In our season seven finale of Class Disrupted, Diane Tavenner and I reflected on the conversations about AI that we’ve had over the second half of this season—from those focused on new schooling models to new tools. This felt like a no holds barred conversation—candor and lots of back and forth as we talked about how much of the education sector is still operating under the traditional model, the challenges of creating learning experiences outside of that model, and the tensions between innovation and entrenched systems. We discussed issues like policy backlash against technology in schools, the value of outcome-based contracts, and whether microschools or low-cost private schools can actually drive large-scale change.
This was a long conversation—but I thought one of our best. We look forward to your thoughts.
Diane Tavenner
Hey, Michael.
Michael Horn
Hey, Diane.
Michael Horn
We’re in person.
Diane Tavenner
I know.
Michael Horn
Two episodes in a row.
Diane Tavenner
It’s amazing. For those who are only listening, we actually are sitting together here in Boston, which is super fun. It’s good to be on this coast with you. And for me, the best part of this is I got to have family dinner last night at your home. And I woke up this morning thinking about it. I just felt really happy. It was joyful. I think you know that my favorite night of the week is family dinner night at my house.
Michael Horn
Yeah.
Diane Tavenner
And I hustle home, so I don’t miss it. And it was so fun to be with your girls and with Tracy. And I was thinking a lot about those few hours and, like, the learning that was happening at that table. I mean, we got to play this board game that your girls had, like, invented, and the dialogue and the conversation and the curiosity and that, you know, they were making their own lunches, which called back to, like, Rhett.
Michael Horn
Yeah. But he was doing that from a young age. I remember…
Diane Tavenner
Yeah, yeah, yeah.
Michael Horn
Preparing meals and shopping and everything.
Diane Tavenner
Totally. And so there was. That was really fun. And to be back in that sort of younger age, because mine are older, I should say. The thing I kept thinking is, like, this is what learning is, and what if. What if this is what it felt like? Which doesn’t. I wish everyone could have this experience.
And yeah. So that. That was good.
Michael Horn
No, it’s really cool. I will say. We also then had a conversation this morning about. She was like, so what night is family dinner night in the Tavenner household? And should we think about creating a similar structure as our girls? Hold up. So you being there also caused us. Because I think it’s intentionality also.
Diane Tavenner
It is.
Michael Horn
Right. And that intentionality we both brought to raising our kids and now thinking about reinventing schools. And I think it was interesting because in our previous episode, we were live as well.
Diane Tavenner
Right.
Michael Horn
We were at the ASU-GSV Summit with Reed Hastings. And it was interesting to hear Reed, I think, for the first time in my memory, and I think that’s what he talked about. We actually need to reinvent the classroom model, the school model itself. We can’t just be tinkering, if you will, toward utopia or layering over here and whatever else. That’s obviously the tune we’ve been singing for most of our professional careers.
Diane Tavenner
Right. But it was fun. It was exciting to hear him talk about that and reflect on the work that he’s done in different, you know, elements of education. And to really start to zero in on. We have to work beyond just the classroom if we really want to transform the experience for young people. And so, I mean, I guess it always feels good to be confirmed and validated.
Michael Horn
Sure. We feel like geniuses now.
Reflecting on mid-season insights
Diane Tavenner
Yeah. So I think our plan for this today is a callback to mid season, where we did this in person also kind of mid season. Oh my gosh, we’re learning so much. Let’s get together and just reflect and process and talk about what we’re learning after that first half of the season where we were really talking to sort of thought leaders and people who are thinking big picture and who understand AI and in AI and so we thought we’d do that again, where the back half of our big AI exploration season has really been with practitioners, I would say. And so what are they? What are we making? We had all these really fascinating conversations. This is our first time to really sit down and like, sort through them and process them and understand what we heard. And then we’ve also been thinking about next season and so we’ll share a little bit of that
Michael Horn
Tease a little bit. It’s been interesting and I’m glad. Process is the word that I’ve been thinking about because I think when people appear on our podcast, this sounds silly, but it’s not an endorsement of their model or their tool or whatever it is by us. It’s a genuine learning opportunity for us. And obviously we were both interested in new school models.
Diane Tavenner
Right.
