Information these days is cheap. When unfolding events travel across social media at the speed of light, breaking news is stale news by the time trained journalists have checked their facts and crafted their story arcs.

Yet while Twitter and YouTube have made information a commodity, insight is still a scarce and valuable resource. It’s one thing to know what’s happening, it’s another to know what to make of it. The hottest hot take isn’t necessarily the shrewdest. 

In that vein, I consistently find that Clayton Christensen’s innovation theories offer a powerful lens for making meaning of the happenings in the world. The theories can help us see when hyped trends lack the necessary conditions for success and when under-rated innovations have potential to change the world. Below are two recent headlines that have caught my attention, followed by some analysis based on innovation theory.

“ChatGPT Is Landing Kids in the Principal’s Office, Survey Finds”

The74 recently shared results from a survey conducted by the Center for Democracy and Technology into how students are using ChatGPT. There’s a lot of interesting information in the survey results, but the bit that grabbed my attention was the finding that students are using ChatGPT more often as a support for navigating anxiety, mental health, and issues with family and friends than as a tool for cheating on assignments. This pattern is a striking example of nonconsumption, and a telltale indicator of a potential path for disruptive innovation.

Nonconsumption refers to situations where conventional products or services are inaccessible due to factors like high cost, inconvenience, or complexity—and it offers one of the prime initial footholds for disruptive innovations. Early versions of disruptive innovations aren’t as good as established solutions on conventional metrics of performance. But they have other benefits—such as affordability, portability, or ease of use—that make them accessible in situations where there is nonconsumption. Nonconsumers are delighted by crummy early versions of disruptive innovations because having something is better than nothing.

Student mental health is a growing crisis, but therapy is largely inaccessible for many students. 

Therapy is expensive. And even when cost isn’t a barrier, therapists aren’t usually available in life’s challenging moments when the need for support is most poignantly felt. Meeting with a therapist can also carry a stigma that makes in-person meetings difficult; and finding a therapist who you connect with can also be a struggle. In short, there’s lots of nonconsumption of therapy among students, and now ChatGPT seems to be filling that void. 

There’s plenty to be concerned about when it comes to students using ChatGPT as a substitute for therapy. For one, ChatGPT explicitly discourages sharing sensitive personal information on the platform due to privacy concerns. There’s also reason to worry whether ChatGPT’s advice will be helpful, rather than harmful, to students, given recent reports of strange and disconcerting conversations some users have encountered with chatbot platforms. There’s also a broader concern that personal interaction with AI could crowd out human-to-human interactions. 

Nonetheless, given the rampant nonconsumption of much-needed therapy, the prudent response to these concerns should not be to discourage or prevent students from using ChatGPT for mental health, but to create AI-based therapy models that are safe and effective. 

“Will Virtual Reality Lead More Families to Opt Out of Traditional Public Schools?”

A recent EdSurge article discussed a Florida-based virtual charter school, Optima Academy Online, that utilizes VR technology as the primary mode of instructional delivery. The school, which follows a classical education model, employs VR to enrich the learning experience. Does this forebode a future in which VR-based virtual schools will disrupt conventional schools?

Jobs to Be Done theory offers a framework for discovering why people break from the status quo to adopt new products or services. Using Jobs to Be Done, education leaders might ask “What are the struggling moments in the lives of students or their caregivers that lead them to choose VR schooling over conventional schooling?” Answering this question will be a key to the future success of VR-based schooling models. (For those interested in this topic, we have a report coming out within the next few months on the Jobs to Be Done of families who switch schools.)

I suspect there are pockets of students for whom VR-based schooling could be a great solution—such as those who are medically homebound, who struggle with social anxiety at school, who live in remote parts of the country, or who highly value the flexibility and customizability of home-based learning. For these students, VR-school is a valuable improvement over the virtual schooling options that have been around for some time. 

But unlike the therapy example, there is very little nonconsumption of schooling. For students who are satisfied with their brick-and-mortar school, VR experiences can’t beat hanging out in-person with friends. Nor can VR fulfill the custodial care role for working parents. If VR is the primary value-add for these schools, they are likely to remain niche solutions for a small subset of students and families. 

“How AI could save—or sink—creative writing in schools”

One of my recent blog posts argued for using tools like ChatGPT to help students develop their writing skills. In my own writing, I’ve found ChatGPT to be incredibly helpful for sketching my ideas into a first draft and for getting quick feedback. 

After sharing that post, I received an email from Lucas Jacob, the director of writing, communication, and media literacy at La Jolla Country Day School. He respectfully pushed back on my arguments and made a compelling case that introducing tools like ChatGPT to emerging writers would likely keep them from developing the essential skill of learning how to structure their writing. 

In truth, the use of AI for writing instruction is uncharted territory, and we don’t yet have proven methods for using it effectively for writing instruction while mitigating its potential risks. The topic deserves much more debate, experimentation, and research before any practices are promoted as universal solutions.

Clayton Christensen had a motto: “anomalies wanted.” It was a reminder that his theories were imperfect models of reality, but that they got better as he and others wrestled with facts that were not yet well accounted for by his theories. In the spirit of “anomalies wanted,” I invited Mr. Jacob to share his critiques in the comments section of my post. I encourage readers to check out his comments and join us in this important discussion.

Do you have an alternative take on any of these stories? Let me know what you think by leaving a comment below, emailing me at, or finding me on Twitter (@ArnettTom) or LinkedIn.


  • Thomas Arnett
    Thomas Arnett

    Thomas Arnett is a senior research fellow for the Clayton Christensen Institute. His work focuses on using the Theory of Disruptive Innovation to study innovative instructional models and their potential to scale student-centered learning in K–12 education. He also studies demand for innovative resources and practices across the K–12 education system using the Jobs to Be Done Theory.