top of page

Machine Learning: How EdTech is Transforming Schools


Imagine this: no homework, no report cards, no classrooms. It’s often a shared dream among young students. But here’s the thing: this is no dream. In fact, this is becoming the reality for more and more young people.


As originally reported by The Economist, education technology, or edtech, is transforming how a child goes to school. Mach

ines have overtaken every aspect of our lives and that includes how we learn.

Take ten-year-old Amartya. He attends Khan Lab School (KLS) in Mountain View, California. He says he is good at “maths” but admits his writing could use a little work. So, what does he do? He practices his grammar online, books an online tutorial with an English teacher and later he will email a correspondent for further help.


This sort of standard is normalizing. Students no longer need to spend all day in a classroom to be taught. They do not need to divide themselves by age, share a common space or live by a schedule to call it education. Instead, schools like KLS push students to pursue individual goals.


Khan students use software built by in-house developers to take tests and watch video lessons. Half the teachers act more like tutors and life coaches. They are still integral to the academic work but they have more time to mentor pupils on character traits like curiosity and self-awareness. The machines handle the grunt work.


Information technology has reshaped our lives in ways we have yet to fully appreciate. It’s only natural that this would permeate into our education systems. According to the research firm Technavio, the combined value of edtech markets in both North America and Europe is set to grow from $75 billion in 2014 to $120 billion in 2019.




The forms of research where the brunt of this money is funneled are two-fold: Artificial Intelligence (AI) and psychology. Basically, AI lets machines learn the student and personalize how they study. Research drawing on psychology and cognitive science provides the practical insight into the “science of learning.”

The late American psychologist Benjamin Bloom describes it as the way to overcome the failings of the factory model functioning today. It’s known as ‘adaptive learning’ software or ‘machine learning.’ Essentially, AI allows a computer to pick up on patterns they were not actually programmed to notice. For example, Mindspark, an Indian company, utilizes this technology by drawing on a bank of 45,000 questions and the two million answers generated each day. With over a decade’s worth of student data, it’s developers can anticipate common mistakes and write code to diagnose those errors.

J-PAL, a group at the Massachusetts Institute of Technology (MIT) looking for evidence of what actually works to alleviate poverty, reviews dozens of randomized controlled trials involving edtech. One study looked at an after-school program in India that used Mindspark for 4.5 months. They found that the progress made in language and math by those students was greater than almost any study of education in poor countries.

It was also achieved for a fraction of the cost of attending a government-run school.


Now, obviously edtech is not perfect. It can’t really improve the argument in a history essay or give recommendations on how to be more humorous in drama class. But it can help teachers facilitate assessments in these different subjects. “Comparative judgement” algorithms can rank pupils, allowing the teachers more time to focus on what’s important: individualized attention.

This brings us back to KLS. Edtech gives students like Amartya more control over his learning, which, many experts argue, will fuel greater motivation. According to Facebook Founder and CEO Mark Zuckerberg, “personalized learning is way better.” But this does not mean traditional standards are out the window.

If schools can combine personalization and structure, it is hard to imagine a world where students don’t prosper. The question now is how education software will fit in to the inevitable future of learning.


6 views0 comments
bottom of page