Charting the battle of man vs. machine in Education
How the tech-empowered teacher will lead the future of Education
“Any teacher that can be replaced by technology should be!”
Arthur C Clarke knew how to push buttons with his words. The battle of man vs. machine has long been fraught with tension (not helped by Hollywood’s apocalyptic depictions). IBM’s Deep Blue served up the landmark moment for Artificial Intelligence in 1997 by defeating reigning chess champion, Garry Kasparov. Since then, the relative progress of human and computer chess players is a grim tale for those hoping that humanity will strike back.
Here’s one chart that maps the evolution of each group over time, according to the Elo rating in Chess.
While humans are getting better at chess, machines have marched on at breakneck speed, leaving their organic counterparts flailing behind.
With the rise of intelligent tutors and other so-called teaching machines, AI now appears to have Education in sight. It seems that each time AI reaches a new milestone, the education community looks on nervously, pondering the consequences for teachers. So, what would such a progress chart look like for teaching?
Any analysis must start with the objectives of teaching. The Elo ratings are accepted as a good measure of Chess ability. There is no straightforward equivalent for teaching. So let’s instead evaluate teaching by a holistic measure that we’ll call “Quality”, keeping the inverted commas to remind us that this is not an exact science. What follows is a conceptual model of how teachers fare against computers in the task of educating young minds.
“Quality” as cognitive
Any sensible definition of “Quality” must entail the teaching of basic cognitive skills. This has been a struggle for schools as long as they have existed. Children are extraordinary for their uniqueness — they develop these skills at their own individual pace. The traditional one-size-fits-all model that teaches to the “average student” is flawed precisely because no such student exists.
It’s not all bad: with research, experience and professional development, teachers are evolving all the time. But can teaching machines do better?
BF Skinner, an early pioneer of adaptive tutoring, certainly thought so. The behavioural psychologist’s electromechanical teaching machines of the 1960s were built to automate instruction. Skinner declared his goal as “allowing teachers to teach more students than ever before” — efficiency was his main driver.
Within this framing, digital teaching machines pose an existential threat to teachers. They embed these basic skills with brutal efficiency, providing targeted support that accounts for students’ prior learning and adapts with every interaction they have. It’s only a matter of time before teaching machines surpass human teachers, if indeed they haven’t already.
As technology grows exponentially more powerful, and AI adopts more nuanced forms, teaching machines will also penetrate deeper layers of cognition such as problem solving and reasoning. And since these non-routine cognitive skills are heavily dependent on the fundamental building blocks of learning that escape many classrooms, teachers will lag even further behind.
Perhaps then, the “Quality” curve will resemble the one seen in Chess:
This thinking is common in EdTech. Jose Ferreira, CEO of the analytics platform Knewton, speaks of his product as a “mind-reading robo tutor”. With machines like these, who needs human teachers, right?
Wrong.
The human side of “Quality”
This picture only tells a partial story. It is based on a mechanistic view of Education that measures teaching only by cognitive outcomes. But what about the human dimension of learning? Surely that should be factored into “Quality”. Perhaps that will be the trump card for human teachers.
This “human dimension” is difficult to define, but it has clear markers: the quality of their face-to-face time with teachers, how they interact with their peers, how they connect their learning to the world around them…the list goes on. And neither human teachers or teaching machines have nailed it.
One teacher, thirty students, a set of curriculum standards and limited time: that is not a recipe for humanised learning.
Many teachers — even the best willed — will fail to establish a human connection with their students because they can not overcome the rigid constraints of formal schooling. Grading, administration and test prep are all privileged ahead of students’ personal needs and preferences. The prevailing didactic model of direct instruction, in which the teacher lectures to students, is deeply impersonal and strips the classroom of its humanity by reducing children to passive consumers of knowledge.
Teaching machines are designed to behave like humans. Adaptivity itself is a human-centred design principle; by having lessons delivered according to their needs and preferences, children are made to feel like someone — or something — gives a damn. Couple this with the affordances of the digital medium — rich, engaging, interactive content — and it is hard to argue that computerised learning is not also deeply human. The student-computer connection will only deepen as AI propels teaching machines towards more humanised behaviours.
But, to state the obvious, a teaching machine is still just a machine.
If we have to draw the line somewhere, it’s in recognising that a child’s relationship with a machine will never be as deep or meaningful — as human — as with an actual human being who gives them the attention and nurturing they deserve.
Our chart needs an upgrade.
Individually, human teachers and teaching machines are inherently limited — in math speak, their “Quality” is bounded. Teaching machines may reach a higher limit, but their impact as self-contained tools will remain stifled.
Arthur C Clarke’s maxim was on the money, but it is due an extension:
“Any teacher that can be replaced by technology should be! And any technology that ignores the human dimension of learning is redundant.”
This doesn’t roll off the tongue as easily, but it neatly places the luddites and overzealous technologists in their place by reminding us that the challenges of Education can not be met by humans or technology alone.
The tech-empowered teacher
But what happens when human and computers come together? The recent history of Chess actually delivers an optimistic view.
As documented in The Second Machine Age, shortly after Kasparov’s defeat the chess community organised freestyle tournaments, with teams composed of both human chess players and computers. By now it was clear that through sheer computational power, even a standard chess machine would obliterate the best human players. What surprised everyone was that the best teams were not those with the most powerful computers. Nor did they have world-beating grandmasters. In fact, the first winning team comprised a pair of amateur chess players who used three basic computers. That’s right: amateur players and basic computers. How so? According to Kasparov himself, it was their understanding of chess as a process that set the pair apart:
“Weak human + machine + better process was superior to a strong computer alone and, more remarkably, superior to a strong human + machine + inferior process.”
The amateurs triumphed because they expertly fused human judgement with technological prowess to achieve the best combined performance. There’s a powerful lesson here for educators and innovators of what can be achieved when humans stop racing against machines and start racing with them. This idea underpins the the enterprise of Blended Learning, a pedagogical framework that aims to enrich the teacher-student relationship by wrapping technology around educational goals.
There is nothing more human that deepening the student-teacher relationship, and it’s here that technology brings its richest promises. The teaching machine is no longer just an automated instructor, but a teaching assistant. The teacher is empowered by having a teaching assistant for every child, feeding insight back to them and freeing them up to attend to each child’s needs. However fanciful this vision may sound, it is already playing out in classrooms all across the world.
It turns out that the battle of man vs. machine is a false dichotomy for educators and innovators.
The future of Education is unknown, but what is certain is that it will be led by the tech-empowered teacher; their potential may be limitless. Our chart is now complete:
The work of educators and innovators is to find out just how high that limit is.
I am a research mathematician turned educator working at the nexus of education, innovation and technology. Come say hello on Twitter or LinkedIn.
If you liked this article you might want to check out two of my other pieces: EdTech’s culture problem and Relationships before scale.
