The $30 Billion Rethink: Why the World Needs a Different Model of EdTech — and What NDX Is Building

A recent LinkedIn post by Cristóbal Cobo, Senior Education and Technology Policy Specialist at the World Bank, has been circulating widely across the education community. Drawing on new reporting and a Stanford-led study, Cobo summarises an uncomfortable truth: U.S. schools spent roughly $30 billion on educational technology in 2024 — about ten times what they spend on textbooks — and yet measures of student numeracy, literacy and cognitive performance are moving in the wrong direction.

We think Cobo is right. And we think the conclusion is not that technology in education has failed. It is that one particular model of edtech has failed — and a different model is now both possible and urgently needed.

The diagnosis: spending without learning

The numbers Cobo highlights are striking. U.S. K-12 districts are projected to nearly double their edtech spend by 2033. Almost 90% of schools now issue a device, with some programmes starting in kindergarten. Nearly two-thirds of elementary-age children spend up to four hours a day in front of a laptop. And the federal government calculates that 92% of districts used pandemic relief money to buy edtech — money that is now gone.

Against that backdrop, the cognitive picture is sobering. Gen Z is, by several measures, the first generation to test lower than its parents on IQ, numeracy and creativity. Research increasingly links heavy unstructured screen time and multitasking to weaker executive function and academic performance. A Stanford-led study confirms what most teachers already suspected: the impact of edtech varies enormously by product, by feature and by the specific skill being taught.

In other words: the money flowed, the devices arrived, the apps multiplied — and the learning did not follow.

Why this happened

The standard edtech model of the last decade was built on four assumptions: more apps are better than fewer, individual screen time is the unit of learning, generic global content can be poured into any classroom, and connectivity is something districts can simply assume. Each of those assumptions is now visibly breaking down.

The average U.S. school district touched more than 2,500 distinct edtech tools in a single year. Teachers were asked to integrate, manage and assess across a fragmented landscape that no curriculum was ever designed for. Lessons became a sequence of logins. Pedagogy became an afterthought. And as Cobo notes, the student quietly turned into a consumer — of content, of notifications, of someone else's product roadmap.

Where we agree — and where we go further

At NDX we agree with the core of Cobo's argument. Edtech-as-a-shopping-list does not work. Devices alone do not teach. Screen time without structure is not learning. And public education systems cannot be sustained on emergency funding cycles that buy hardware once and leave teachers, content and maintenance to chance.

Where we go further is in the response. The lesson from the $30 billion decade is not that countries should pull back from technology in education. It is that the next generation of education technology has to look fundamentally different: integrated rather than fragmented, sovereign rather than borrowed, curriculum-aligned rather than generic, and built around the teacher rather than around the app store.

A different model: what NDX is actually building

NDX builds Educational Intelligence for K-12 classrooms in emerging markets. That is a deliberate choice of words. We are not selling an app. We are not adding a 2,592nd tool to a district's catalogue. We are building the integrated platform that ministries of education can actually run — combining classroom hardware, offline-first content, curriculum-aligned AI and structured teacher training into a single coherent system.

Five design principles separate this model from the one Cobo critiques:

• One platform, not a thousand apps — devices, content, lesson tools and AI live in one curriculum-aligned environment, so teachers stop juggling logins and ministries stop paying for overlap.
• Teacher-first, not student-as-consumer — our tools cut lesson prep, marking and reporting time so teachers get hours back each week and stay in charge of the lesson.
• Curriculum and language native — content is aligned to the national curriculum, in the language of instruction, with examples that match the classroom in front of the teacher.
• Offline-first and sovereign — lessons run whether or not the internet is stable, and national data, language and curriculum stay under national control.
• Built for the budgets that actually exist — designed for the devices, timetables, power and connectivity emerging-market schools really have, not for an idealised U.S. district.

Why this matters now

Cobo's warning is aimed largely at high-income systems that have already spent the $30 billion. For emerging markets, the strategic moment is different — and better. Africa, MENA, South Asia and Asia-Pacific have not yet made the mistake of buying 2,500 disconnected tools per district. They have the chance to skip that decade entirely and adopt an integrated, sovereign, teacher-centred model from the start.

That is the opportunity NDX is built around. Across our deployments — from Cameroon's national Intelligent Classroom pilot with MINESEC, to engagements with ministries across West Africa and the Gulf — we are seeing the same pattern repeat. When devices, curriculum-aligned content, offline AI and teacher training arrive together, teaching gets easier and learning gets visibly stronger. When any one of those elements is missing, schools end up exactly where Cobo describes: spending more, achieving less.

The takeaway

The $30 billion number should be a wake-up call, not a verdict. It tells us that the first generation of edtech was structured around the wrong unit of value — apps and devices instead of teachers, lessons and outcomes. It also tells us that there is now both the evidence base and the political appetite to demand something better.

That is the category NDX is building: Educational Intelligence — a single, sovereign, curriculum-aligned platform that ministries can lead, teachers can actually run, and students can learn from, every day, with or without the internet. We are grateful to Cristóbal Cobo and the World Bank team for naming the problem so clearly. Naming it is the first step. Building the alternative is the work.