Article
The Edtech Blind Spot: Crowded Into a Small Corner of Education, While the Biggest Pains Go Unsolved
What is a school still for, once learning can happen anywhere? That question runs underneath Jennifer Carolan's recent interview in The Humanist, and I haven't been able to put it down. I came to edtech the way Carolan did, through the classroom. She taught before spending two decades backing education companies at Reach Capital, I spent a decade in education before leaving to build for it. So I didn't read it as market analysis. I read it as someone who already knows, firsthand, how disconnected the system really is.
She refuses the easy stories about where this goes. Edtech got burned and never worked. AI tutors will finally fix everything. Schools are too broken to save. The classroom is sacred and technology ruins it. She picks none of them. Her actual point is quieter and truer: schools have been slowly losing their monopoly on knowledge for a long time, AI is just speeding it up, and the part of school worth protecting was never the information anyway.
Here's the thing I've staked a company on: a learning outcome is never just a classroom problem. Schooling used to be the same thing as learning, that's over. Kids learn everywhere now, and what happens in any one room is shaped by a dozen things outside it: how the school is run, whether departments talk to each other, whether anyone can actually see how the place is doing. The failures we pin on the classroom usually start somewhere up the chain. There are too many disconnections at the ecosystem and organizational level, and until those get solved, no classroom can flourish, even a great teacher's impact leaks away before it can compound.
Which forces the harder question, the one I think actually matters: if learning no longer lives in school, is school still a meaningful place to be? I believe it is, just not for the reason it used to be. Delivering information has left the building; the internet, and now AI, do it better and faster. What's left is the thing schools alone do: they form people, and they push them. School is where a kid is challenged to think at a higher standard than they'd hold themselves to alone, where they learn to belong to something, to work across difference, to become a citizen. Carolan says it cleanly, schools are "not just educational institutions; they are microcosms of society," and "one of our most important reminders of why we still need each other." The role isn't disappearing. It's shifting, from delivering content toward preparing "citizens who can thoughtfully participate in shaping the future."
Which is exactly why the disconnection matters so much. Carolan sees the same fault line: "like so many institutions right now, our schools are fighting for relevance," with students who "feel disconnected from what they're being taught." I'd push it one level up, what a student feels in the room is the downstream symptom of a system that's come apart around them. The part of school worth keeping is precisely the part a broken system can't deliver.
The decline is real, and it's structural
The decline Carolan points to isn't just a feeling, it's in the numbers, and it isn't only an American story. Birth rates have fallen below replacement across most of Asia and Europe; Japan's population is projected to drop from 128 million to under 53 million by 2100. Fewer children means shrinking and closing schools, from rural Japan to the rural US, where public enrollment fell by more than a million students between 2019 and 2023 and is set to keep sliding through 2031.
At the same time, families who can are opting out. International schools have now passed 15,000 worldwide, teaching 7.6 million students, enrollment up 13% in five years, with 58% of them in Asia. In the US, that same exit runs through education savings accounts, where roughly 40% of K-12 students can now spend public dollars outside the public system. Different mechanism, same move: leaving.
Faith in the next rung is slipping too. In the US, the share of adults who call college "very important" fell from 75% in 2010 to 35%, and when the median Black borrower still owes 95% of their loan two decades after starting, who can blame them. And the people holding the system together are stretched thin: UNESCO warns of a worsening global teacher shortage heading into 2030.
The line that stuck with me was that the system was "built for stability." That's exactly the problem, and she's blunt about where it leads, it's a losing battle, she writes, "a Sisyphean folly to believe that schools, built for stability, can outpace this fast progressing technology." Standardization, seat-time, everyone moving in lock-step, those were the point once. Now parents want the opposite: pace that fits their kid, support for how their kid actually learns. The factory model isn't failing because people stopped caring. It's failing because what families want changed and the system can't bend. And teachers are the ones absorbing the difference.
The real problem is upstream
Bending isn't really a classroom act. A school flexes, or doesn't, at the level of how it's organized, governed, and connected. Brookings, in its work on systems thinking, calls these the "levers of organizational change" and argues they sit inside the institution, not the lesson, which is why the digital revolution is, in its words, "a radical change that many education systems fail to harness."
And the levers don't stop at one school. The same disconnection runs through the whole system: between schools and the accreditors and regulators that judge them, between a school and the network it belongs to, between all the parts that are supposed to add up to a child's education and rarely do. This is the layer that decides whether everything beneath it works, and almost nobody builds for it, because it's a hard, boring bone: organizational plumbing, not a slick classroom app.
The money didn't leave education. It lost patience for vague.
Carolan pushes back on the idea that everyone's abandoning edtech, and she's pointing at something real. The capital didn't leave, it moved. The dollars that are left are going somewhere specific: AI, workforce training with clear payoff, anything with a measurable outcome.
In my view, even that money isn't aimed deep enough. Measurable isn't the same as fundamental. Tutoring, training, point solutions, they're easier to fund because the outcome fits on a chart, but they still sit on top of the same disconnected system. The capital got more disciplined. It didn't get more structural.
