Three Years After ChatGPT: What Has (and Hasn’t) Changed in Our Schools

Three years ago, the release of ChatGPT captured the world’s attention. It wasn’t the first artificial intelligence tool, nor the first machine-learning breakthrough, but it was unquestionably the moment when AI became part of the public imagination. It was heralded as the technology that would change everything — and for a moment, many believed it would. I was at NEOM at the time and we had already integrated AI as a core competency in our framework. I remember a flurry of texts from colleagues around the world sharing their experiences of urgent faculty emails, hand-wringing administrative meetings, and parent concerns that all signaled that education would never be the same.

And yet… three years later, most of us would say that while the impact has been real, it has not been the transformation some feared — or some of us hoped for.

This tension reveals something important about our schools. As Malcolm Gladwell reminds us, tipping points don’t happen just because a technology exists; they happen when social systems are ready to respond. And in our schools, our systems — the structures, mindsets, and habits that define how school “works” — have never been built for rapid innovation. They gravitate toward stability, predictability, and tradition—appealing comforts that quickly become liabilities in a BANI world (See my last Blog).

Why the "Revolution" Felt Smaller Than Expected

Part of the story is that we consistently overestimate the short-term impact of emerging technologies. We imagine change as a clean break — an overnight shift. But as the Heath Brothers and James Clear often point out, sometimes the most meaningful change is the opposite: a series of small moments, nudges, and re-designs that accumulate over time.

The other part is that our school systems were never designed with agility in mind. So while AI has advanced rapidly, schools have been slower to evolve. As Brené Brown might say, fear of uncertainty often leads us to "armor up" rather than open up — and armored leadership rarely paves the way for innovation.

Still, we have seen some action and impact:

• Teachers experimenting with AI for feedback, differentiation, and planning.

• Students using tools to explore, design, draft, revise, and iterate.

• Schools beginning to imagine new workflows, new literacies, and new expectations for what it means to learn.

  
  

Jon Gordon writes often about the power of intentional positivity — the idea that progress requires choosing purpose over panic. Three years later, the educators who have leaned into experimentation rather than fear are the ones showing us what’s possible. We have to get more comfortable being uncomfortable but also remain positive forces driving experimentation and action in our schools.

A Better Question for Year Three

So as we mark the third anniversary of ChatGPT, perhaps the most important question isn’t What has AI done to schools? but What have we done with it?

Simon Sinek reminds us to “start with why.” For our schools, the why has never been to adopt technology for its own sake. Our why is authenticity in learning. Agency for every student. Systems that can respond — not react — to a world changing at exponential speed.

Which means the real work ahead is not about the next tool or model. It’s about redesigning schools so that emerging technologies can actually help us build the environments our students need and deserve.

Three-Year Check-In Questions for Schools

As AI moves into its next chapter, here are the questions that matter most:

• How are we using technology to deepen—not shortcut—our mission, vision, and values?

• Where are we expanding authentic learning and student agency in the learning?

• What outdated systems and/or structures are keeping us from becoming more adaptive, agile, and future-ready?

• How are we equipping students with not just tools, but mindsets—curiosity, courage, and discernment—in a world shaped by AI?

• And perhaps the most important: What bold experiments are we willing to try this year?

Adam Grant often calls on us to “think again” - treat our ideas as hypotheses, our practices as prototypes, and our systems as works in progress rather than monuments to tradition. Three years into the rise of AI in education, that is precisely the invitation before us. AI isn’t asking us to predict the future; it’s asking us to participate in designing it. It pushes us to reexamine long-held assumptions about learning, teaching, assessment, and the role of schools themselves. It challenges us to let go of practices that no longer serve students and to experiment courageously with what might.

But none of that begins with technology.

It begins with the questions we ask.

Schools that thrive in a BANI world won’t be the ones that cling to old certainties—they’ll be the ones that cultivate confident humility, that model rethinking as a skill, and that create cultures where inquiry is not a disruption but a discipline.

AI gives us unprecedented tools, yes—but more importantly, it gives us a reason to ask better questions:

• What if learning were designed around curiosity rather than compliance?

• What if assessment captured growth instead of gaps?

• What if school structures were personalized for student needs instead of forcing students to flex around them?

• What if we let go of “the way we’ve always done it” long enough to imagine something better?

  
  

Grant would argue that rethinking is not a threat; it’s a responsibility. AI simply accelerates the urgency.

Looking Forward

ChatGPT or the rise of AI haven't transformed education over the past three years.

For my entire life, countless “revolutionary” technologies—from filmstrips and televisions to scantrons, smartboards, laptops/tablets, Google, and MOOCs—have been hailed as the thing that would finally transform education.

If we want future technologies to truly elevate learning rather than simply layer onto old habits, school leaders must carve out real space to examine the systems and structures we’ve inherited. Many of them were designed for a different era, with different expectations, constraints, and definitions of success. Reimagining them isn’t about chasing the latest tool—it’s about creating the conditions where any advancement, including AI, can be harnessed for deeper inquiry, richer collaboration, and more meaningful engagement. The schools that thrive in the years ahead will be the ones willing to step back, question the architecture of “how we’ve always done things,” and intentionally design environments that are fl

exible enough, human enough, and future-ready enough for whatever comes next.