, but it also delivers endless distractions just a click away. So what happens? Well, Lawrence hold has documented the 5% problem that a lot of these ed tech programs work in research, but only a small fraction of the students use them as intended. And as I've always written to me, that's why the learning model matters way more than the technology. Layering tools simply onto traditional schools without addressing distraction, it just never gonna produce the gains that we see in some research. And I've personally come to believe that I've also missed something in a lot of my writing. So a mea culpa here, which is that I think the model has to actively account for distractions because when you give students these general purpose devices that many places are, the result is just billions of distractions. And it's what makes what Alpha School is doing so interesting, I think because what they're telling us is that they're not using smartphones, they're not using just plain old Chromebooks during their core two hour learning time. Instead they have this vision model on the computers that is actually monitoring what students are doing, then the guides can actually act on that data, Then the incentives align so that if students are not doing their core learning, then they're not going to get their time back for the areas of passion projects and so forth that they're really excited about. And it seems to me that these guardrails plus the incentive structure earning that time back is what makes that powerful combination. So look, ed tech, it may be powerful in theory, but students aren't learning in theory when it's just simply layered onto systems not built to optimize learning. Students predictably are shifting away from effortful learning to effortless scrolling.
