Mr Edison says that the radio will supplant the teacher…

Teachers will be relegated to the backwoods.

With fire-horses.

And long-haired women.

Or perhaps shown in museums.

Education will become a matter of pressing the button.

Perhaps I can get a position at the switchboard.

This was penned by a disgruntled US teacher in the aftermath of the First World War. It records one of the technological waves that would, enthusiasts predicted, transform education. The forecast, like its successors, proved false.

The poem transports you to a semi-steampunk world, with rows of smartly dressed women manipulating machines with buttons and copper wires. Of course, this vision now seems ludicrous (if a good setting for a young-adult novel, complete with long-haired women and abandoned fire-horses).

The technology the anonymous writer cites is antiquated. Just as the vision of the Skinner box - a futuristic 1950s idea of a “teaching computer” - seems ludicrous. It is harder to take seriously the idea of a classroom in which every child peers into a behaviourist box than a Victorian classroom with blackboard and chalk.

Content is controversial, it is expensive to make, it is inherently subjective. In other words, it is often a bad business model. But it really matters

Yet the reasons for psychologist BF Skinner’s invention were identical to those given today. Children are disengaged. Classrooms are inefficient with students all moving through their lessons at the same pace. Feedback is too rare and dependent on the overworked teacher’s marking. The Skinner machine was absurdly clunky by today’s standards - relying on levers and punch cards - but its goal was not.

Many have claimed that technology will now achieve its potential because it is finally the right technology. Specifically, that the combination of portable devices (such as tablets), the internet, adaptive engines and machine learning will make possible what radio, television, interactive whiteboards and educational games could not. Even the disappointment of massive open online courses haven’t dented this enthusiasm.

I am sceptical of this claim. New technology is unquestionably extraordinary (though so was the radio). Artificial intelligence and machine learning (ML), in particular, seem to be on the brink of transforming our lives.

Some education applications that use ML - such as language teacher Lingvist - are exciting. But it seems to me that technology enthusiasts have consistently ignored two things. The first is content - something that every ed-tech investor groan.

Content is not yet king

Content is controversial, it is expensive to make, it is inherently subjective. In other words, it is often a bad business model. But it really matters.

Quality is important. Just as the difference between a great novel and a bad one is important, or a good and bad song on the radio. Or the difference between Plato and your average PhD philosophy student. But technology is quality agnostic. It can use unreliable measurements (which would render ML useless). It can deploy incoherent content. It can view engagement as the end goal, rather than learning. This can negate its potential.

The second is time. Time is one of the great promises of technology: more efficiency, less workload. But I have been to very few classrooms with student-facing technology that have really saved time. Boring things such as collecting devices, switching them on, making sure connections aren’t faulty and work isn’t lost, removing distractions - all eat away at efficiency gains. All are solvable, and have been solved in many schools, but are remarkably hard to solve consistently at scale. You cannot wish away hundreds of thousands of physical classrooms, each with 30 children.

It is these issues that have led some proponents of traditional education (although I dislike this term, since this approach is based on more up-to-date science than alternative views), including knowledge-based curricula, tough behaviour policies and frequent assessment, to abandon technology. They think, quite reasonably, that the content is less good than they could produce themselves and that it wastes more time than it saves (though they are more sympathetic to technology that is not student-facing).

This has been compounded by a land-grab of technology by the progressives (again, a term I dislike, since they don’t seem to have changed their minds for decades and this is not, usually, how we define progress).

This is, on the face of it, odd. Our teacher poet was defending workers’ rights from the invasion of technology. The Skinner machines were criticised for their behaviourist model (with some justification). The anti-technologists were traditionally on the left.

Myths about brains

Perhaps it’s a linguistic phenomenon - technology is progress, so it belongs to the progressives. Or a reflection of the left’s abandonment of workers’ rights in favour of other causes. Or simply an assumption that those who believe in knowledge love the 1950s and hate anything that has happened since. I suspect, though, that it is more because of the two most recent radical technological transformations. The first, the computer, has led us to completely mischaracterise the brain. The second, the internet, appears open-ended and is therefore compatible with a world in which students can be let loose to discover whatever they want.

The progressive model of education doesn’t work. We know, from ever-improved experiments in cognitive psychology, that the brain is nothing like a computer and requires coherently taught knowledge

I think this land-grab is a great shame. The progressive model of education doesn’t work. We know, from ever-improved experiments in cognitive psychology, that the brain is nothing like a computer and requires coherently taught knowledge. It cannot just explore as an empty vessel or be taught the processes of cognition without facts. That’s not how we evolved.

At the same time, technology does offer opportunities. Not as a pure medium, but through what it can convey and how. It doesn’t matter if you have a radio if no one ever composes great music.

In the next few years, I hope to see superb online content, backed by the best assessments, that is thoughtful about how to save teacher time and constantly improve learning. In other words, I want the traditionalists to claim back technology - by thinking of “what” first and “how” second.

Rachel Wolf is a former Downing Street adviser and director of consultants Public First