Changing the education paradigm with AI

“Artificial intelligence systems are being developed to act as teachers’ aides, leaving teachers more time to give individual students personalised attention”

The world’s current education paradigm relies on an outdated and inefficient model with one teacher helping an entire classroom of students master the same material at roughly the same pace in a predetermined amount of time. The model also assumes that student motivation is relatively constant, and roughly the same for each student, explains YJ Jang, CEO and founder of Riiid.

Even exceptionally talented teachers can overcome only some of these issues, but exceptionally talented teachers are, by definition, rare.

Today, however, artificial intelligence systems are being developed to act as teachers’ aides, increasing efficiency by automating routine tasks, leaving teachers more time to give individual students personalised attention, and maximising motivation through analytics that give them a better understanding of individual students’ comprehension and engagement. A human-centered AI system can deliver the right content at the right time to each student as they progress, minimising stress, frustration and boredom. When students do well, learning is fun.

This has benefits far beyond education. Learning success raises a child’s self-esteem, just as too much failure lowers it. Grades, which today often act as a public judgement of a child’s worth, can become private motivational goals, improving as a student follows the optimal learning path mapped for them by the AI system. Good self-esteem is a prerequisite for positive social engagement. Education is a fundamental building block for a healthy society.

Education has two fundamental building blocks: content and what scientists call knowledge tracing – keeping track of how much a student knows. The best education is one-on-one because the teacher understands where the student is in their learning journey and delivers the appropriate content. The goal is to feed the student content that is challenging, so the student doesn’t get bored, but not too challenging, so the student doesn’t get frustrated and quit.

But one-on-one teaching is expensive, available only to the affluent or to parents with enough time to homeschool their children. The rest of the world is stuck with the imperfect classroom model.

Since there are generally too many students for a teacher to keep close track of, assessment is done through periodic tests. The result is a cycle of slack boredom and intense cramming, the inefficient rhythm of most people’s secondary education.

An AI system can act as both personal tutor and teacher’s aide, delivering the right content to the student, while keeping the teacher informed of each students’ progress. The teacher can then spend more time helping individual students.

The technology is still in its infancy, but the science behind it is mature and proven to work.

For example, test-publishing or preparation companies and AI for education solution providers like us are also actively working together to incorporate AI technology into assessment and lead the transformation in education. The vision is not just simply to prepare students for tests but work to have AI integrated in students’ formational education. AI systems can tell students, teachers and parents how a student is doing – what they are strong at and where they need improvement and support – at any point, in real time.

What we need now is widespread adoption and technology needs to penetrate into public education. It is already happening without you realising it.

Montour School District in Coraopolis, Pennsylvania is running a pilot program with Carnegie Mellon University in which software sends real-time information about student’s math progress to computer glasses worn by the teacher. The teacher sees a smiling emoji over the heads of students who are doing well and a frowning emoji over the heads of students who are struggling. With a tap of the finger, the teacher can see which problem the student is working on.

Mesquite, Texas’ school district, meanwhile, is partnering with companies to launch a tool that tracks student performance to help teachers create individualised learning plans, recognizing that they need to move beyond standardised learning and address the uniqueness of each child.

Policymakers, teacher associations, school districts and individual schools should make efforts to understand how the technology works, and companies that use AI in education must be proactive in explaining it. What education needs is a serious dialog among these stakeholders to shift the paradigm. It’s time to teach with AI.

About the author: YJ Jang is CEO and founder of Riiid, an A.I. education company that builds systems to power learning.