The Future’s So Bright: AI In the Classroom
AFTER SPENDING A full day working with students one-on-one, most teachers have to grade tests and prepare lesson plans at home on nights and weekends. Is it any surprise that the 2018 Teaching and Learning International Survey by the National Center for Education Statistics found that educators in the U.S. feel more overworked than their counterparts in other nations?
Since adding more hours to the day isn’t an option, artificial intelligence may be the answer, by not only helping teachers with time-consuming tasks like grading but also by personalizing curriculums to help students succeed. “Educators—both K-12 and in higher education—are being asked to do more with less,” points out Dan Ayoub, general manager for Microsoft Education. “With AI, there is tremendous potential for true, personalized education and to streamline everything from research to grading to analytics.”
Companies have been trying to figure out how to harness learning with artificial intelligence for years. But it’s only been in the past decade that progress has been made. Algorithms are slowly being introduced in classrooms, with early adopters using them to both free up their time for actual teaching and to do new and exciting things with students.
Jennifer Jones, founder and CEO of Green Ivy Schools, a progressive and innovative school network in New York City, believes AI will differentiate learning over the next decade. “It’s already been heading that way over the past 10 years,” she says. “There are a handful of apps being used a decent amount across the country to monitor students’ progress and their reactions, pace, and depth of learning. Some teachers don’t always even realize they are using AI, even though they very much are.”
From established companies like Microsoft and Amazon to startups like Cerego and Ponddy, companies around the globe see the vast potential for using AI in the education sector. Here’s a look at a few of the interesting ways AI is changing our classrooms— with plenty of room to grow.
Picture a typical K-12 classroom: Twenty-plus students, sitting in desks, being taught the same lesson from a teacher. But it’s long been known that not everyone learns in the same way. Researchers are applying that thought to AI, trying to understand how different brains take in and process information and how to make learning easier and more fun for students. Microsoft HoloLens, for instance, is changing the ways STEM skills are taught. The 3D reality program allows students to fully immerse themselves in a subject rather than just learning the facts. “It’s the difference between viewing pictures of an aortic valve and being able to see how it actually works using mixed holographic reality,” says Ayoub.
Along with science and math, language is another area that’s ripe for AI-enabled learning. Ponddy, a San Jose–based startup that makes personalized language learning technology, uses natural language-processing techniques to identify key learning elements. “The traditional way to teach a language is to teach vocabulary, grammar, and language structure. But after the course, students can’t communicate—that’s a huge problem,” says Franz Chen, the company’s CEO. “With Ponddy if you know a word, then you are introduced to other words that are similar—egg and apple for breakfast, for example. It’s easier to absorb when things are put together in your context.”
Ponddy also helps teachers by creating curriculum faster. “Putting together a curriculum is hard and laborious and takes about one year per grade. But we can create curriculum in minutes instead of years,” says Chen. Right now, Ponddy is running a pilot program where students can create their own content and share it with peers, while a teacher acts as a guide. “The results will be presented in May, but it looks very good,” says Chen.
Cerego, a startup based in San Francisco, uses spaced rehearsal, where learning is spread out over time rather than in one session, to help students learn information faster and retain it more effectively. “Even a little bit of spacing—over a few days, rather than in one cram session—makes a massive difference in how much a learner remembers that material weeks or months later,” says CEO Paul Mumma. Like Ponddy, Cerego also helps teachers create lessons more quickly because AI is used to auto-generate multiple choice questions. “This allows teachers to focus on the actual teaching and training needed for students, especially at the one-on-one level, instead of spending time creating study sets,” says Mumma.
Making the Grade
One of the most time-consuming parts of a teacher’s day is grading papers and tests. But new platforms like Bakpax use technology to read students’ handwriting and auto-grade their schoolwork. Besides saving teachers hours, the program easily keeps track of how students are performing overall, and students get instant feedback and grades, which is a huge motivator. “Every teacher who has had experience with these basic AIs will tell you that it frees up time,” says Jones. “It is also delightful for students, because they love learning games. It’s very engaging for students, because they have a deep affinity for iPads.” In this day and age of instant gratification, students receiving quick feedback is something many companies are focusing on.
In the fall, Microsoft launched the AI-powered Presentation Coach in PowerPoint. The program records students as they present slides and offers a dashboard with immediate suggestions on how they can improve their presentations, including different word options and advice on pacing. Regardless of which algorithm is used, facial recognition systems generally compare the image taken in step 1 with a database in step 5. For example, according to Governing magazine, at least 39 states use facial recognition software linked to their Department of Motor Vehicles.
Tell Me More
Voice recognition software, like that found in Amazon Alexa, is also being introduced on campuses. In 2017, Arizona State University gave 1,500 engineering freshmen living in the school’s engineering dorm an Amazon Echo Dot. The Dots gave the students access to information and services for their new lives at school, and they could sign up for courses to learn how to build their own voice-user interfaces with Alexa. “With voice-enabled devices becoming more prevalent in our connected world, it only makes sense to bring these capabilities to our campus,” John Rome, ASU’s deputy chief information officer, told ASU Now.
While there is no question that AI will have a major impact on the education sector in the coming decade, the ways it will change teaching depends on whom you ask. Jones believes that AI will eliminate certain teaching positions, specifically within basic math, science, and language courses. “There will be no basic math instructor in 10 years,” she predicts. “That seems cold, but it’s not—it’s just not a great use of a human being. It’s better for all of us if those teachers are teaching something other than basic math.”
But all of the technology professionals we talked to did not agree. “Educational technology should be used as an aid to the current education and training systems, not in place of them,” says Mumma. “A robot or AI will never be able to provide that level of instruction, or the level of empathy, that comes with teachers.” Ayoub agrees. Nothing can replace the human interaction between teacher and student. It’s important to remember these technologies help educators make the most of their time, not replace them, so they can do more of what they love: teaching.