You may remember when, a few months ago, Pepper the Robot gave testimony to the Education Select Committee. That day, Times education editor Sian Griffiths asked Twitter how long it would be before teacher shortages were filled by robots. A tongue-in-cheek comment, surely? If you believe that, then you won’t be comforted by what Professor John Hattie had to say in Edinburgh this week.

The author of Visible Learning, a book with Teflon-level resistance to valid criticism, which has granted Hattie a sort of canonisation in his own lifetime among the education community, was speaking to the Visible Learning World Conference. Yes, as well as radicalising individuals and semi-autonomous cells of true believers, the book has spawned an eponymous world conference. (No, it’s nothing like a religion. Honest.)

At this conference, Hattie reported upon a visit to Asia where he’d seen a lesson taught by a robot. Afterwards, the students allegedly said in conversation with him that they preferred the robot teacher to its human antecedent. Hattie used the speech to undermine the idea that student-teacher relationships, and student-student relationships, are crucial to education. In a clarion call to policymakers to take AI seriously, he stated that it has the power to “reduce some of the problems that we have that are related to human interaction and all the biases that relate to it”.

It is hard to provide a counterfactual to Hattie’s claims. Asia is a big place with more education systems than countries, and it’s impossible to know what questions he asked to elicit this response from the students. But be in no doubt, policymakers everywhere are listening. Attendees to his research-informed education mega-church have flown home to destinations worldwide, and policy researchers everywhere are already drawing up reports on AI in education for their respective ministries.

Who among them will dare run counter to the tide? Who will ask if the “problems we have that are related to human interactions and all its biases” aren’t in fact modelled and disseminated by policymakers themselves? Who will wonder whether the students don’t prefer the robot to an over-worked and micro-managed human, but might prefer a relaxed and self-efficacious professional to a talking algorithm?

Nobody. That’s who.

Faith in fakes

According to Hattie, the students particularly liked that they could ask the same question multiple times without being judged. In my 14 years of classroom practice, that kind of judgement was far more likely to come from peers than from me, and I would always challenge it. When it did come from me, it was because of pressures put on me from outside the classroom, and that’s a large part of why I stepped away from the job. I could neither fully be the human I wanted to be, nor the machine my leadership team needed me to be.

When Jim Callaghan, in 1976, first made school improvement a national issue and a Labour priority, and brought politicians and the public into the so-called "secret garden" of education, his specific concern was the curriculum. Here we are in 2019, and Ofsted and the profession seem only just to have discovered the idea.

In 1976, Callaghan worried that industry complained young people left school without the “basic tools to do the job”. Today, that complaint is only louder.

In fact, the only concern of his that has been effectively addressed is the gender gap in science attainment. Unfortunately, this seems to have been finally achieved due to a 10 per cent drop in boys’ performance in the 2015 Pisa test, rather than a rise in girls’ performance. Additionally, high achieving pupils in science are more likely to come from advantaged socioeconomic backgrounds, be white, and attend selective or independent schools. Not what one might imagine a Labour success to look like.

In short, national policy making has been nothing but an abject failure in state education by almost anyone’s honest calculations. It has only been successful in making teachers do a whole lot more of the things politicians need them to do to make them look good – namely provide reams of data – and a whole lot less of what Callaghan already knew worked best, but which everyone in education seems to have forgotten.

Like the villagers in Beauty and the Beast, who remember nothing of the castle in the mountains just above them because of the witch’s curse, political incantations have erased from our collective consciousness the importance of parents. In their place stands the taxpayer who pays our wages, ephemeral, changeable, and driven only by short-term efficiency.

From the parent point of view, political manifestos and Ofsted judgments have increasingly replaced direct engagement, except in the most superficial ways – a situation the UN recently bemoaned.

A Technological Utopia?

Rather than accept Hattie’s claims at face value, the political left has an opportunity to reclaim the educational narrative from the prophets of technological utopianism. Artificial intelligence is not just about making machines that do what humans do, but better. It sets about making humans more like machines too. It’s a self-fulfilling prophecy. In the face of this deprofessionalisation, human teachers leave in their droves. They leave the poorest schools first, and they leave a gap for learning machines to fill.

Once the 40-year political project of national accountability delivers an education system founded entirely on the teacher-student relationship, then what? Once efficient, centralised scripted lessons are in place everywhere. Once effective use of interleaving and spaced practice streamlines our impoverished idea of curriculum. Once direct instruction is the only instruction. Once teacher education and recruitment are completely deregulated and a stable, sustainable level of staff turnover is achieved. Once segregation of school provision and the judgement of school quality are made algorithmic. Once the taxpayer, finally, can be sure not a penny is spent in error. Then what?

Well, then teachers will be the only source left of problematic human bias. Then, racial and class divides won’t be evident in exam outcomes alone, but in the nature of educational provision itself. Then, the machines will be doing all the technical jobs anyway, and industry’s demands on the school system today will turn out to have been as irrelevant to the future as those that troubled Callaghan are to us.

If I were a gambler, I’d put money on robot teachers replacing the workforce in comprehensive schools and further education colleges, but not their independent or selective counterparts. There, humanity will still be valued, as all things seem to be, by having a price tag attached to it.

And if I were a politician, I’d say that wasn’t a vision of the future most parents would be happy with.

JL Dutaut is co-editor of Flip the System UK: a teachers' manifesto (Routledge). He is currently on a career break from teaching to research school accountability systems around the world. He hasn't found one he likes yet, and he doesn't think you would either