The "edtech" label is cold right now, but that was never the real problem. The real problem is that the most valuable work in education is also the least visible, and capital only funds what it can see. None of this is to say the obvious, classroom-scale problems aren't worth solving, they are.
Go a layer deeper, into the system itself, and the generalists back off. From the outside the system level is invisible, and if you can't see the problem you can't underwrite it. So the capital pools around what's legible, what fits on a chart, and the hardest, most fundamental work stays starved. The founders building it have nowhere obvious to turn but the handful of specialist investors who know education from the inside, and there are so few of those that the rest of us get lost in the noise.
Why I went for the hard bone
I decided to go straight at that hard bone anyway. Not because it's the fundable thing, but because it's the thing that actually has to be fixed, and because the disconnection I care about doesn't change when you cross a border.
The work itself is slow and unglamorous. The foundational layer of a school, whether it can give an honest, ongoing answer to "how are we doing?", takes years inside the system just to see clearly, and there's no shortcut around that kind of domain knowledge. But it's the layer where impact actually compounds. Get it right and a school can finally see itself: improve on purpose instead of by accident, and put its energy where it changes outcomes for kids.
It's the layer where the disconnection actually lives, and left alone, it eventually eats the classroom too: every disconnect upstream lands as one more login, one more form, one more thing stealing a teacher's time. So the work isn't to add another tool to the pile. It's to subtract, to take chaos out instead of layering more on.
The people holding it together
The whole point of school is, and will stay, the human, communal part. If it were only about information, we could all just retreat into our own corners with our own devices. What we believe at Novana is the opposite of that retreat: that the system should help a school make sense of what it's doing, what it's aiming for, and how to actually get there, so the human part has the room to happen.
And here's where I land hardest with Carolan. She believes in tools that lift teachers up instead of replacing them, because "the human connection between teacher and student remains the most powerful lever for learning." She's even done the math: a teacher works around 53 hours a week, and roughly a third of that goes to lesson planning, grading, admin, and communication, the exact work AI is now good at. "What if AI afforded teachers more time to create and lead magical, human-centered learning experiences," she asks, "and less time on tasks that cause burnout?"
For me that question isn't a nice sentiment, it's the spec. Every hour we take off a teacher's plate is an hour handed back to the part of the job no software will ever do. The goal isn't to automate the teacher. It's to stop making the teacher do the robot's work, and to clear the disconnection above her so she can actually teach.
Teachers and students are the center of education, and because they're also its most visible part, they're where edtech keeps piling on, addition after addition, overwhelmingly. What almost nobody builds for is the people accountable for the whole thing. Everyone knows the classroom is hard; far fewer see how heavy it is to lead a school, and fewer still the layers above it: the groups and networks running dozens of schools at once, and the accreditors, regulators, and education systems steering thousands. It's the part of education most outsiders have never even heard of. I watched it across my whole career, leaders at every level answerable for a quality they could never clearly see, steering on fragmented information, worn down in a way no one names. How this enormous system actually navigates, day to day, gets a fraction of the attention it deserves. That's the silent bleeding all the way up, and the part I couldn't walk away from.
Before you build the future of school
Carolan ends by telling founders to go become a substitute teacher. Sit in a classroom three days straight. It's boring and exhausting, she warns, and that's the point.
I didn't have to take that advice. I lived it first, a decade in education before I wrote a word of a pitch deck. And it does something no market research can: it makes you allergic to the wrong kind of ambition. Once you've watched a great teacher lose a whole evening to compliance paperwork nobody will ever read carefully, you stop wanting to build the thing that impresses investors and start wanting to build the thing that gives the evening back.
That's the real opportunity right now, not just the risk. Building has never been easier, students ship working AI tutors in week two of class, but easy to build was never the same as worth building. The frontier that's genuinely open isn't one more tool for the classroom; it's everything upstream of it that almost no one has touched. So if you're a founder drawn to education: go deeper. Past the classroom, into the system. The hardest, most fundamental work is still sitting there, unbuilt.
And if you're an investor tempted to read the funding winter as proof that edtech was the wrong bet, I'd resist that conclusion. Education is one of the largest, most durable markets there is; we all know it. The money pulled back not because the opportunity shrank, but because the easy version got crowded and the real version is hard to see from the outside. That's a reason to learn the space more deeply, not to walk away from it. The biggest opportunities here are exactly the ones that take the most understanding to spot.
Because the problem was never just the classroom. It's everything we built around it and never connected. Fix that, and you don't just tidy the system, you give every classroom inside it a real chance to work. That's the work still waiting, and it's where the impact has been hiding all along.
Further reading: Jennifer Carolan's interview in The Humanist ("What Edtech Got Right, What It Missed, and What AI Changes Now") and her Reach Capital essays "AI and the Role of the Teacher" and "School AI Policies Mask a Bigger Question"; Brookings, "Systems Thinking to Transform Schools"; EdSurge, "Too Many Tools, Not Enough Impact"; EdWeek, "The Ed-Tech Backlash Is Here."