If schools continue measuring knowledge in isolation while the workforce rewards AI-powered co-creation, today’s graduates risk obsolescence.
GUEST COLUMN | by Erin Mote
KORAKIT INTRAPRASERT
In the traditional world of educational technology, we are accustomed to an adoption process. A school district identifies a need, evaluates software, negotiates a contract, and rolls out a pilot program. This is a controlled, linear, and intentional process.
But as I recently argued at Stanford and in my Congressional testimony last year, Generative AI is not an adoption technology. It is an arrival technology. This concept, developed and articulated by MIT researcher Justin Reich, is critical for the work we need to do in education over the coming three to five years.
‘Generative AI is not an adoption technology. It is an arrival technology. This concept, developed and articulated by MIT researcher Justin Reich, is critical for the work we need to do in education over the coming three to five years.’
The distinction is not merely semantic; it concerns the difference between adopting a new tool and adapting to a new environment. You don’t ‘use’ arrival technology any more than you ‘use’ the air you exist within it. To understand why AI is causing such profound fear among the public—despite experts’ optimism—we must examine the history of arrival technologies and the four dimensions that define them.
Data from the Pew Research Center highlight a widening perception gap: while 56% of AI experts believe AI’s impact over the next 20 years will be positive, only 17% of the U.S. public shares that optimism. In fact, 35% of U.S. adults anticipate a negative impact, citing concerns that AI will weaken core human skills and social connections. This disconnect is even more pronounced in specific sectors; for instance, while 73% of experts see AI’s impact on jobs as positive, only 23% of the public agrees.
Arrival technologies are defined by their ability to bypass traditional gatekeepers. They are a class of innovations—in this case, Generative AI—that do not simply “pass through” the economy as tools but instead “arrive” as a permanent, foundational layer of human infrastructure.
– Electricity did not wait for a procurement officer to decide if lightbulbs were “pedagogically sound”; it arrived and immediately reordered the hours of the day, the design of buildings, and the very nature of work.
– The Internet dismantled the monopoly that libraries and textbooks had over information.
– Social Media arrived and rewired the social fabric of an entire generation before a single school board could draft a “responsible use” policy.
Like these predecessors, Generative AI didn’t ask for permission. It wasn’t adopted; it happened. Data from the 2024-25 school year confirms that it is already a permanent feature of our environment: roughly 85% of teachers and students reported using AI, even though it was never “officially” rolled out in most of those contexts.
It is a fixture of the classroom that does not require approval from the school board. This reality demands that we design for foreseeable disruption rather than merely incremental improvement, as the technology is already deeply embedded in the daily workflows of the majority of our learners and educators.
The Four Dimensions of an Arrival Technology
To navigate this shift, we must understand the four characteristics that make an arrival technology uniquely challenging—and uniquely powerful.
1. Maximal Systemic Disruption
In an adoption model, we fit a tool into our existing systems (e.g., using a digital whiteboard to do what we did on a chalkboard). Arrival technologies, however, break the systems themselves. If a student can generate a B+ essay in thirty seconds, the “system” of the take-home essay is no longer a valid measure of learning. This disruption feels so tangible because our educational systems have historically been built to measure what you know, rather than what you can create with what you know. We aren’t just changing the tool; we are being forced to interrogate the fundamental architecture of assessment and instruction.
‘We aren’t just changing the tool; we are being forced to interrogate the fundamental architecture of assessment and instruction.’
2. Reordering of Everyday Life
Arrival technologies don’t stay in the “enterprise” (the school or office). They leak into the home, the grocery store, and the playground. Because AI assists with everything from drafting emails to coding apps, it reorders how students spend their time outside of school. Education can no longer exist in a vacuum; if the world outside is AI-augmented, a “policy-gap” inside the school creates a dangerous disconnect between academic life and reality.
3. Access Defined by Economic Wealth
While adoption technologies are often provided via school-funded 1:1 initiatives to ensure equity, arrival technologies often follow the path of private wealth first. Those who can afford the “pro” versions of models, or who have the hardware to run them locally, gain a compounded advantage. In education, this means “arrival” can exacerbate the digital divide unless we proactively build public infrastructure to ensure equitable access.
4. The Potential for Managed Harm
The data reveal a stark “Policy Gap”: while 85% of the community uses the technology, only 31% of public schools have written policies governing its use. Arrival technologies can cause harm if not managed appropriately. From deepfakes to the erosion of core human skills, the risks are real. However, the EDSAFE framework argues that safety is not the counterpolarity to innovation. Rather, safety is the guardrail that allows for growth. By establishing benchmarks for Safety, Accountability, Fairness, and Efficacy (S.A.F.E.), we move from a reactive posture of “banning” to a proactive posture of “steering.”
The New Baseline for Survival
And the clarion call for action is now. The most urgent reason to understand AI as an arrival technology is the risk of obsolescence. If we fail to design systems that account for the rapidly changing composition of technology and life, we lose our ability to answer some of the most fundamental questions about the purpose of industrial-era education: employability and workforce readiness.
If schools continue to measure knowledge in a vacuum while the workforce measures the ability to co-create with intelligence, our graduates will enter the labor market fundamentally unprepared. We cannot afford to wait for the “perfect” adoption cycle; the composition of work has already changed. To ensure students are ready for the world that has arrived, we must design for a future where technical fluency and ethical co-creation are the new baseline for survival in the global economy.
‘To ensure students are ready for the world that has arrived, we must design for a future where technical fluency and ethical co-creation are the new baseline for survival in the global economy.’
If we treat AI as an adoption technology, we will spend the next decade trying to “integrate” it into 20th-century classroom models. If we recognize it as an arrival technology, we acknowledge that the landscape has changed forever.
Innovation moves at the speed of trust. To build that trust, we must move past simple adoption and toward a systemic redesign—one that prioritizes AI literacy, robust public infrastructure, and the protection of fundamental human rights. We are not designing for a new tool; we are designing for a world in which schools must not only transfer knowledge but also empower imagination and connection.
—
Erin Mote is CEO of InnovateEDU and a national voice on AI policy, digital equity, and student data protection. She has testified before Congress on the future of education and technology. Connect with Erin on LinkedIn.
0 Comments