leaders In header

Future of Education: AI in Schools

Across the world, teachers report both a lack of time, and a huge amount of stress.  This is resulting in working hours on the rise and burnout an ever increasing problem, especially as the administrative burden shows no sign of abating.

By: Adi Gaskell, Katerva’s Futurist 


The practice of education has remained largely unchanged over the past 150 years.  Teachers stand before a class of around 30 pupils and bestow their wisdom upon the class.  This heavily condensed version of Baumol’s cost disease goes at least some way towards explaining why education (and healthcare) have become progressively more expensive, even while the price of things such as computers, cars, food and clothing have plummeted.

Platforms such as those used by the Massive Open Online Courses (MOOCs) promised to disrupt this model by flipping the classroom and allowing world leading tutors to deliver lectures to tens of thousands of pupils at a time.  It remains to be seen if this is scalable, however, and despite the London School of Economics recently launching seven online undergraduate degrees in areas such as economics and data science, the disruption has been relatively slow.

“These degrees provide the means for talented students to access the academic excellence and social engagement that LSE is globally renowned for,” said Professor Paul Kelly, Dean of University of London Programmes at LSE. “LSE is a truly international institution and we consider University of London students to be important members of our wider academic community. This partnership with 2U will enable the University of London and LSE to ensure the continued evolution of the student experience to remain engaging and relevant for our times.”

Tech in the classroom

While such an approach has merit for universities who not only have world-class faculty but often an ever expanding global reach, it’s less beneficial for primary and secondary schooling, where the problems are often altogether more down to earth.

Across the world, teachers report both a lack of time, and a huge amount of stress.  This is resulting in working hours on the rise and burnout an ever increasing problem, especially as the administrative burden shows no sign of abating.  Indeed, the OECD report that the 50 hours per week put in by the average teacher is on the rise. It’s no surprise that 81% of teachers in the UK are considering leaving the profession.

While automation has enabled many industries to overcome Baumol’s cost disease, the reality for education is likely to be one where AI and associated technologies augment rather than replace teachers.  This impact is likely to be especially keenly felt in the administrative tasks that can take up to 40% of a teacher’s schedule, yet could efficiently be automated.

Humans in the classroom

Given that the very things that make great teachers so good are the very things that are often so hard to automate, this augmentation narrative should come as no surprise.  Nonetheless, it’s what materialized in a recent survey from McKinsey of a few thousand teachers from Canada, the UK, US and Singapore. The survey quizzed the teachers on the amount of time they spend on 37 different activities, from planning the lesson to grading assignments.  Alongside each activity, the teachers were asked to say whether they want to spend more or less time on them, as well as the technologies they’re currently using in the classroom.

The findings were pretty uniform, with teachers from all four countries reporting that they spend too much time preparing, evaluating and administering the classroom, and not enough on actual teaching.  

Preparation was arguably the biggest time sink, with teachers averaging 11 hours per week preparing activities for the classroom.  McKinsey argues that smarter use of technology could reduce this by almost half. Where they believe technology will have less impact is on the areas where teachers directly interact with the students.  They suggest that activities such as coaching and social/emotional skills development are not areas of strength for technology today.

Integrating at scale

As well as the challenges involved in providing useful technology in the classroom itself, there are also considerable barriers in terms of integrating new technologies at scale.  Curricula are difficult to change, and the time and money required to train teachers on new technology is often sorely lacking.

Where things may shift, however, is in the methods of teaching itself, with many educationalists advocating the flipped classroom, that sees lectures consumed at home, or in study sessions outside of core class time, and the teachers then providing personalized coaching and support to students based upon their interactions with the material.

This flipped approach also relies heavily on technology to provide evaluation and feedback so that when the teachers are called into action, they know precisely what their students know so they can provide the support required.  Computers have long performed this role in simple, multiple-choice environments, but neural network technology and natural-language processing is making it feasible for AI to assess detailed texts and provide feedback across a growing range of subject areas.

Last, but by no means least, technology promises to make a significant dent in the administrative tasks that take up so much of the modern teacher’s time.  Indeed, McKinsey believes that it could knock administration time down from 5 hours per week to just three. Collectively, the 13 hours per week that they believe technology could save, will not only significantly reduce the stress and burnout associated with teaching, but also make teachers more effective in guiding our young people into the future of work.

If you loved this article–our ‘Future of’ series will explore new topics released weekly every Thursday–stay tuned!