What a soulless dystopic world we have ahead of us if the proponents of Artificial Intelligence are to be believed.  If we share their understanding of education processes, we believe that all knowledge can be systemised, programmed and parcelled out to the empty vessels - students - in the sterile virtual classrooms of the future. 

There are many things wrong with the assumption that a good education merely consists of lectures streamed from all over the world and lessons in which students sit in front of a machine that will set them work, then mark and level it before churning out the next task for completion. Something seems to be missing here - unless the students of the future are rather more biddable and less questioning than the current ones.

Diagnosis of academic achievement and levels of understanding seems so blindingly straightforward in the AI world.

Organic learning

Those of us who have been in the education business since the introduction of the National Curriculum - and before - will remember that, to date, the levels of achievement in all subjects have been revised several times, and that even APP was very limited in its scope.  In fact, levels should have been dispensed with completely since the sterling work of the Commission on Assessment Without Levels report.  Shouldn’t we be taking note that learning is not all about progressing from one perceived “level” to another, but something rather more organic and less easily definable?

Even Ofqual, in consultations prior to the introduction of 9-1, admitted it could not provide level descriptors for every single level.  Teachers may have noticed that every time specifications are revised, there are only official descriptors for the 9-1 equivalents of A, C and F.

I would be interested to see how anyone could list the vast array of skills and elements of knowledge required in all its permutations for GCSE or A-Level English literature and provide fail-safe judgments for every eventuality. We humans are still struggling to agree.  The latest Ofqual research using the senior examiners and subject experts to review “problem scripts”, found that there were some scripts on which agreement of the mark could not be reached. If the subject experts and the most senior examiners are struggling, can we really expect that artificial intelligence will be able to find a pathway through such complex nuanced marking?

Emotional and intellectual

In the real world, education is a messy business with all kinds of false starts, misunderstandings, and revisions.  Even the most able of students don’t grasp everything the first time. Real learning graphs go up as well as down and students revise their methodology to tackle increased levels of difficulty. Teaching is as much a negotiation as it is delivery of subject-specific material. There are many times when teachers re-phrase, re-order, and re-write instructions and explanations in order to advance students’ knowledge and understanding.

Inconveniently, students have individual preferences as to how their education is provided, hence the many pupil voice surveys.  Education is as emotional as it is intellectual. The cleverest and the most resistant students often want to know “What’s the point?” as they encounter new material, questions and problems. We all know how much of a lesson can be given over to giving a rationale for embarking on a topic or a question. A touch of humour helps the knowledge go down and a collaborative exercise raises more interesting questions and takes students through vital emotional learning about the stages of a group, for example.

Assessing artificial intelligence

Although the proponents of educational research approach teaching as a scientific exercise, all of us know in our heart of hearts that no method or approach is completely classroom proof. Teaching remains as much of an art as it does a science. Maybe the citizens of the future will need the skills to work alongside the artificial intelligence built into many job roles, but they would be wise to acquire the interpersonal ones first if they are to rise in any profession.

And that means that literature, the arts, and humanities will still be vital. I don’t think I will need to retire to the bunker anytime soon, but should that day come, like my hero, the Savage in Brave New World, I will be shouting for my Shakespeare.  Let’s see how even the most advanced AI copes with assessing an A-level coursework essay on that. 

Yvonne Williams is Head of English in a secondary school in the south of England and has co-authored an article How accurate can A-level English literature marking be? Published in the current National Association for the Teaching of English’s English in Education journal.  

Want to keep up with the latest education news and opinion? Follow Tes on Twitter and Instagram, and like Tes on Facebook