AI, Education, and a Pattern We’ve Seen Before

April 11, 2026

"The future ain’t what it used to be." ~ Yogi Berra

Seems appropriate that I’m heading to ASU+GSV tomorrow and finally putting some thoughts down on AI in education.

It’s a topic I haven’t written about yet, at least not directly, but it’s been showing up more and more in conversations, in the work, and in the questions people are asking.

A close friend sent me something recently because she said it was starting to show up in my work and in how I think about education and systems. I spent some time digging into it, did some additional research, and it got me thinking about something familiar. We’ve been here before.

There have been multiple moments where technology showed up with the promise of simplifying things. In many ways it did, but it also introduced new layers of complexity. If you go back to the early days of EdTech, most systems lived in silos. You had your Student Information System, your Learning Management System, curriculum tools, and assessment platforms all operating independently. I used to say all the time when I was in a district that we were data rich and information poor.

So naturally, the response was to try to connect it all. We went through the phase of building data warehouses, followed by a major push around interoperability. There were a lot of organizations doing really important work in that space, and a lot of that work is still ongoing today. In some ways, you’re just now starting to see pieces of it come together. But it took years, and even now, while things are better, the experience for the end user is still often fragmented.

That’s why AI is so interesting. For the first time, there’s a real possibility to move from fragmentation to simplification in a way that feels seamless, and the idea of true personalization at scale becomes much more realistic. At the same time, I’m starting to see a familiar pattern emerge.

A lot of these AI tools are being built in their own silos. Now you’re seeing conversations around partnerships and data access, which is a good sign, but we’ve been here before. Getting all of that data into one coherent system is incredibly difficult, especially at the scale education operates. Unless you move higher up, at the state level or beyond, and start thinking more holistically about data, it’s hard to fully solve, and even then there are real challenges that come with that approach.

There’s some great work happening across the country, but there’s still a long way to go. So when I look at AI, I see a huge opportunity, but I also see the risk that we fall back into a pattern we’ve already lived through.

I do think AI has the potential to simplify education systems, and that’s what makes this moment different, but it will only happen if the education community is intentional about how it’s built and applied. For me, it always comes back to the same question: are we truly keeping students at the center?

Almost 20 years ago, I started using the term student-oriented architecture. The idea was simple. Every system, every integration, every decision should be built around what’s best for the student experience, not just what’s easiest for the system. My team will probably remember me drawing the “generic kid” on the whiteboard in just about every meeting because I wanted that student in the room when decisions were being made.

That same thinking needs to apply here.

If AI is layered on top of disconnected systems without that alignment, it risks becoming just another layer. But if companies and schools work together, and actually leverage the interoperability work that’s already been done, there’s an opportunity to do this differently.

There’s also a responsibility that comes with it. AI is only as good as the data behind it, and while more data can lead to better outcomes, it also brings real implications around privacy, data protection, and trust. We have to get that right, and it’s not something any one group can solve alone.

It’s going to take the entire ecosystem. Schools, companies, and policymakers all working together with a shared understanding of what we’re trying to accomplish. That’s where things get challenging. Policy plays an important role in protecting students and families, and it should, but policy doesn’t move at the same pace as technology, and right now AI is moving fast.

So the question becomes whether we can create enough alignment across that ecosystem to move forward responsibly without slowing innovation to the point where we miss the opportunity, because there is real upside here.

If we get this right, AI could accelerate progress in ways we haven’t seen before. But there’s another side to this that is just as important.

If AI becomes the place people go instead of the systems themselves, the risk isn’t just technical, it’s directional. You can move faster and become more productive, but if you’re not clear on where you’re going, you just end up getting there faster.

There’s always a human element to education. There’s pedagogy, research-based instruction, and the reality of what actually works for kids in a classroom. AI isn’t going to replace that, but it will amplify whatever system it sits on top of.

If the system is strong, aligned, and focused on students, AI can accelerate that in a powerful way. If it’s not, it can just as easily amplify the wrong things. That’s why this moment matters.

AI won’t fix education. It will amplify what’s already there. The question isn’t how fast we move, it’s whether we’re pointed in the right direction.

It also doesn’t just amplify outcomes. It amplifies expectations, and not always in ways organizations are ready for.

That’s a conversation worth unpacking, and one I’ll come back to next week.