Michael Horn
Which was three episodes that we had. And then we were interested in the ed tech tools that are coming up and using AI in different ways. And we were trying to process. We had this framework that you introduced in the last time we were in person in Cambridge or in Boston about the sort of horizon one, horizon two, horizon three, schooling models and the way I think we’ve sort of categorized it, there was H1 is sort of the industrial model that we know and you had H3 as sort of this post industrial model. And H2 is sort of this hybrid that is elements of both maybe and so forth.
Diane Tavenner
And so I love that metaphor of the steamship that is outfitted with sails still because it has to be running both because steam is not reliable enough. You can’t really get all the functionality you need and trust it. So you have to still bring the H1, if you will, into it. I’m glad you brought up the framework because that really shaped how I saw the second half of the season and had all those conversations. I’ve actually been thinking a lot about that framework and how my understanding of it is sort of, I don’t know, maybe becoming a little bit more nuanced as we talk with people and think about these different models. And so that might be a place, a place to start.
Michael Horn
Yeah, I think let’s dig into that because. And you can take it wherever you want. I think it’s a framework that I have found both useful and I’ve been struggling with, like, what is this H3 really gonna look like? Where does it really depart from the grammar of schooling as we’ve known it? And do we really think we’re gonna actually get there? Or are there gonna always be this sort of H2 pulling us some of the elements, like where. And then I guess I’ll say the last thing, which will probably come up as we talk about the school models that we had on the podcast. Where, you know, where are parents in all this? Like, where is their headspace? Cause like what we want, where parents are, where schools are. And I, for those that can’t see my hand gestures, I actually think that’s the continuum at the moment. Like historically, I think parents were more conservative than the schools, I think that may have flipped.
Diane Tavenner
Well and I think there’s some evidence of that with the growth in microschools and sort of alternatives like, people really sort of self bundling, if you will, you know the homeschooling.
Michael Horn
I was shocked by the way from the homeschooling I read this morning in the 74. 10 million. I know 10 million families. They reported from this RAND survey that Johns Hopkins had done. I think it was Angela had done there. And that’s a big number.
Diane Tavenner
It’s a big number. And I think it speaks to this idea of where are parents and I agree with you, they’ve historically been more conservative. I mean, I wrote prepared to try to, you know, I call it a love letter to parents, like to introduce them to some of the ideas of what’s possible. Because my, my feeling was like they want one thing for their kids, but they don’t know how to get it. And so they kind of double down on, you know, the existing model, if you will. I actually think we’re starting to see, at least for some, a break from that. And not just a small amount, but, you know, a growing.
Michael Horn
In some cases quite big. Yeah, in some cases there’s, there’s steps and I actually think that comes up. I don’t know if you want to stay on the H1, H3 or go into the model.
Diane Tavenner
I think what I would say is just a couple of reflections and observations. So kind of how I’ve been using that framework in real life. So one I find that it is helping me be engaged in conversations with educators and developers. All the people who are in, who care about this work and who are working in it, who are thinking about AI in that if I can just set the framework out up front and say look like my passion is developing the post industrial model and H3 and honestly I don’t think I’ve seen it yet and I don’t think it exists yet and that’s not a bad thing. That’s fine. But let’s not pretend that certain things are H3 or that they are a real new model. We might have that conversation in the models conversation when I don’t think they are. And you brought this up the last time we talked like we also need tons of kids are in H1, they’re going to be in H1 for a while.
How do we use AI to make those experiences better? So I think when I’m able to say to people which conversation are we having? Where are we working right now? And the key for me has been how do we not constrain or compromise work in developing the post industrial model or the H3 place by working on H1. And that is a real risk I think because I mean there’s so many like elements of H1 policy and norms and practices and tools that just codify and solidify that model and then therefore prevent unintentionally the invention of new models.
Michael Horn
And we talked about that some in the mid season in terms of you named assessment and special education in particular in those I think there’s a lot more and frankly in the last
Diane Tavenner
I’m going to add a new one
Michael Horn
Since we recorded I think we have seen the growth, the fastest growth in my memory of a series of policies around screen time. Cell phones were already emerging but now into screen time in general that I don’t love but well intentioned in H1 that are like diametrically opposed to I won’t call them H3 but I will say like these new models of schooling that are appearing by education entrepreneurs in sort of this H2, H3 nebulous area.
