The Digital Delusion summary

A classroom platform can be polished, adaptive, and packed with features, yet still fail for a basic reason: human learning has biological requirements that software can’t redesign. Once you look beyond test scores and ask what allows knowledge to take hold, three recurring problems appear.

Start with attention. Conscious thought can follow only one active task at a time. What feels like multitasking is actually rapid switching, and every switch carries a cost: attention briefly drops, errors become more likely, and memory formation weakens. That matters enormously for students because screens have been practiced primarily as entertainment devices. Children aged eight to 18 are estimated to spend roughly 2,500 hours a year using digital media in fast-switching ways, compared with around 450 hours using devices for schoolwork. During screen-based study, the average student drifts toward unrelated content in under six minutes.

Phones have become such a normal part of daily life that simply having one nearby can feel harmless. In a classroom, though, their influence doesn’t begin only when the screen lights up. A student can leave a phone untouched and still have attention pulled toward it, while the expectation of checking it between lessons keeps that pull alive across the school day. That’s why simple “out of sight” rules often fail.

The academic cost is substantial. A meta-analysis found that greater smartphone use was linked to noticeably poorer academic performance. Large studies in several European countries also found that schools removing phones saw learning gains equivalent to roughly 11 extra hours of teaching each week. These policies can also return time to teachers by reducing repeated battles over device use.

Artificial Intelligence – or AI – creates a strange problem for education: the better the finished work looks, the less it may reveal about what the student actually knows. An elegant essay or flawless solution can now be produced with little of the struggle that builds understanding. When the tool does the thinking, polished performance can hide weak learning underneath.

That concern becomes sharper when you look at the early research. At first, the results seemed encouraging: two reviews appeared to show that AI improved learning. But once duplicate studies, missing details, and reporting errors were corrected, much of that apparent benefit faded. The evidence went from promising to weak, suggesting that the excitement around AI has moved much faster than the research supporting its educational value.

The real question isn’t whether technology belongs in schools, but when it actually improves learning. That means changing the default: start with what students need, then make every device justify the time, money, and attention it consumes. Technology should enter a classroom because it solves a defined learning problem better than the alternatives, not simply because it is available, fashionable, or convenient.

Parents can begin by making screen use visible. Track what devices are used at home, how long they are used, and whether they support worthwhile activity. At school, ask for independent evidence before accepting a new platform and request offline alternatives when appropriate. This matters because one-to-one device programs are now widespread, even though research suggests they produce only modest learning gains. Collective action can strengthen these requests: parents who share clear values and work with sympathetic teachers are harder to dismiss than isolated complaints.