For teachers on the ground, the promise of a digital classroom is at odds with their daily reality. (Photo: Gallo Images / Papi Morake) There was much fanfare in certain parts of the internet when Nature retracted the study, The effect of ChatGPT on students’ learning performance, learning perception, and higher-order thinking: insights from a meta-analysis, due to concerns around discrepancies in the meta-analysis and a lack of confidence in the conclusions.
The problem is that the flawed study has been cited some 262 times in peer-reviewed papers published in the same journal, in less than a year. It seems that the science world was desperate for hope about the role large-language models could play in education.
Now seems like a good time to mention another dramatic turnaround in a decision to deploy technology into the classrooms alongside the earth’s most intellectually vulnerable inhabitants (read: children).
Three years ago this month (May 2023), the Swedish minister of education announced that the government would progressively decrease the digitalisation of Swedish schools and reintroduce paper-based reading. This came after the country went all in on digital education in 2009. The reversal has a €104-million budget attached to the ambition of “one textbook per pupil and subject”.
Back to basics
/file/attachments/orphans/GettyImages-71805388_920179.jpg "Introducing AI and tech without deep pedagogical roots risks deskilling the next generation, warns Cambridge University Press group managing director for international education, Rod Smith. (Photo: Tim Shaffer / Microsoft via Getty Images)")
When Daily Maverick asked Cambridge University Press group managing director for international education, Rod Smith, about the global pivot to introduce technology and AI into the curriculum, he was cautious and measured.
“Quite often these debates can be led by the technology first rather than by the pedagogy first or by the education first,” said Smith in expertly chosen words.
Smith offered a warning that introducing AI and tech without deep pedagogical roots risked deskilling the next generation, saying that a curriculum instead must empower learners to control technology.
“If we are educating young people to be replaceable by [AI] we’re teaching them the wrong things... we’re not developing the skills to enable them to be master of the technology rather than being replaced by the technology.”
Cambridge, like most education businesses looking for scale in the modern economy, is focusing on what it calls an “evidence-based” approach to AI rather than rushing to adopt new tech for its own sake. They are still building out AI capabilities, don’t get that part wrong.
Their primary focus is on how technology affects learning and brain development, stating that their thought leadership is “very much based around what is the evidence telling us about how this is working?” and “what is it the evidence telling us about the impact on young people and how they’re thinking and their cognitive capabilities”.
As easy as do, re, mi
Smith saw the trap coming in our questions and quickly clarified that Cambridge is not trying to pivot into an EdTech vendor that sells hardware or proprietary software packages:
“We are not in the hardware sales process...”
Cambridge operates globally and, under his stewardship, has a modulated approach to development. “We’ll go at the pace … that people want us to go at... when you’re in international, you realise that some countries are ready to go digital and some will continue to need to have, you know, print”.
Smith actually defended the pedagogical value of physical books over screens, explaining that “actually there are some things that print does very well, better than digital; in fact, in terms of the way information is presented and assimilated by young people”.
But his words around advancements in technology – that are actually benefiting physical books – resonate with a choir of teachers who have grown weary of technology’s relentless march.
“Even there the new technologies are making it cheaper to print, cheaper to produce things like that, so I think there will be a benefit there to that format as well.”
For teachers on the ground, the promise of a digital classroom is at odds with their daily reality. The National Professional Teachers’ Organisation of South Africa (Naptosa) has repeatedly voiced frustration over top-down technological and curriculum shifts formulated without adequate consultation or an understanding of classroom conditions, so the idea of new textbooks should read like a Scandinavian fairytale.
Organisation president Basil Manuel has previously questioned: “How far removed from the classroom are those so-called experts and education policy makers?”
They rightly argue that adding complex technological and assessment requirements burdens an already overloaded system:
“Instead of delivering quality education, teachers have been reduced to crowd controllers in chaotic classrooms… Listen to teachers – those who live the reality of the classroom daily and understand that genuine transformation begins with creating the right conditions for teaching and learning.”
Enter to learn
The current target of teacher frustration is Minister of Basic Education Siviwe Gwarube, and she is thankfully, largely on their side:
“Foundational learning is central to the work we have undertaken in the past year, as a child who cannot read, write and calculate with meaning and confidence will struggle in every other subject... we launched a strategic shift to improve foundational learning. This strategy, endorsed by Cabinet, places focus on expanding access to quality ECD and early-grade literacy and numeracy.”
The DBE’s broader strategy continues to view technology as a vital pillar of the classroom, but calls on all stakeholders to maintain the “Five Ts” (Time-on-Task, Teacher preparedness, Textbooks, Technology and Testing). Summarising the Ramaphosa administration’s vision for this dual approach, Gwarube explains:
“The achievements of 2024/25 reflect a decisive reorientation of the basic education system: evidence-led planning, strengthened foundations, inclusive access, sustainable infrastructure and professional teaching. Together, they advance the GNU’s vision of an equal, just and prosperous society by investing in the most important national resource – our children.”
When screens do work
While the Swedish government pulls back from screens and our government is erring on the side of balance, there are two South African school networks that argue that technology, when integrated properly into a blended learning model, is the key to delivering high-quality, individualised education at scale.
At Apex and Spark Schools, technology is not a distraction, but the primary method for content delivery, which frees up teachers to focus strictly on their actual job: pedagogical intervention.
Sanri Festen, head of blended learning at Apex Education (with schools in the Western Cape areas of Stellenbosch and Eerste River), describes this dynamic:
“The blended learning model allows technology to deliver the lesson for the first time [to] introduce the content and then let the teacher and the facilitators be the ones to close the learning gaps... bringing technology into the class has been a major shift because now we can collect data in the moment – we have real time data points on our students and we can respond and support them in the moment.”
This writer has personal experience with the pioneering (in South Africa, at least) Spark Schools version of blended learning – both his children were subject to it in the early stage of primary school (getting into schools in Cape Town is tough) – with great results. But CEO and co-founder Stacey Brewer explains it best:
“We were already getting scholars to split time between digital content (that adapts in difficulty to their learning pace) and classroom interaction based on best practice tuition.”
Cost benefit
For these networks, technology is a crucial mechanism for making premium education affordable. Both schools justify their use of technology by explaining that their blended learning models combine the best elements of technology with in-person access to excellent teachers.
This does allow them to provide quality education at an affordable price (between R800 a month and R2,000 a month – Apex targets a lower end of the market than Spark) to communities that need it most.
South African parents know that we are starved for affordable school models for high-quality education, and Apex compares its response to that need to designing an electric car from first principles rather than merely fixing broken petrol cars. They also argue for bravely embracing the enormous benefits of technology to achieve this goal.
On the hardware side, Spark uses Acer Chromebooks for its learning labs, and Apex sources its Chromebooks from Asus. In both cases, Google is the real winner because the young minds are subjected to their idea of computing at the foundation phase.
But the hardware battle is a topic for another day.
For the purposes of this article, at the start of the 2026 academic year, the Gauteng Department of Education reported 1,381 Grade 1 learners remained unplaced, part of a total of 4,858 unplaced learners (including Grade 8).
While Sweden’s massive investment in reverting to paper and textbooks is grounded in the legitimate dangers of screen distraction, cognitive offloading and diminished reading comprehension, the success of Apex and Spark Schools reveals a different, local truth.
South Africa is the most advanced economy on the continent, but that is not a prophylactic for the emerging-market tax: making compromises. DM
Tech in Education Part 2 will uncover the technology vendors who exploit this structural weakness.